IIIIIIIllllllliffz;w 0(de '9‘“ / IHIWJIHHIIIHllllli’lllllllllHIHIIUIIIIIIIHHIIIIII 31293 00762 0697 LIBRARY Michigan State University This is to certify that the dissertation entitled TWARD AN UNDERSTANDING OF ABUSE IN THE DATING RELATIONSHIPS 0F COLLEGE STUDENTS: THE RELEVANCE OF THE FEMINIST/ SOCIAL CONTROL ‘MODEL presented by Julia Anne Parisian has been accepted towards fulfillment of the requirements for Ph.D Psychology degree in KM / WM/ Major professor September 24, 1990 Date MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771 REMOTE STORAGE PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE 2/17 203 Blue FOR!“ S/DateDueForrns_20|7.indd - 09.5 TOWARD AN UNDERSTANDING OF ABUSE IN THE DATING RELATIONSHIPS OF COLLEGE STUDENTS: THE RELEVANCE OF THE FEMINIST/SOCIAL CONTROL MODEL Volume I BY Julia Anne Parisian A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1990 ABSTRACT TOWARD AN UNDERSTANDING OF ABUSE IN THE DATING RELATIONSHIPS 0F COLLEGE STUDENTS: THE RELEVANCE OF THE FEMINIST/SOCIAL CONTROL MODEL BY Julia Anne Parisian This study utilized a mail-out-mail-back survey technique to explore heterosexual college students' experiences with violence and abuse in their dating relationships. The final sample included 252 Juniors and Seniors (163 women and 89 men) from a large western university. The goals of the study were: (a) to explore the relationships between sexual (8A), non-sexual physical (NsPA) and psychological abuse (PsyA) in dating relationships; (b) to examine gender differences; and, (c) to assess the relationships between abuse and a number of attitude and dominance variables which the Feminist/Social Control Model views as important. Over 44% of the women and 36% of the men reported experiences with either SA or NsPA in their dating relationships. The three types of abuse were found to be significantly related to each other, displaying important similarities and differences in their relationships to other variables of interest. It was concluded that, at a minimum, future studies should adopt an expanded view of "dating abuse" which always includes the assessment of 99gb SA and NsPA. Although men and women did not report significantly different levels of involvement (as either victims or aggressors) with NsPA, men did report inflicting more and sustaining less SA in their dating relationships. In addition, the emotional, physical and relationship consequences of dating abuse were found to be mere severe for women. Specifically, women were found to lose power over more domains of relationship decision-making as they reported increased involvement with dating abuse. They also suffered more moderate and severe injuries from NsPA experiences and reported much more emotional distress (and injuries) as a result of their experiences with SA. Some results of this study were consistent and others inconsistent with the predictions of the Feminist/Social Control model of interpersonal violence. The findings regarding the relationship between dating abuse and patterns of male dominance in decision-making, however, provided strong support for this model's perspective and cannot be explained within the context of other existing models of intimate violence. To John Jeppesen iv ACKNOWLEDGMENTS The ‘doing' of this study was influenced and aided in many ways by many people. I wish to thank those who've been most helpful, without whom I could not have accomplished the task. First among them is my dissertation committee chairperson, Dr. Bob Caldwell. His consistent guidance, support, availability and (especially) patience were crucial to my continued progress through this endeavor. He is a gifted teacher and mentor who has always maintained high standards, great faith, and the amazing and rare ability to encourage me to grow in the direction of my own interests. I owe him a lot. I also wish to thank the other members of my committee, Dr. Ellen Strommen, Dr. Dozier Thornton and Dr. Carol Mowbray. Though I'm sure I tested their patience at times, their thoughtful feedback was greatly appreciated and improved the quality of this project immensely. I am also grateful to Dr. Jane Moorman, the director of The Counseling Center at The University of California at Berkeley, and to Mim Brown, also on the staff there. Without their help, I could never have accomplished the political, bureaucratic and administrative aspects of this project. I am grateful, also, to the many students who took time out of hectic schedules and shared their sometimes painful experiences with me with such candor. The person who has been closest to the center of this cyclone has been my husband, John Jeppesen. My thanks and love go to him for teaching me so much about the meaning of "support" during the past several years -- standing by through long months of my absence and distraction, through the ebb and flow of my enthusiasms and despair, through the general havoc that this process has wreaked on our personal, social and economic lives. His encouragement and tenacious belief in me have carried me through many of the rough spots. My dissertation support group buddies (east, west and in between) got me through the rest of the rough spots. For their inspiration and friendship, my appreciation and love go also to Drs. Leah Gensheimer, Sharon Green, Andrea Solarz, Sully Sullivan, Sara Wood-Kraft, Polly Fretter and Cheryl Olson. They added a lot of light to the process. vi TABLE OF CONTENTS Contents Page LIST OF TABLES ......................................................... x LIST OF FIGURES ...................................................... xii INTRODUCTION ........................................................... 1 REVIEW OF THE LITERATURE ............................................... 7 Feminist/Social Control Model of Aggression ...................... 7 The model .................................................. 7 The model’s relevance to the study of dating violence .................................................. 11 Theoretical Positions Reflected in Dating Violence Literature ...................................................... 26 Theoretical models ........................................ 26 Sexual Aggression ............................................... 41 Prevalence ................................................ 41 Methodological considerations in research on the prevalence of sexual aggression ....................... 45 Factors associated with experiences of sexual aggression ................................................ 53' Effects of sexual aggression and victimization ............ 65 Non-Sexual Physical Aggression in Dating Contexts: Prevalence ...................................................... 71 Factors Associated with Experiences of Non-Sexual Physical Aggression ............................................. 74 Introduction .............................................. 74 Relationship characteristics .............................. 87 Childhood experiences with violence ....................... 91 Attitudes toward violence ................................. 92 Sex-roles, gender identities, and beliefs about control ................................................... 95 Effects of Non-Sexual Physical Aggression ...................... 108 Gender Differences ............................................. 112 Criticisms of the Conflict Tactics Scale ....................... 119 The Relationship Between Sexual and Non-Sexual Physical Aggression ..................................................... 121 Psychological Abuse; ........................................... 124 RESEARCH GOALS AND HYPOTHESES ........................................ 127 Achieving a Comprehensive View of Dating Aggression ............ 129 Hypotheses, research questions and analyses .............. 131 Assessing the Impact of Attitudes and Patterns of Dominance ...................................................... 134 Sex-role stereotyping - Discussion and hypotheses ........ 134 vii Table of Contents (cont’d.) mm; m e Acceptance of Aggression - Discussion and hypotheses ............................................... 135 Resources and dominance in decision-making: Discussion and hypotheses ................................ 137 Assessing the overall impact of attitudes, resources,and dominance - Hypotheses ..................... 140 Gender Differences - Evaluating the Performance of the Conflict Tactics Scale .................................. 142 METHODS .............................................................. 149 Subjects and Procedures ........................................ 149 Measures ....................................................... 154 Demographic and personal information ..................... 156 Measure of non-sexual physical abuse ..................... 156 Measure of sexual abuse .................................. 160 Measure of psychological abuse ........................... 162 Psychological and physical status variables: The Brief Symptom Inventory ........................ 166 Other measures ..................................... 170 Resource variables ....................................... 171 Relationship variables ................................... 178 Attitude variables ....................................... 178 Decision~making/power variables .......................... 182 Additional variables ..................................... 184 RESULTS .............................................................. 185 A Comprehensive View of the Phenomenon of Dating Aggression....195 Comparing the Feminist/Social Control model’s relevance to each form of DA ............................. 200 Emotional consequences of experiences with SA, NsPA, and PsyA ........................................... 226 Physical consequences of SA and NsPA ..................... 229 The relationship between DA experiences and measures of psychological symptoms and social adjustment .......... 230 Assessing the Impact of Attitudes and Patterns of Dominance....241 “',_¢;===xc5ex role stereotyping and dating aggression .............. 241 Acceptance of aggression and adversarial beliefs ......... 246 Resources and dominance in decision-making ............... 250 Assessing the Over All Impact of Attitudes, Resources and Dominance .................................................. 267 Relevance of the Feminist/Social Control model to understanding men’s abusive behavior ..................... 268 Relevance of the Feminist/Social Control model to understanding women’s abusive behavior ................... 272 Gender Differences ............................................. 275 Comparing the CTS and The New Violence Measure ........... 277 DISCUSSION ........................................................... 281 Achieving a Comprehensive View of Dating Aggression ............ 281 The Conflict Tactic Scale and The New Violence Measure ......... 286 The Relevance of the Feminist/Social Control Model ............. 299 Conclusion ..................................................... 324 viii Table of Contents (cont’d.) Contents EASE REFERENCES ........................................................... 333 APPENDIX A ........................................................... 345 ix LIST OF TABLES Table Number and Captjgn Page Number 1 Men’s Experiences with Sexual Aggression: A Summary of Findings ........................................................ 54 2 Women’s Experiences with Sexual Aggression: A Summary of Findings ........................................................ 56 3 Men’s Experiences with Non-sexual Physical Aggression: A Summary of Findings ............................................. 75 4 Women’s Experiences with Non-sexual Physical Aggression: A Summary of Findings ............................................. 78 5 Experiences with Non-Sexual Physical Aggression: A Summary of Findings (Gender not controlled in analyses) ................. 81 6 Items From The Conflict Tactics Scale ............................. 85 7 Distribution of Sex, Class Standing and Race in the Final Sample ......................................................... 155 8 Descriptions of Dating Abuse Variables ........................... 157 9 Item/Scale Correlations for the Inventory of Psychological Abuse .......................................................... 164 10 Inventory of Psychological Abuse Inter-scale Correlations ........ 167 ll Descriptions of Psychological and Emotional Status and Injury Variables ...................................................... 168 12 Descriptions of Resource and Relationship Variables .............. 173 13 Item/Scale Correlations for the Resource Items (Standardized Scores) ........................................................ 176 14 Item/Scale Correlations for the Resource Difference Variables (Standardized Scores) .......................................... 176 15 Descriptions of Attitude and Decision-Making/Power Variables ..... 179 16 Descriptive Statistics for Dating Abuse Variables ................ 186 17 Descriptive Statistics for Psychological/Emotional Status and Injury Variables ............................................... 188 18 Descriptive Statistics for Resource and Relationship Variables...l90 19 Descriptive Statistics for Attitude and Decision Making/Power Variables ...................................................... 191 20 Men’s and Women’s Rates of Inflicting and Sustaining Sexual and Non-sexual Physical Abuse .................................. 194 21 Correlations Between Non-Sexual Physical Abuse, Sexual Abuse, and Psychological Abuse Measures: Women’s and Men’s Data ....... 196 22 Correlations Between Psychological Abuse Scales and Other Dating Aggression Scales and Variables - Victimization Data....199 23 Inter-correlations of Predictor Variables Used in Discriminant Function Analysis - female Respondents ......................... 201 24 Inter-correlations of Predictor Variables Used in Discriminant Function Analysis - Male Respondents ........................... 204 List of Tables (Cont’d.) abl 25 26 27 28 29 3O 31 32 33 34 35 36 37 38 39 4O 41 42 43 44 45 46 Numb n ai asylum: Mean Score on Predictor Variables for Women in Various Offender Groups ................................................ 209 Mean Score on Predictor Variables for Men in Various Offender Groups ................................................ 211 Discriminant Analysis Summary: Non-Sexual Physical Abuse - Offender Groups, Nomen ................................. 214 Discriminant Analysis Summary: Non-Sexual Physical Abuse - Offender Groups, Men ................................... 215 Discriminant Analysis Summary: Sexual Abuse - Offender Groups, Women ................................. 216 Discriminant Analysis Summary: Sexual Abuse - Offender Groups, Men ................................... 217 Discriminant Analysis Summary: Psychological Abuse - Offender Groups, Women ....................................... 218 Discriminant Analysis Summary: Psychological Abuse - Offender Groups, Men ......................................... 219 Predictor Variables’ Correlations with the First Discriminant Functions for Each Type of DA ..................... 222 Results of Repeated Measures Analysis of Variance Comparing Men's and Women’s Emotional Reactions to Three Forms of DA ..... 228 Relationship of Non-sexual Physical Victimization Experiences to Psychological and Social Adjustment Variables ............... 232 Relationship of Sexual Victimization Experiences to Psychological and Social Adjustment Variables ............... 233 Relationship of Psychological Abuse Victimization Experiences to Psychological and Social Adjustment Variables ............... 234 Relationship of Non-sexual Physical Abuse Offender Experiences to Psychological and Social Adjustment Variables ............... 237 Relationship of Sexual Abuse Offender Experiences to Psychological and Social Adjustment Variables ............... 238 Relationship of Psychological Abuse Offender Experiences to Psychological and Social Adjustment Variables ............... 239 A Comparison of the Mean Rates of Inflicted Dating Aggression Across Groups Defined on the Basis of Sex Role Difference #1...244 A Comparison of the Mean Rates of Inflicted Dating Aggression Across Groups Defined on the Basis of Sex Role Difference #2...244 Results of Analysis of Variance - Inflicted Sexual Abuse By Resource Balance, Decision Making Power Ideals, and Sex ..... 253 Results of Analysis of Variance - Total Actual Amount of Power Over Relationship Decisions By Combined DA Victimization Group, Level of Relationship Involvement and Sex ............... 258 Results of Analysis of Variance - Ideal/Actual Difference Scores (Total Scale) By Victimization Group and Sex ........... 261 Correlations of the CTS-based measure, The New Violence Measure, Dangerousness, and Injury (Dichotomous) with Brief Symptom Inventory, Emotional Impact, and Actual Decision- Making Power Scales ............................................ 278 xi LIST OF FIGURES Figure Number end Captieg Page Number 1 Riggs’ and O’Leary’s (1989) Model of Contextual or Background Variables Associated with Courtship Aggression .................. 29 2 Riggs’ and O’Leary’s (1989) Model of Situational Variables Associated with Courtship Aggression ............................ 31 3 Comparison of the Emotional Reactions to Victimization Experiences Across Three Types of Dating Abuse ................. 228 4 Rates of Inflicting Non-Sexual Physical Abuse Among Students with Differing Levels of Relationship Involvement and Differing Levels of Acceptance of Violence in Intimate Relationships .................................................. 249 5 Mean Rate of Inflicting Sexual Abuse for Groups Defined by Resource Balance, Power Ideals and Sex ......................... 254 6 Mean Actual Relationship Power Scores for Groups Defined by Level of Relationship Involvement and Sex ...................... 259 7 Mean Actual Relationship Power Scores for Groups Defined by Level of DA Victimization and Sex .............................. 259 8 Mean Total Difference Score for Male and Female Victim Groups....262 9 Mean Scores for Desired Power in Decision-Making for Students Who Do and Do Not Inflict Abuse: Non-Sexual Physical Abuse ..... 264 10 Mean Scores for Desired Power in Decision-Making for Students Who Do and Do Not Inflict Abuse: Sexual Abuse .................. 265 11 Mean Scores for Desired Power in Decision-Making for Students Who Do and Do Not Inflict Abuse: Psychological Abuse ........... 266 xii INTRODUCTION The prevalence of violence between intimate partners has been well documented over the past two decades (Finkelhor, Gelles, Hotaling & Straus, 1983; Gelles & Straus, 1988). It is estimated that at least 1.8 million American women experience severe beatings annually (Straus, Gelles, & Steinmetz, 1980). In addition, 46% of American women will be sexually victimized at some time in their lives (Russell, 1984). Nearly 40% of these assaults will occur in a dating or marital relationship (Russell, 1984). The effects of this violence is obviously devastating. The toll to women's physical health due to the resulting injuries is profound. Various estimates suggest that between 20 to 50% of all women's visits to emergency rooms are due to domestic assaults (Stark, Flitcraft, & Frazier, 1979). Furthermore, assault by an intimate partner is one of the most frequent causes of violent death for women. Walker (1989) reported that over half of all women homicide victims are murdered by current or former intimate partners. In April of 1989 it was reported on a local television news program in San Francisco that domestic violence had recently become the leading cause of homicides in that city. 2 The increasing social concern about this problem since the mid-seventies has resulted in a substantial body of research focused on the causes and factors associated with domestic violence. In the past few years this concern has expanded to include the study of physical aggression in dating relationships. In 1981, James Makepeace published one of the earliest studies of courtship violence. He reported that over 20% of the respondents, who were undergraduate students at a midwestern university, had had at least one experience with physical violence in a dating relationship. This prevalence rate led Makepeace to conclude that courtship violence was "a major, hidden social problem" (p. 100). Numerous studies conducted since then have confirmed and expanded upon these early results. Subsequent studies have, in fact, tended to find even higher rates of dating violence. In reviewing the literature on the prevalence of courtship aggression among college populations, Riggs and O'Leary (1989) recently concluded that, "as many as 50% of college students experience instances of courtship aggression at some point in their lives and 30% of these students may currently be involved in aggressive relationships” (9. 55). The literature on sexual aggression corroborates this picture of violence and abuse in dating relationships. Evidence indicates that, for many young people, dating relationships - rather than providing an arena for romance or developing intimacy — provide, instead, experiences with 3 aggressive or coercive sexual interactions, and all too often, with violent attacks. In fact, it has been found that a significant proportion of all of the sexual aggression which is known to occur takes place in the context of dating relationships (Kanin, 1984; Ross, 1985; Russell, 1984). One recent national study of sexual aggression and victimization, which surveyed 6,159 college students from 32 different campuses around the country, found that over 25% of the women respondents had experienced the most severe forms of sexual aggression, an attempted or completed rape (Koss, Gidycz, & Wisniewski, 1987; Warshaw, 1988). Of those who were raped, 88% knew their attacker and 57% reported that the assault occurred on a date (Warshaw, 1988). In this same study, 25.1% of the male respondents reported having engaged in some sort of sexually coercive behavior with a female partner since the age of fourteen. 80 it is clear from the results of studies of both sexual aggression and of other forms of physical aggression in dating relationships that, at least among college populations, young people are at sugstantial risk for experiencing violence in their dating relationships. The goal of the current study was to gain further understanding of the phenomenon of dating aggression among college students at a large western university. The study addressed the experiences of both the aggressors and the victims of dating aggression, and males and females were queried both about inflicting (or initiating) violent acts 4 and about experiences with sustaining violence at the hands of their dating partners. This investigation attempted to expand upon previous work in the field in three areas. First of all, dating aggression was defined more broadly than in previous studies reported in the literature. The vast majority of previous studies have focused exclusively upon a limited range of physically aggressive behaviors which has excluded sexual aggression. It was felt that adopting an expanded view of dating aggression would result in a more accurate reflection and understanding of the nature of aggressive interactions in dating relationships. In this study, dating aggression was defined to include three components: (1) sexual aggression, (2) other forms of physical violence and, (3) psychological abuse, all of which were measured separately. The similarities and differences between these three types of abuse were explored, as were the relationships between them. In addition, the psychological impact of each was assessed. The second goal of this study was to explore the extent to which one theory of domestic assault, the feminist/social control model, is pertinent to the study of dating aggression among a college population. This theoretical perspective maintains that norms and patterns of male dominance in social relationships play a central role in maintaining aggression and violence in intimate relationships. This theory has not previously been applied in a systematic way to the study of dating aggression. The predictions of this theory of 5 intimate violence were examined by measuring three pertinent variables: (a) dominance patterns in decision-making in the relationships between the respondents and their partners, (b) patterns of resource distribution between dating partners (resources measured included perceived economic, scholastic and social assets) and (c) attitudes toward male-female relations. The extent to which these variables were associated with each of the three forms of dating aggression was examined. A more detailed discussion of the specific predictions regarding these variables is presented below, after a review of the literature. The third goal of this study was to explore, in greater detail than in previous studies, gender differences in experiences with dating aggression. The feminist/social control model maintains that the aggression in dating relationships is largely initiated by men as an attempt to achieve dominance and power over their partners. Female-to- male aggression, from this perspective, is seen as primarily self-defensive in nature, and it is expected to have much less serious physical and emotional consequences for men than male violence is expected to have for female partners. Previous research in dating aggression has relied exclusively on The Conflict Tactics Scale (CTS), a measure developed by Straus (1979) for use in studying domestic assault among married and cohabitating adults. This research has, for the most part, found no consistent gender differences in the number or types of inflicted or sustained 6 violent behaviors in dating relationships - that is, on the whole, women report inflicting as many, if not more, of the same kinds of aggressive behaviors on their partners as men report. But several authors in the domestic violence field have criticized the CTS, suggesting that it is insensitive to contextual factors that would yield a more accurate picture of gender differences (Dobash & Dobash, 1983, 1988; Saunders, 1988; Walker, 1984, 1989; Yllo, 1988). For instance, one criticism of the CTS has been that it does not allow for a distinction between self-defensive actions and other violent behaviors. One of the goals of this study, then, was to explore the differences between men's and women's experiences of dating aggression, with the criticisms of the currently employed methodology in mind. This involved: (a) augmenting the CTS with additional items which were designed to be more sensitive to the contextual features central to understanding gender differences, and, (b) exploring gender differences in the patterns of association between dating aggression and the measures of dominance in decision-making, of resource distribution between dating partners, and of attitudes toward male-female relationships, which were mentioned above. Before turning to a more detailed discussion of some of these issues, a review of the literature will be presented. First, the feminist/social control model will be discussed. Then the theoretical perspectives which have been more prevalent in the literature on dating aggression will be 7 reviewed. Finally, the literature pertaining to each of the following four areas will be presented in turn: (a) sexual aggression in dating relationships, (b) other forms of physical aggression, (c) what is known about the degree to which these two phenomena are related, and, (d) psychological abuse. Review of the Literature FeministlSociel Control Model of Aggreseion The model. The feminist/social control model's analysis of violence in intimate relationships is closely associated with feminist activists' attempts, during the past two decades, to address issues of violence against women. Their efforts have been responsible for the development of battered women's shelters and rape crisis centers throughout the United States and England. The emergent feminist model of interpersonal violence has subsequently had a significant impact upon the research done in the areas of domestic assault and sexual aggression during the past ten to fifteen years. This model has been articulated most clearly in the domestic violence literature where a number of authors have advocated this perspective in discussing issues related to battered women (Adams, 1988; Allen & Straus, 1980; Bograd, 1988; Brown, 1980; Dobash & Dobash, 1979, 1988; Gondolf, 1988a, 1988b; Pagelow, 1981; Straus, 1980b; and Walker, 1989). 8 Among the earliest and strongest proponents of this perspective were Rebecca and Russell Dobash, Scottish sociologists who have conducted research and written extensively on the subject of wife battering. In their book, Violence egeigst wives (1979), these authors asserted that wife battering does not stem from pathological, deviant processes centered either in the individuals involved or in their relationship, but instead develops out of and is maintained by widely accepted societal norms which promote male dominance over women in social relationships. Pagelow (1981), concurred that "traditional ideology" concerning appropriate male and female roles is the fundamental cause of men's violence against their wives and lovers. As Pagelow described it, this traditional ideology includes patriarchal assumptions about male dominance in family relationships. However, it goes beyond that, including a ”broad range of internalized beliefs" which shape the social relationships between men and women in nearly all aspects of society: "It is a ranking of human beings based on (expectations of) male superiority and female inferiority, and designates greater status and power for males than females, regardless of other attributes, skills, knowledge or accomplishments" (p. 40). This ideology is seen as (a) fostering expectations of male control and dominance over women in relationships and, (b) legitimizing men's aggressive attempts to achieve this 9 control when and if their relationships with women do not fit with these expectations. From this perspective, men's aggressive and violent behaviors in intimate relationships are seen as purposive, instrumental actions motivated by a desire to achieve or maintain dominance and control over their partners. Further, both the goals and the aggressive means are seen as being promoted by social norms which prescribe appropriate behavior for men and women. As Dobash and Dobash (1979) put it, "...men who assault their wives are actually living up to cultural prescriptions that are cherished in Western society -- aggressiveness, male dominance, and female subordination -- and they are using physical force as a means to enforce that dominance" (p. 24). Thus, the male culture is further seen as promoting and encouraging the use of aggression and violence among men both in general and more specifically as they are directed against women (Dobash & Dobash, 1988). One of the main tenets of the feminst/social control model, then is that male-to—female violence in intimate relationships is caused by dysfunctional social norms. It does not stem from deviance or psychopathology at an individual or relationship level. It is not ”abnormal", in that sense; but, rather, is systematically supported and condoned by the larger patriarchal social order. To account for individual differences in the expression of violence and aggression in intimate relationships, Pagelow 10 (1981) asserts that different individuals will adopt these patriarchal norms to varying degrees. As with all social norms, they will be strongly aspired to by some, therefore serving as important standards for behavior for those individuals, while others will not adopt the norms to the same degree. Social learning theory would predict that the extent to which a particular man accepts these patriarchal norms could be expected to be determined by a number of variables, including, the degree to which important role models in the his life have accepted these norms, the extent to which his current social group supports them, the outcomes of his past attempts to express these norms, etc. One of the necessary conditions for violence to occur in intimate relationships, then, is that the male accepts these norms of male dominance. But, in addition to that, violence would be expected only if and when this individual feels that his position of dominance is threatened (Gondolf, 1988a). This would be expected to occur when there are conflicts about control within the relationship. This would depend upon, among other things, the extent to which his partner also accepts the norm of male dominance. In addition, independent of any particular conflicts about control, his position of dominance might be threatened if there are imbalances between the partners' available social, economic, or personal resources which are contrary or obstructive to his goal of dominance (Allen & Straus, 1980; Brown, 1980). 11 A further prediction made by some proponents of this model is that violence will be used by a man only if he has no other means available to him to regain or achieve the position of dominance and control which he wishes to have. Thus, violence is seen as a man's ultimate way of regaining or achieving dominance in the relationship -- a method of last resort (Allen & Straus, 1980; Brown, 1980; Straus, Gelles & Steinmetz, 1980). Another important, and perhaps obvious, assumption made by the feminist/social control theorists is that men are the primary perpetrators of violence in intimate relationships, while women are the victims. It is predicted that, when women do engage in violent behavior in such relationships, they do so only, or at least primarily, in self-defense (Dobash & Dobash, 1979; Walker, 1989). The model's relevanee t9 the study of dating violence. As mentioned above, the feminist/social control model of intimate violence has had significant influence on research in the areas of sexual aggression and domestic violence. Of course it is not the only theoretical model which informs the work in these areas, but a substantial number of studies of sexual aggression and domestic violence have explored variables pertinent to the predictions of the feminist/social control model. As a result of these studies, a fair amount of evidence in support of the feminist/social control model has emerged. (This work, as it pertains to the central 12 questions in this study, will be reviewed in the sections below.) The feminist/social control model, has 39; been equally influential in guiding research in the area of dating aggression (apart from sexual aggression). Few studies of non—sexual dating aggression have taken a feminist/social control perspective in exploring issues or variables central to this model. A number of factors may account for this. First of all, the feminist/social control model was developed largely out of work with clinical populations -- with battered women who required shelter from their assailants or with couples who had come to the attention of legal, mental health or medical professionals. These clinical samples have consisted, for the most part, of victims and assailants who were married, or living together. In addition, the battered women in these settings have often experienced the most severe and lethal forms of violence in their relationships. The feminist/social control model, then, was derived in this specific context to explain and describe the phenomena of chronically, severely battered women. Some authors, have questioned the model's applicability to less committed relationships and to less severe forms of violence (Makepeace, 1989; Riggs & O'Leary, 1989). Even proponents of the feminist/social control model are not unanimous about its applicability to dating aggression. Dobash & Dobash (1979), for instance, are careful to distinguish between marital and dating relationships in their 13 discussion of the feminist/social control model. They maintain that the norms in support of male dominance and control, and, more specifically, of the violent and aggressive means of achieving that goal do not strongly shape male-female relationships until marriage. They state: It is still true that for a woman to be brutally or systematically assaulted she must usually enter our most sacred institution, the family. It is within marriage that a woman is most likely to be slapped and shoved about, severely assaulted, killed or raped. Thus, it is impossible to understand violence against women without also understanding the nature of the marital relationship in which it occurs and to which it is inextricably related (Dobash & Dobash, 1979, p. 75). Others who advocate the feminist/social control model of interpersonal violence take a different stance, though (Kelly, 1988; Ptacek, 1988b; Russell, 1984). These authors suggest that the pattern of male-to-female violence goes beyond the confines of the family, encompassing a wide range of sexually and physically abusive actions in all kinds of relationships -- all of which are seen as different expressions of basically the same social problem. From this perspective, incest, wife battering, stranger and acquaintance rape, and sexual harassment in the work place are all seen as stemming from a common underlying cause: Social norms that encourage male domination and aggression toward women. Male aggression toward dating partners would 14 be included here as well. Thus, these authors would argue that the feminist/social control model would apply as well to the phenomenon of dating aggression as it would to wife battering. Two related positions are also found in the literature pertaining more directly to violence in dating relationships. As the preceding discussion implies, the question of whether or not the feminist/social control model is applicable to dating violence hinges, in part, upon whether or not one views dating violence and wife battering as the same phenomenon. In regard to this question, on the one hand, Roscoe and Benaske (1985) maintain that they are. In their study of battered women they found that over half of the women who sought shelter services reported that they had experienced violence in dating relationships also. This led these authors to conclude that dating violence and marital violence are two phases of the same phenomenon. They suggest that both instances of interpersonal aggression be conceptualized as "relationship violence". On the other side of the issue are Makepeace (1989), Riggs & O'Leary (1989), and DeMaris (1987). These authors have all advocated for a separate theory of dating aggression, maintaining that the violence in these less committed relationships is qualitatively different from that which occurs in marital relationships. Makepeace (1989) states that the application of "‘borrowed' spousal abuse theories" 15 (p. 103) to dating violence is not adequate to the task of understanding the phenomenon. A recent review of the literature by Sugarman and Hotaling (1989) compared dating violence to that which occurs in marital relationships. These authors concluded that, while the evidence indicates important similarities between the two, it also indicates substantial differences. Among the similarities that they report is the fact that roughly similar lifetime prevalence rates have been found for dating and marital violence (excluding from consideration, for now, cohabiting couples, who have higher rates of violence than either of these two groups). They also find that similarities exist in the reasons or causes for violence given by both married and dating individuals in a number of studies. For instance, disagreements over alcohol or drug usage and jealousy have been found to be related to violence in both types of relationships (Sugarman & Hotaling, 1989). But a number of other explanations for violence have been found to be pertinent for married individuals but not for those who are dating. These include disputes over money, children, and the wife's employment. In comparing the risk markers which have been associated with violence in the two contexts, Sugarman and Hotaling (1989) identify three that are associated with both marital and dating violence: higher levels of male sexual aggression, higher stress levels, and lower income levels. They also 16 identify the following as variables that have failed to be consistently associated with violence in both areas: being the victim of violence during childhood, race, sex-role attitudes (which will be discussed in detail below), and, psychiatric symptoms. However, Sugarman and Hotaling (1989) also discuss a number of variables that have been consistently related to violence in either dating or marital relationships, but not in the other context. For instance they report that lower educational attainment, religious incompatibility, and having witnessed marital abuse as a child have been consistently related to experiencing marital abuse, but not to dating aggression. Also, low self-esteem and positive attitudes toward violence have been associated with violence in dating relationships, but not with marital violence. Although Sugarman and Hotaling suggest that methodological inconsistencies between the research in the two areas may account for some of these differences, their final assessment of the reviewed data is that, ”it is premature to conclude that dating and marital violence represent the same phenomenon” (p. 26). Thus, the empirical evidence is ambiguous concerning the equivalence of dating and marital violence. So it seems clear that concerns about dissimilarities between the two phenomena may be partially responsible for the fact that the feminist/social control model, derived so directly from work with clinical populations of battered women and their 17 assailants, has not been fairly considered by investigators in the dating violence area. However, enough evidence of some similarities between the violence in the two contexts exists to preclude an off-handed rejection of the domestic violence theories on the basis that marital and dating violence are qualitatively two totally separate phenomena. Another factor which has contributed to the rejection of the feminist/social control theory in the dating violence literature stems from the methodology employed to measure violence in the studies in this area. Virtually without exception, all of the investigations into dating aggression have utilized some version of The Conflict Tactics Scale (Straus, 1979) to assess violence. As mentioned earlier, this measure has been criticized for its lack of sensitivity to the contextual features of the violent interaction that are important in revealing and understanding gender differences.‘ One result of these weaknesses may well be that the CTS actually meek; gender differences in the experience of interpersonal violence. The apparent lack of gender differences in inflicting and sustaining dating violence, then - resulting, perhaps, from the exclusive reliance on the CTS - could lead to a premature rejection of the social control model, with its emphasis upon gender differences in the experience. ' A more thorough discussion of the problems and criticisms of the CTS will be presented in a section below. 18 As a case in point, Riggs and O'Leary (1989) advocate for the rejection of the feminist/social control model precisely because of "the large number of women who report aggressive behavior against their dating partners" (p. 56). Burke, Stets, & Pirog-Good (1988) make a similar argument in drawing conclusions from a review of studies of dating violence that utilized the CTS. They view abuse as ”a mutual problem of both sexes....because research finds no difference in the physical abuse rate by sex" (p. 74). But of course, since the CTS does not allow for a differentiation between self- defensive and offensive acts, nor take into account differential consequences of aggressive acts, it makes gender comparisons difficult, if not meaningless. Additional influences that may mitigate against a more serious consideration of the feminist/social control model in the area of dating aggression stem from the prevailing culture of professional and academic psychology itself, which is, of course, imbedded firmly within the larger cultural and social environment. Thomas Kuhn (1970) pointed out that, historically, the accumulation of scientifically derived knowledge has not progressed by means of a smooth, purely rational process. Instead, he maintains that the paradigms that have been dominant within a given field at a given time have had a powerful influence on the research questions that were asked, the methods of inquiry used, and, even, upon the ways in which the results have been interpreted and received. The French historian/philosopher, Michel Foucault (1972) took 19 this notion a step further with his concept of the "episteme”. Foucault (1972) describes the episteme as, "something like a world-view, a slice of history common to all branches of knowledge, which imposes on each one the same norms and postulates, a general stage of reason, a certain structure of thought that the men of a particular period cannot escape..." (p. 191). He maintains that the dominant episteme within a culture at a given time will act as a powerful determinant of the scope and content of the scientific activity carried out in all branches of science within that culture. The episteme, in contrast to Kuhn's notion of a paradigm, would be expected to be a more pervasive, generalized, and frequently unconscious and/or unexamined set of assumptions that provides a framework, for all individuals in a culture, from which to organize and perceive reality. It defines the acknowledged universe of possibilities and determines basic, shared assumptions about ”truth". Foucault maintains that the episteme will function as an essentially conservative cultural influence on all areas of scientific activity. Since the content of the episteme is a reflection of the larger social/cultural context, the direction of its influence is expected to support the existing power structure which shapes social relationships in the culture. This tendency of scientific activity to conserve and protect the status quo - as it is reflected in the activities of the social sciences - has been addressed by a number of 20 authors recently - especially in the field of community psychology (Albee, 1986; Bleier, 1988; Heller, Price, Reinharz, Riger & Wandersman, 1984; Rappaport, 1977, 1981; Ryan, 1976; Weisstein, 1971). Both Rappaport (1977) and Ryan (1976) discuss how the endeavors of social science, imbedded as they are in the prevailing social structure, are inevitably permeated by the majority norms and values of that system. This mitigates against the likelihood that social science will act as a force for social change. Ryan maintains that those who are in the position to assess and define "social problems" (e.g. politicians; tenured, funded social scientists; human service professionals) are also (by definition) in positions where they have benefitted from the prevailing system of rewards and punishments that shape social relationships. The self-interests of these groups will make it very difficult for them to define social problems in ways that would call for change in the very power structures from which they have benefitted. Rappaport (1977) maintains that in the event that analyses of social issues ge_cha11enge or question the status quo, they are branded by the larger scientific community as "political" or "value laden”. This criticism, of course, reflects the position that social science research can and should be value free - which, at least on the basis of the historical analyses of Kuhn (1970) and Foucault (1972), is a questionable assumption. Rappaport (1977), and a number of feminist theorists (Bleier, 1988; Bograd, 1988; Dobash & 21 Dobash, 1979; Hare-Mustin & Marecek, 1988, 1989; Yllo, 1988) concur that scientific activity is not, nor can it be, conducted in a value-free or context-free manner. These authors maintain that the impact of social norms and values on social science research is unavoidable, inevitable, and most problematic when it is unstated or denied. From this perspective, the analyses of social problems and the research activities which stem from them are inherently political and value laden. They are only labeled as being so (and criticized for it) when they challenge the existing social structure. The prevailing episteme that is currently shaping and influencing the social sciences generally, and the field of psychology in particular, mitigates against serious consideration of the feminist/social control theory of violence in intimate relationships. Two aspects of the episteme which are especially powerful in this regard, and which will be discussed separately below, are: (a) the traditional emphasis on searching for individually-based deficits and causes of social problems, both within the field of psychology and in the larger mainstream ideology; and, (b) the widespread acceptance of the male-dominant, hierarchical ordering of relationships in all social institutions ranging from the family to universities and government organizations. Rappaport (1979) states that, "psychology is full of paradigms of the person” (p. 19). He maintains that even innovative, community or environmentally-oriented approaches 22 to addressing social and psychological problems are always, ”seen through the eyes of a psychological-scientific community trained within a tradition, a set, a perspective, a world view, or if you will, a paradigm, which sees everything in terms of individuals and their adjustment to a single standard” (Rappaport, 1979, p. 19). This dedication to analyzing issues in terms of individually-based deficits and causes discourages the consideration of models, like the feminist/social control model of violence against women, which emphasize culturally-based causes of social problems (Sarason, 1983). Such models, since they are critical of the status quo, are threatening to those who have benefitted from it. They are less likely to be taken seriously or given as much attention. On the other hand, models which assume an individual focus are much more comfortable and fitting with most people's own private world views. They keep the responsibility for change centered clearly upon the individual exhibiting the "problem", instead of revealing social changes that might be required for the solution of the problem (Ryan, 1976). Psychologists, on the whole, are undoubtedly much more comfortable considering, prescribing or engineering changes at the individual level for a variety of reasons - some conscious and others not. What seems clear is that this bias is supported by the prevailing episteme which shapes the social sciences. The second aspect of the episteme which mitigates against the careful consideration of the feminist/social control 23 model is the pervasive, entrenched acceptance of the male- dominant, hierarchical ordering of relationships throughout all social institutions. 'This problem, as it relates to the practice of the social sciences and human service fields, has been discussed in recent years by a number of feminist researchers and theorists (Bograd, 1988; Brodsky 8 Hare- Mustin, 1980; Chesler, 1972; Dobash & Dobash, 1979; Frieze, Parsons, Johnson, Ruble a Zellman, 1978; Mowbray, Lanir, a Hulce, 1986; Russell, 1984; Yllo, 1988; Yllo & Bograd, 1988). The values and norms supportive of the hierarchical ordering of male-female relationships pervade this culture. They are reflected in the content of the traditional sex role stereotypes, the widespread male dominance in decision-making roles and predominant male control of the resources which define social power. They saturate every social institution, significantly influencing the relationships between individuals within each. To expect that the social sciences could operate independently of these norms and values would be naive and misguided. They shape the way that social issues are perceived (Hare-Mustin & Marecek, 1988), the ways that mental health and illness are defined and treated (Brodsky & Hare- Mustin, 1980; Chesler, 1972; Mowbray, Lanir & Hulce, 1986), the methods employed in conducting scientific investigations (Bleier, 1988; Bograd, 1988; Dobash & Dobash, 1979; Frieze, Parsons, Johnson, Ruble & Zellman, 1978; Yllo, 1988), the conclusions and inferences drawn from research results 24 (Frieze, Parsons, Johnson, Ruble & Zellman, 1978; Russell, 1984), and the extent to which those results are published and cited. All of these authors have detailed specific ways in which the patriarchal norms which are operant in all aspects of our society have acted to systematically bias the activities of social science. They detail the many ways in which our values, and, more importantly, our discomfort when these are challenged, have shaped our choices of theoretical models employed, the interventions used, the research questions asked, the variables measured, and the instruments used to measure them. These values have also biased the attention and emphasis given to certain results, and the frequency with which subsequent investigators do or do not address issues raised, etc. As in the case of the bias in favor of individually- focused paradigms (and, for that matter, of ell aspects of the episteme, in Foucault's (1972) view] these norms in support of male dominance and aggression act in a conservative direction. They present strong obstacles to the acceptance of (or serious consideration of) theoretical positions which are critical of the underpinnings of the dominant social order. The feminist/social control model of male violence against women is, of course, just such a model. The issues which are central to the feminist/social control model -— the norms of male dominance over women in social relationships and of male aggressiveness -- are at the 25 core of our accepted social fabric. This model calls into question the nature of the most personal and intimate aspects of our daily lives - our relationships with family and friends. It forces us to examine the meaning and impact of the norms and values which determine the communication patterns and power structures in all of our social relationships. It requires us to ask questions that are extremely uncomfortable to ask and to seek answers that we might not want to face -- answers that might encourage us to change basic aspects of our own lives. The feminist/social control model doesn't allow us to hide behind distinctions between "pathological” others and ourselves. Instead, it implies that we might all be responsible for doing something about the violence in our society. This is not an easy perspective to entertain. It is more comfortable to assume a gender-neutral, individually-focused model of ”dysfunction" to explain the phenomena of intimate violence. Such a model does not entail such a heavy price: It does not challenge us personally. It is likely then, that these aspects of the prevailing episteme have discouraged an adequate consideration of the feminist/social control model in the research on dating violence. It has only achieved influence in the areas of domestic violence and sexual assault, because the work in both of these areas has been infused with the perspectives and experiences of activists who are committed to counteracting these values and norms. Such activists are, 26 perhaps, more likely to be outside the dominant power structure - and therefore more impervious to its influence - than are social scientists. (Of course, they are influenced by different values and norms, which will shape their work just as strongly.) The study of dating violence, though, has been conducted more exclusively within the domain of traditional academic circles, without the involvement of community activists. In the following review of the literature pertinent to dating violence, evidence will be presented that suggests that the feminist/social control model has not yet received a fair test in this research and that the recent calls to dismiss it from consideration (Makepeace, 1989; Riggs & O'Leary, 1989) are unjustified and premature. But, before reviewing the literature pertaining to dating aggression, the theoretical perspectives which have shaped the research in this area will be discussed and the most recent and well-articulated theoretical model (Riggs and O'Leary, 1989) will be presented. Theoretical Positions Reflected in Dating Violence Literature Theoretical models. For the most part, prior research on dating violence has been descriptive in nature (Firestein, 1987; Riggs & O'Leary, 1989) and investigators have not explicitly stated their theoretical or value orientations. Most studies have employed multivariate correlational survey strategies with college student samples with the dual goals 27 of assessing the incidence of violence in dating relationships and discovering factors which are significantly associated with these experiences of violence. But most of these studies have not been clearly derived from a particular specified theoretical perspective. Dobash and Dobash (1979), in discussing a similar "atheoretical", multifactorial trend in the domestic violence literature, point out the pitfalls of such an approach. As discussed above, theories and values inevitably shape research questions, whether they are explicit or not. Among other things, they determine the particular variables which are measured and the measures themselves. In the case of research into dating and domestic violence, these theories are often unacknowledged and unarticulated. Different researchers operate from different biases and expectations -- which undoubtedly affect their results. Dobash and Dobash maintain that the accumulated effect of this approach in any field would be an aggregated knowledge base which is a morass of confusing, misleading results which are largely haphazard and incoherent. These difficulties are exacerbated in the dating violence literature by the fact that unfounded causative conclusions are sometimes drawn from the correlational data (see for example, Burke, Stets s Pirog- Good, 1989). This, of course, results in the possibility that the consequences of intimate violence are misinterpreted as its causes, adding more confusion to the field. 28 very recently Riggs and O'Leary (1989) put forth two theory-based models designed to provide a much needed theoretical foundation for the dating violence literature and to facilitate the understanding and prediction of this phenomenon. Their models are based on social learning theory and draw heavily upon the previous work in the field, incorporating many of the constructs which earlier investigations have found to be significantly related to dating aggression. Riggs and O'Leary present two separate models. One is intended to predict ghg will be violent in dating relationships (see Figure 1). This one focuses on what Riggs and O'Leary (1989) call "contextual" or background variables. This model contains seven constructs which these authors predict will increase the likelihood that a particular individual will engage in courtship aggression. These are: (1) prior models of aggression in intimate relationships, (2) childhood experiences of physical aggression from parents, (3) attitudes of acceptance toward the utilization of physical aggression, (4) psycho- or neuropathology, (5) arousability/emotionality, (6) personality characteristics including aggressiveness and impulsiveness, and, (7) the utilization of aggression in previous relationships (including relationships with siblings, friends, strangers, 29 3 Accept aggression as a response to ,J frustration/conflict Parental aggression V toward child ’ 7 \ / Prior use of aggression ’- ps LIYL A 4—: Courtsz aggression , 6 l //~‘"w r , I 5 4 /:// muemmq/ >1 / 4.. emoUonalIty Figure 1. Riggs' and O'Leary's (1989) model of contextual or background variables associated with courtship aggression. Note. From "A Theoretical Model of Courtship Aggression" by D.S. Riggs and K.D. O'Leary, 1989. In Violence in Dating Relationships: Emerging Social Issues (p. 59), edited by M.A. Pirog-Good and J.E. Stets, New York: Praeger Publishers. Copyright 1989 by Maureen A. Pirog-Good and Jan E. Stets. 30 etc.).2 Riggs and O'Leary maintain that all of these predictors will have direct effects upon courtship aggression except for psycho- and neuropathology which, they believe, "are related to courtship aggression only through their impact on more general levels of aggression” (i.e. factors "5", "6”, and "7” above). These authors' second model is intended to predict ghee or in what specific situations an individual who is prone to using violence in intimate relationships will engage in aggressive behavior (see Figure 2). This model incorporates a number of ”situational” constructs. Riggs and O'Leary (1989) describe the situational model as follows: There are five major predictors of courtship aggression in the second component of the model: (1) the expectation of a positive outcome to the aggression, (2) stress,:3 (3) the use of alcohol, (4) the partner's use of aggression, and (5) relationship conflict. Five constructs are related to relationship conflict (relationship problems, problem—solving ability, couple communication, . relationship satisfaction, relationship level/intensity) rather than to the aggression itself. (p. 63) 3' Numbers instead of letters are used to enumerate the elements of this model (contrary to APA guidelines) to be consistent with Riggs and O'Leary's figures presented in Figures 1 and 2. " In their discussion of ”stress“, it is clear that Riggs and O'Leary are referring to stressful events or stressors rather than levels of emotional distress. 31 j I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I .I I r ------------------- '1 I i I II I 1 | 3’ iil ommmwomuumme-¢1II I II '7 I I I So : 2 I I | Communication ' Stress : : I I I|I III I I I III 5o 5 Relstionehip pmblans Relationship conflict Courtship aggression n» Consequences llr I I i I) I . , , I 5b I I I : nummaammuwma ::: I I I : :. g I 50 RelationshIp __ 3 4 . ‘J : | : level [intimacy Alcohol use Panners aggression I ' I I I ' . I I. ____________________ .J g L ___________________________________________ .I Figure 2. Riggs' and O'Leary's (1989) model of situational variables associated with courtship aggression. Note. From "A Theoretical Model of Courtship Aggression" by 0.8. Riggs and K.D. O'Leary, 1989. In Violence in gating Relationships: Emerging Social Issues (p. 60), edited by M.A. Pirog-Good and J.E. Stets, New York: Praeger Publishers. Copyright 1989 by Maureen A. Pirog-Good and Jan E. Stets. 32 In addition, the situational model predicts that the consequences of one act of aggression in a dating relationship will have an important impact upon the likelihood of subsequent aggressive behaviors. As depicted with the broken lines in Figure 2, Riggs and O'Leary (1989) predict that this impact will be mediated through influencing (a) the outcome expectations for the utilization of aggression, (b) the subsequent consumption of alcohol, (c) the overall level of conflict experienced by the couple, etc. They maintain that exploring the consequences of aggression is an important, though complicated, matter. They point out, for instance, that one aggressive act may have multiple consequences for a given individual in a given instance. Also, it may have different consequences for the two partners. Furthermore, the same consequence may have very different meanings or values for different individuals. Riggs and O'Leary (1989) maintain that the background variables and the situational variables are interactive - especially in regard to the impact of the background variables (e.g. past role models of aggression and experiences with violence in previous relationships) upon the current outcome expectations for aggression (a situational variable). They delineate other specific points of interaction between the two models as well. In general, they suggest that the background variables "may serve to increase the chances that a conflict situation becomes aggressive” (p. 68). Stated differently, this view suggests that the 33 background variables may act to suppress or elevate an individual's threshold for violent behavior under conditions wherein the situational variables are operating to increase the likelihood of aggression (e.g. in a relationship with chronically high levels of conflict, at a time when the individual has experienced a number of outside stressors, when both partners have been drinking and the individual's partner has acted aggressively). Riggs and O'Leary (1989) have made an important contribution to the field in putting forth these two models of dating aggression. As stated above, the work in this area has suffered from the lack of a theoretical framework. These models provide a sizeable, empirically verifiable set of constructs, with specified relationships between them which can be helpful in guiding and organizing future research. The two models are particularly useful in distinguishing between individually-based, background variables on the one hand, and situation-specific variables that increase the likelihood of dating aggression, on the other. Such a distinction is useful in beginning to organize and make sense of the long lists of variables which have been found to correlate with dating violence. However, in spite of these strengths, the models suffer from a number of serious problems. First of all, the models lack conceptual clarity in a number of important respects. At the outset, Riggs and O'Leary (1989) do not define "courtship aggression", although 34 they appear to use the term to refer exclusively to physical violence. However, they do not indicate whether or not the models are to be applied to sexual aggression as well as to other forms of physical violence. While they do mention some studies of sexual aggression in their initial literature review, their subsequent discussions and review sections do not include studies of sexual aggression. They also do not make any mention of verbal aggression or psychological abuse, although these could also obviously be important components of ”courtship aggression". The models suffer from other conceptual problems as well. Some of the constructs included in the models are not well defined nor clearly differentiated from one another. For instance in the model of background or "contextual" variables (Figure 1), it is not clear exactly how psychopathology (construct 04), personality characteristics such as aggressiveness and impulsiveness (construct #6), and "arousability/emotionality" (construct #5) are distinct from one another. In their discussion of ”arousability— emotionality", Riggs and O'Leary (1989) suggest that other theorists have linked ”arousal or emotional responsivity" (p. 62) to aggression. But, by placing this construct in this model, as opposed to the model concerning situational variables, they imply that this is a stable person-centered characteristic rather than a situation-specific variable. From that perspective, it is difficult to see how such a 35 characteristic would be distinct from personality variables or psycho- or neuropathology. Aside from these conceptual and definitional issues, there are two other major problems with Riggs and O'Leary's (1989) theoretical position, when viewed from the perspective of the feminist/social control model. First of all, these models are exclusively focused upon the individual (or relationship) level of analysis, excluding from consideration any cultural or social conditions which might underlie the problem of dating violence. This is in direct conflict with the feminist/social control theorists who, it will be remembered, view male-to-female violence as a pattern that is 95;! understandable when viewed in light of the social and cultural forces which perpetuate it -- namely, the widely accepted norms supportive of male dominance and aggression in intimate relationships (Bograd, 1988; Dobash & Dobash, 1979). Since our theoretical perspectives shape our perceptions of the problem and, therefore, the interventions derived from them, Riggs and O'Leary’s exclusive attention on blaming (and then fixing) individuals is problematic. As discussed above, this focus is bound to result in more "comfortable", individually-focused remedies. However, if the predictions of the feminist/social control model are accurate, it will also result in misdirected interventions and, therefore, in a passive perpetuation of the social conditions which foster dating violence. 36 The second main problem with these two theoretical models is that they are gender undifferentiated. They make no distinctions between men's and women's experiences of violence in dating relationships. Instead, they imply that these two phenomena are expected to be similar, if not identical. Riggs and O'Leary (1989) maintain further that, even if this is not so, these models can nonetheless be useful in exploring any gender differences that might exist. But, such an assertion trivializes the importance of gender differences, and, therefore, of the potential role played by the social norms targeted by feminist/social control theorists as the underlying causes of violence in intimate relationships. It should be noted at this point, though, that the specific predictions of the feminist/social control model can be viewed as being consistent with social learning theory's explanation of aggression (Bandura, 1977, 1983). That is, the effects -- at the psychological and relationship levels - - of the social conditions of sexism, as well as of norms which are supportive of male dominance and violence in intimate relationships, could be explained by the constructs and processes contained in the social learning theory's explanation of aggression at the individual level. For example, widespread economically based sexism would be expected to result in men and women having differential access to resources, which are central to the ability to access and control important reinforcers and wield influence 37 in social relationships. This would be an important factor in understanding power differentials and dominance patterns in intimate relationships. If men have more economic resources, they would generally have more power to reinforce and shape the behavior of those they were intimately involved with, which would tend to give them more power in those relationships. Sexism, and gender differences might further be reflected in and reinforced by the existence of different patterns of contingencies, available reinforcers and/or different schedules of reinforcement for males and females exhibiting the same behavior. Aggressive behavior may be rewarded in boys and punished in girls, for instance. Similarly, dominance behaviors in social situations may be differentially reinforced for males and females. All of these factors would be expected to affect the relationships in an individual's immediate environment, thereby determining the learning experiences based on their own directly reinforced performances as well as those based on observing the experiences of important models in the immediate interpersonal environment (e.g. parents, siblings, and friends). In addition, similar patterns might be reflected in media images presented throughout the culture ~- on television, in magazines, music, fiction and textbooks, for example. To the extent that all of these sorts of differences operate similarly throughout large segments of the culture, they would be expected to result in consistent 38 gender differences at a societal level. These could be reflected in consistently different habit patterns for males and females, different outcome expectations for similar behaviors, different standards for self-evaluative reactions and self-goal setting, etc. The above is an extremely brief and incomplete discussion of some of the ways that social learning theory could be used to explain -- at an individual level -- the social forces that feminist/social control theorists see as central to fostering and maintaining violence against women in intimate relationships. The goal here is not to provide an exhaustive translation of the predictions of the feminist/social control model into social learning theory terms, but to make the point that the model's predictions are consistent with and can be explained by social learning theory's constructs and processes. Similarly, a number of the feminist/social control model's predictions could be subsumed under the constructs which appear in the Riggs and O'Leary models, which are based on social learning theory. For example, power struggles caused by the male's goals of dominance could be included under ”relationship conflict" in Figure 2. Other factors which are emphasized by the feminist/social control theorists are less easily incorporated in the models as they stand, but could be included by making minor additions. For example, dominance expectations, and gender role expectations could be added to 39 the model of contextual or background variables depicted in Figure 1. However, to do so would not address the two major problems with these models that were discussed above: (a) their exclusive attention to the individual level of analysis to the exclusion of the societal level, and, (b) the fact that they are gender undifferentiated. These, then, are the major, or ‘irreconcilable' differences between the feminist/social control model of violence in intimate contexts, and the explanations of that phenomenon put forth by adherents of social learning theory, like Riggs and O'Leary. Even though social learning theory can readily account for the gender differences and social factors which the feminist/social control theorists see as central to the phenomenon of male violence against women in intimate relationships, its lack of explicit attention to these issues tends to obscure and minimize their importance and focuses attention on individually-focused analyses and interventions. Hare-Mustin and Harecek (1988) discuss the consequences of obscuring and minimizing gender differences in psychological analyses of phenomena such as this, which they call, a ”beta bias”. They suggest that this is common in some areas of psychological theorizing and maintain that it serves to "draw attention away from women's special needs and from differences in power and resources between women and men" (p. 460). They go on to assert that, "In a society in which one 40 group holds most of the power, seemingly neutral (approaches) usually benefit members of that group..." (p. 460). Dobash and Dobash (1979) agree with this perspective. They discuss this problem as it relates to the domestic violence literature and assert that gender neutral treatments of violence in this context are, ”not specific enough and miss the point" (p. 20). They maintain that such approaches present skewed and distorted pictures of intimate violence. They further suggest that, "the more general and abstract the approaches to interpersonal violence become, the less useful they are in the understanding of violence" (p. 25). The Riggs and O'Leary (1989) models of dating aggression are examples of such an approach. They abstract the phenomenon of interest from the socio-cultural context which fosters it, thus ensuring that those linkages and causes will be obscured. In summarizing the review of the theoretical models put forth by Riggs and O'Leary (1989), the position taken here is that, while these two models make some very important contributions to the dating violence literature, they are seriously flawed in their inattention to the probable gender differences predicted by feminist/social control theory. Compounding this problem is the fact that these models omit from consideration the constructs which feminist/social control theorists pinpoint as crucial to understanding and exploring, not only gender differences, but the most fundamental causes and dynamics of violence in intimate 41 relationships. From the feminist/social control perspective, a valid model of dating aggression would have to address (and emphasize) the impact of the social norms which are believed to perpetuate intimate violence —- namely, the norms of (a) male dominance in relationships with women and (b) male aggression as an acceptable means of goal attainment in these relationships. One of the major goals of the present investigation was to assess the extent to which these issues are, in fact, related to the aggression which occurs in dating relationships. If they are important, an individually-focused pathology or deviance model, such as Riggs' and O'Leary's, only helps us to avoid confronting the 'social issues which may have to be addressed if the violence between men and women in intimate relationships is to be alleviated. Before discussing the specific research questions and methodology employed in this study, a review of the literature pertaining to the following areas will be presented: (a) sexual aggression, (b) other forms of physical aggression in dating relationships, and, (c) psychological abuse. Sexual Aggression Prevalence. Research with college students assessing the prevalence of sexual aggression (SA) has revealed that it is an extremely common phenomenon. In the multi-campus national study conducted by Koss et al. (1987) 53.7% of the college 42 women reported having experienced some form of SA since the age of 14 years. The mildest form of SA was "unwanted sexual contact", which included fondling or kissing ”subsequent to the use of menacing verbal pressure, misuse of authority, threats of harm, or actual physical force" (p. 166). This was reported as the most serious experience of SA by 14.4% of the women. The most severe forms of SA, attempted rape and rape, were reported by 27.5% of the women (attempted rape = 12.1%; rape = 15.4%). The men in that study were asked to indicate if they had perpetrated any forms of SA. Over one quarter of the men indicated that they had engaged in some form of SA since the age of 14, and fully 7.7% of them admitted to raping or attempting to rape a woman. This discrepancy between the rates of SA experienced by women and admitted to by men is consistent across studies. In discussing this difference, Koss et al. (1987) maintained that their data ruled out the possibility that a small number of sexually aggressive men accounted for the victimization of large numbers of women, since the average number of incidents per respondent reported by men and by women were equivalent. These authors hypothesized that either a large number of the women in the study were victimized by men who were not college students, and therefore were not represented in the sample, or that men systematically under-reported the amount of SA that they were involved in. In addressing this latter possibility, Koss et 43 a1. cited studies with convicted rapists which indicated that these individuals, in general, tend to perceive their coercive and threatening actions toward their victims as well within the bounds of mutually consenting sexual activity. Observations made by Kanin (1984), in his study of 71 self-identified, college student rapists (all of whom had committed rape in the context of dating relationships) expand upon this idea that men who perpetrate SA may misjudge and underestimate the threatening and coercive aspects of their own behavior. He noted that, even though his subjects often admitted to using clear and extreme levels of physical force, including holding the woman down and physically restraining her, they seemed somewhat surprised at the level of fear and submission expressed by their victims. In another study of the prevalence of SA in dating situations among college students, Muehlenhard and Linton (1987) found that 77.6% of their sample of 341 women reported having experienced some form of "unwanted sexual activity”, either in high school or in college. Unwanted sexual activity was defined on the questionnaire as "when the female does not want to engage in some sexual activity, and she makes this clear to the male either verbally or nonverbally, but he does it anyway. The unwanted sexual activity could be anything ranging from kissing to sexual intercourse” (p. 188). The results indicated that 14.7% of the women had experienced sexual intercourse against their will. Over half (57.3%) of the male respondents (N=294) in this study 44 admitted to forcing some sort of unwanted sexual activity with a partner, and 7.1% of them admitted to raping a woman. The different incidence of overall SA found in this study, as compared to the Ross et a1. (1987) results, is most likely due to measurement differences. The definition of unwanted sexual activity utilized by Muehlenhard and Linton (1987) could have included less coercive forms of SA than that used by Ross et al. (1987), which would account for the greater prevalence of SA reported by the former authors. The incidence of rape reported in the two studies is very similar for both men and women. In a separate study of college women at a major midwestern university, Koss (1985) found an overall rate of SA that was very similar to that found in her national survey. Over half (54.5%) of these women reported having experienced some sort of SA, ranging from the mildest form, which was defined as engaging in "sexual intercourse when they did not desire it subsequent to the use of extreme verbal coercion, insistent arguments, false promises or threats to end the relationship by the man" (p. 196) to the most severe forms, which included experiences which would fit the legal definition of rape. The incidence of reported rape (12.7%) was also comparable to that reported in previous studies, but the respondents in this study reported a higher incidence of attempted rape: 25.3% reported these experiences. warshaw (1988) reported another important and startling finding from the national survey described above. Very often 45 acquaintance rape involves repetitive incidents. The mean number of rape experiences reported by the women who had been raped was 2.02. Similarly, the men who admitted perpetrating rape reported an average of 2.29 incidents. So, it appears that the most extreme forms of SA are not only very prevalent among this population, but they may be a recurrent component of dating relationships for many college students. Methodological considerations in research on thg prevalence of sexual aggression. The prevalence rates of rape and attempted rape reported above are substantially higher (10 to 15 times greater) than those reported in official victimization studies (Ross et al., 1987). One such study, The National Crime Survey, conducted by the Bureau of Justice Statistics (1984; described in Ross et al, 1987), was based on telephone surveys of a random community sample of households. The authors of the study reported that rape was a rare occurrence - the most infrequent type of violent offense measured. Kass et a1. (1987), and Russell (1986) point out that the way in which questions pertaining to sexual aggression are asked is critical to achieving valid estimates of its prevalence. In the National Crime Survey, the word "rape" is used repeatedly in the questions and the first screening question used to indicate whether or not the respondent may have been raped asks if she has been ”attacked" in some way in the past six months. These questions assume, of course, 46 that rape victims define themselves as such and conceptualize their experiences with sexual aggression as violent attacks. Yet the data are clear that this is not the case. Studies have consistently found that only a fraction of the women who have experienced incidents of SA that are consistent with legal definitions of rape actually label their experiences as "rape" when asked outright (Koss, 1985; Ross et al., 1987; Lane & Gwartney-Gibbs, 1985; Russell, 1982). The proportion of victims who do consider their experiences to be "rapes” have ranged from only 27% (Koss et al., 1987) to 57% (Koss, 1985). Koss (1985) identifies the women who do not label their experience as rape as ”unacknowledged rape victims". These are women who will respond affirmatively to a question that describes a behavioral interaction that meets the legal definition of rape, but who do not respond affirmatively when asked if they have been "raped". Ross et al. (1987), found that when women were raped in the context of dating relationships, they were even less likely to label their experiences as such. Koss' (1985) study of 231 college women revealed that unacknowledged rape victims were more likely to have known their assailants prior to the assault. In fact, all unacknowledged rape victims in her study had known their assailants prior to the rape incidents. While many of the acknowledged rape victims also knew the offender, the proportion was less. In addition, the unacknowledged rape victims tended to have had longer and 47 more intimate involvements with the rapists than did those acknowledged rape victims who were attacked by men who they knew. warshaw (1988) reports that men who inflict rape and attempted rape upon women may be even less likely to define their actions as "rape”. In the national survey conducted by Ross and her colleagues which Warshaw describes, only pp; percent of the men who admitted that they had committed SA actions that conformed to the legal definition of rape defined their actions as such. So, obviously, it is not sufficient to simply ask women if they have ever been raped, or men if they have ever raped a woman to achieve a valid estimate of the prevalence of these occurrences. Russell (1982), and K035 and Oros (1982) have tackled this methodological problem with some success. These researchers have developed measures specifically designed to assess experiences with various forms of SA. Russell's measure is an interview guide intended for use in lengthy face-to-face interviews with women. It consists of a number of open-ended questions which ask about a woman's experiences of‘unwanted sexual encounters throughout her life. Russell maintains that it is extremely important to utilize face-to— face, open-ended interviews in researching issues concerning sexual aggression to maximize the validity of the data. Ross and Oros' (1982) measure, The Sexual Experiences Survey, in contrast, is designed as a paper and pencil questionnaire. Although it also measures a range of sexually aggressive 48 occurrences, it focuses more exclusively upon unwanted sexual intercourse per se than does Russell's measure. These two research instruments share a number of important characteristics. First of all, both of them include numerous separate questions inquiring about the occurrence of unwanted sexual intercourse, with contextual details varying from question to question. This is expected to increase the probability that instances that have occurred are, in fact, reported because the instruments provide multiple cues for memory retrieval. Secondly, although both instruments include one question which asks about "rape" outright (which allows for a distinction between those victims who acknowledge that they have been raped and those who do not), the majority of questions do pg; include the word "rape". Finally, both instruments are based upon a definition of sexual victimization and aggression that conceptualizes these phenomena as a continuum of behaviors which range from milder Ito more and more coercive and violent acts, culminating in what is defined, legally, as rape. This view, which has been advocated by Halamuth (1984) and others contrasts with the traditional view of rape as a dichotomous variable - it either had or had not occurred. A woman clearly had or had not been raped. A man clearly was or was not a rapist. And both phenomena were seen as being very rare occurrences, distinct from "normal" functioning. From this perspective, rapists are seen as pathological, disordered individuals who act abnormally. 49 The perspective reflected in the measures of Ross and Oros (1982) and of Russell (1984) is much more consistent with the feminist/social control model of sexual aggression. This model views sexually aggressive behavior as one expression of the prevalent social norms which condone and support violence and sexually coercive, abusive interactions between men and women. he norms which are believed to be particularly inent to sexual aggression include (a) widespread acceptance of violence in interpersonal relationships, (b) sexism, and (c) sex-role patterns that foster mutual misunderstanding and animosity in the social interactions between men and women in dating situations - especially in those interactions and communications concerning sexual / intentions (Goodchilds & Zillman, 1984). Sexual aggression,/// then, is not viewed as an ”abnormal" occurrence, but rather as a very widespread serious social problem which is prevalent to greater and lesser degrees in many "normal" relationships. The results of the studies which have utilized these instruments have yielded much higher prevalence rates for attempted and completed rape experiences among women than those reported in prior research. Russell (1984) found, in interviewing a random community sample of 930 adult women living in San Francisco, that 44% of them had had experiences of rape or attempted rape at some point in their lives. Koss and her colleagues (1987) found that 27.5% of their national sample of college women reported having experienced these 50 most serious forms of SA. Russell would suggest that the differences in prevalence rates found in these two studies stem from the different measurement strategies employed. She views the utilization of in-depth, in-person interviews as essential to achieving valid incidence rates. But the different prevalence rates could also be due to the different populations sampled (a community sample of adult women as compared with college women), or to other differences in the measures employed. Ross et a1.'s (1987) sample of college student averaged 21 years of age, whereas Russell's study utilized a community sample of adult women, who reported an average age of 43 years. The research instruments used in both studies queried respondents about whether or not they had "ever” had experiences with sexual aggression. Russell's respondents, who reported higher prevalence rates, were reporting about an additional 20 years of experiences, which undoubtedly would account, at least in part, for the higher rates found in this study. In addition, Russell's interview schedule was designed to assess all forms of sexual aggression - including incest and childhood sexual assault. Koss's et a1.'s study, on the other hand, was targeted specifically at experiences with sexual aggression in dating relationships. The items in the Sexual Experiences Survey (the measure used in the latter study) specifically exclude neither experiences with family members nor childhood experiences of sexual assault. Still, it is probable that, without items designed specifically to 51 assess these experiences (which were included in Russell's interviews), they would be under-reported by Ross et a1.'s respondents. This, of course, could also contribute to the lower prevalence rates found in Ross et a1.'s study of college students' experiences with dating aggression. So, it is not clear that the face-to-face interview methodology employed by Russell is the only way to achieve valid estimates of the prevalence of sexual aggression. But, what is clear is that the methodological approaches used by both of these researchers are far superior to those used in prior studies of SA when assessing the prevalence of sexual aggression. Another methodological consideration in this area of research concerns how one should conceptualize the different involvements of men and women in SA. Not surprisingly, men are viewed solely as the aggressors by most researchers, while women are seen as the victims. Some authors have pointed out that, while it is logically possible and conceivable that men could be sexually victimized (even raped) by women, the known incidence of such behavior is so low, that it is not important to explore that phenomenon further. There are, however, exceptions. Gwartney-Gibbs and her colleagues (Gwartney-Gibbs, Stockard, & Bohmer, 1987; Lane & Gwartney-Gibbs, 1985; Gwartney-Gibbs & Stockard, 1989) have taken an atypical stance in this debate. As mentioned earlier, most research on violence in dating relationships has focused only on physical violence which has 52 been defined without reference to sexual aggression. Gwartney-Gibbs and her colleagues have combined measures of both sexual aggression and other forms of physical violence in their studies of dating violence among college students. In research on other forms of physical violence, it has become customary to measure both inflicted and sustained violence for both men and women. Lane and Gwartney-Gibbs (1985) revised the standard measure of SA (The Sexual Experiences Survey, which was developed by Ross & Oros, 1982) so that it, like the measures of other forms of dating aggression, can be answered by men and women from the perspectives of being both the aggressor and the victim. Gwartney-Gibbs strongly advocates for this strategy, maintaining that female to male sexual aggression is a very real part of dating violence and that, in omitting it from study, an incomplete picture is drawn (personal communication, April, 1989). In the studies conducted by this group, although they hay; found female-to-male SA, it occurs very rarely. Lane and Gwartney-Gibbs (1985) found that the most common form of SA that women admitted inflicting was "Saying things you didn't mean in order to have sexual intercourse", which was reported by 2.5% of women. The percentages of women that committed the other types of SA measured in this study ranged from zero to 2.2%. It should be noted, however, that men reported a fair amount of sexual victimization in this study - although 'primarily of the milder forms. The percentages of men who 53 reported victimization ranged from zero to 12.8% for the various forms of SA. 80 there does appear to be justification for adopting the strategy advocated by Gwartney-Gibbs in studying SA among college students. Factors associated with experiences of sexual aggression. Most of the research which explores sexual aggression among college students has been purely descriptive in nature - simply aimed at defining the scope of the problem. All of the research has also been correlational, with very little longitudinal work done. Nonetheless, the more recent studies have attempted to identify factors or "predictor” variables that are associated with SA experiences for women and men as a first step toward trying to understand the causes and dynamics of sexual aggression. Summaries of some of these studies and the variables that have been found to be significantly correlated with sexual aggression are presented in Tables 1 and 2. Some of the studies reviewed have focused on women, some on men, and others have included both women and men. The results that are pertinent to men are presented in Table 1, while those regarding women are presented in Table 2. Studies which included both are listed in both tables. Several points of clarification are useful in understanding the information presented in Table 1. The measures of the SA variables are of two types: Halamuth's measure of self-reported Likelihood of Raping (LR) (Check a 54 .moa>uooeouu v.0; xoa Loco.u_maeu cmouc_oe o» >.ox.. egos use... Am .39. me 03:89; «5.3 cos—ecu 09.35 3 .53.... «to... 0...... 3 .>u..osxoa o.ossw muozou noesu.uue o>.uo>eoacou 0.0: cu >.ox.. egos one... an .u_uu.cooouco >o>esm no ao.gmco.uo.oa o.oeo»\o.es o~_eouuotogo o» >.ox.. egos ore... a~ neoco_eooxw ..aseoc dosxom a~mopv cox Aesop. no co.u¢eruoo .osxea can on >.ex.. eros et- cos o>.auouuoe xddesxem a. .eoeo a neon eue..ou oeu.. onerous a «now .u.ua_cowouce as case: ace cos eoozuon ea.zuco_uo.oe ou.eouoeeogo o» >.ex.. egos of. «not or: cox Ac. .cosoa «season ooco.o_> we uc_uaeooe egos eonsu_uue c.ouc.oe one. on: co: go .aoaxuooeoua odoe-xoa soco.u_oeeu oeoe c.ouc.es «not or: so: no .oaou we o>.ueoansu «cuss coseou enhance o» >.ox_. egos oe- oaea or: cox an .ouo to...oo co an outsourouc. .osxoo u¢e.w no; ease or: so: as .c.u.eo so no...saw L.ocu c. euro.o_> egos atone; ease oz: :0: an .ueooa o.ae so.) 3..» mouco.eo-xoa egos atone; once or: co: Ac n.>usue .cfl- 35¢ on :2: 3532.! 8.1.9.358 egos use use... 3.5. on: co: 3 0...: mo 3.30.. .ces eeeuo A.>.co eeoaoeeoue «toner zoom. on :2.» .283 c. .233- ..o 2.35.6 aces—8.: 28. tone. 9.39. 3.5- ae c2: $3.5m 385 32.93: on: cue .eozoesa .oooos .ocou.e eoxuco uses >>eog so someones auceo_oc. oua A. more a neon one..ou -o.~ a .uoxe.o .aaou .co.auoumu< oc_uoo so woman Locuo use «a oc_uu..wc. ueoaoe oo.o was» “.me_uo.> e. «tarot-n a.ogu sofa .ex.. egos son: et- eedox “a use aeoaaoeooe .ueocueaa e.egu sot; .ox.. egos «to soon so moc.aou «goon oo~.s_uo_> >..osxuu ace e>_uaeeouo >..osxoa oc.>o; aeoaou or: so: an cozy .>o>esm .cos ca use desxoa «to: ucmcmeuasu atone; use omen. .oucoeea egos amoeu.3 cue o>.aaorooe >ddesxom A. co.eee> ue.a_uo: one..ou on. .onn.u.>ocurozo Ax.ouoeoaoa pontoon; 38.9.. soc «.3 .aoexuooeoua o.or-xoa .oco.».uoeu Co uoo:..ox.na case: use co: Ammopv 53...... 3 .33.: 28. 2» 2...... to 8233.... .23... 8.9.3.2 8.. 2 2:82.33 «8:8 em 53...... a x85 no.no_to> uuoo new once a to: s< M mmwammuuuqlmmmmmw .. .. -O 1.132; a ..0 am. ..4 ..-«.- m- .- Le? a..-) a manna 55 .ceeoa oe- ceeu cease ones» so ac.uaouoo egos as ou peso» or. ..ex.. egos ee- .ecoaeoarouc. so oc_uoouue eros et- .ex.. egos ee- aeuoo .ox_. egos a. once: page go .ozoo.e >>ee= £33 a S :8 no... .8. 8 33.: to... 3 cs. 2.» :5 :82 5.. an «a an “N unc_ouo Aneuuauuo xddesuoe :o_e3 .8 35.3»- ces so_ss .aesooo ao.u.>.uuo .esxoa co: “hoe—v couc.4 (m co.=s co oeueu co e>_eu o» >.ox.. ages n. cos on» use» atone; cox AP ucoeem».u up eoo..ou wo~ a prosce.gos: .eoooo ouco.eo> ecu roe uc.ucsouuo e. seen so are >co cosy renown ecoaeoa on ucso. ores ao.ne_eo> rouo.uoea neua_. e>ono as» so eo.uoc.neou <... go .oeuco.eooxo .esxoe egos ueoaoe... Am .coeo: season ouce.o.> docoaeonrouc. we oc_uaoouo egos era... he .cosox neuron >u.._unog too-ego atone; on >.ex.. egos etc... An .eco.uo.ou dosxou c. logo 0» generous. ere oucecwsou so one..ooe use» ueonoe ou >.ex.. ores eta... a~ .>o>rsm . one. o» «cocooaossu e._con .esxom a~mopv co: sh nauseous any .omsoee .asxou «roe ouoeueeoeom (a e. ounces o:: cos «p .ooeo a neon one..ou mm. flowery gusseLe: .oaoa yo o>.ueoanse «ease coseou o>o..on on >.oz.. egos ego uc_ne¢ so uoo:..ox_4 gu.z no.3 co: Aw awe—non .eaor ou >.ox.. on 0a.. n.303 see so noog_.ox.aa co: rogue uegu o>e..on ou >.ox_. egos oe- ue.ao¢ yo noon..ox_4 go_: so.3 co: a. noueoooe.+.om euoddou can “snowy gusee.ox ae.no.en> (u no.3 nouo.oueou acou_w_ ._m oLno.to> .u00 now o. . e.z mmduauuuuqlqmmmmw .e.ueoo. a «Home II. ~ a .mi. Vt 56 .usou.us. o» no_na omen: menu noxuso H.>usue «eon. .onoo.o no.) uou-_uoano one so.u-.s.uu_> nusxeu so noose.noaxw a~ flamers sonata: .xoe «sou.sue< o>.uoz .xs so.o< .>e>ssm use “know. .no- eeu_n: .x~— no.seaa.: .xc— «guano "noses us.unoaon sees: so oeuso.nosxu .esxem sees: .naae.su.3 eon-usounoa .osono u.snue o» us_usouuo ono*a_u ease Co eusone>osa A. a~nopv eono a «sea ooennou ump.n a .uo>u_o .eaox .ausou.us. >m any on «composes neso.uose o>_uooos «nos unseen masons >w no.3 nuoo an .us.uu.> ease ueouonzosxoo use nsona >m ou-Louos sn uounoaot nausea». ou ooseuu_aon we a.o>o. amono_n A» .eosono.> «no: any ueusoaon oases: ones uoouonzosxoe no.= .anousoaao so eonu aes_eoo ueas auto» use ousono.> so a.o>o. euro. uouuoaon sauna >m son .oasOLo uou.s.uo.> osos< no .anouseaao no.3 >oos.us. no.ns seasons uousosou us.uu_> ones uouuonsosnues: .ueus.esu0a one». one es.uo_> ones uoouenzosxue .onousomvo on» no.3 uous.esaoo >.no.n .onon osmuuonsosxoe Laos one us_uo.> one. uouuonzosxoes: .onsonu uou_e.uu.> on» usoe< an ace . >m nu.: "ones .anosunon us.ouo.:osnoe - >m nosxoa Co oneness sou-ens ou uouonon one >m we a.o>e. nono.= as nu.: u>m ouoseuo: u>m £83.85 :3 .98 825.8; nuance on.» we come nousso> ou ueuonon one >m no one>on seno_= an xnnnsxem uoz ”mesons .onsouo xso soosuon ouo_usosop~.u m one. sees: awesome. usesaooe ou.»u.neuoonenu >u..osoanoa so ousu.uuo nonuo oz .~ on use: >o>rsm .asono uo~.s.uu.>-sos nouse.neaxu nesxem sees: sen» gonna) .osxoa nonon.. eno- usonos ue_uu_> ones uoouonxosxo< A. A~mopv .aono a neon euonnou pn~ Ammopv anon A. .so_aoenuuo us_uou mo ease» nenuo use nuon no ues.~ou 3 9.2235 209.... 3 38.: ones 03- o... 3 £323 on: case: 3 case). $023 .ossonu noon n.9nu uses- sneuues use. anodes «m s.ouesu on: sees: “N noose.nonxw nesxem “hoops season .ensoso neon n.0nu uses. so.uou_s.uu.> ones». use A~oopv .uono a neon sees: a .usouooum so_ueenuue odes ones nuns anodes caucuses so (m uu..as_ on: case: a. so so.ano> uo.».uox ooennou amp .enn.u.>osune:o a0.ge.ro> (a nu.: moo-nonnou «sou.~_s.. onno.so> oe.nsm e on a n.nus< anduuuuumqlqmmmum u.. flu . t .m:. . .a. mim ... (1541- . . u- .-, -mfl, in.uu;o3 a «Home Av ~.~l%.~. 57 .sosoa nonoo sonu sees: usosou nousu.uuo s.onu s. .oso.u.uouo «no. one: so.oo~.s.uo.> .osxoo uooso.sooxo on: sees: as .ueonuo so son» oouou (a so sou.o onos u.on sea atone. sees: .oousoaxo ous.os. uonu nooou >.so no os.xoo. son: .oouou nonoo ou uonossoo no cousoaxo ousnos. uos ou soowo one: sou-u (m an .sas on» sh uouo.u.s. sou.o once one: nouou (a on» «on» anodes soso: .N .onosonos us.uou sonuo no.3 son» nouou (a so season-a no.3 .m.m u.mv oouou so.ns nose. anodes sees: A. ..ooe .oaxoo acute...u up o.no..o> mafimmmumm0|4mmmmw 8.8: .32. soon... ouonnou can .o.»eoo. m manna 58 Halamuth, 1983; Halamuth, 1986); and more behavioral self- report measures of actual, specific SA behaviors that respondents have engaged in during some designated period of time. The most widely used measure of this type is Ross and Oros' (1982) Sexual Experiences Survey (SES). Malamuth's measure of LR involves asking male subjects, after they have read brief vignettes describing either stranger or acquaintance rapes, how likely they think it is that they would behave as the man in the story did if they could be assured that no one would know. Respondents rate their responses on a five point scale from "not at all" to "very likely". Subjects responding with anything other than "not at all" are considered to be expressing some likelihood of raping. Individuals' LR scores have been found to correlate with sexual arousal to rape depictions, acceptance of various rape myths, actual aggressive behavior expressed toward a woman in a laboratory setting (Check & Malamuth, 1983), and scores on K033 and Oros' SES measure (Malamuth, 1984). In various studies Malamuth and his colleagues have found that approximately 30% of male college students indicate some likelihood of raping a woman if they believed they would not to get caught (Check & Malamuth, 1983). One of the things that stands out from Table 1, especially when compared to the information about women summarized in Table 2, is that mgpy more attitudinal variables are consistently correlated with men's SA experiences than with women's. Four different attitudes have repeatedly been 59 associated with an increased probability of men behaving in sexually aggressive ways towards women: (a) Men who report more SA express more traditional sex roles than do other men (Check & Malamuth, 1983; Ross et a1, 1987; Koss & Leonard 1984; Huehlenhard & Linton, 1987); (b) they are more likely to endorse common myths that are held regarding rape (e.g., women really enjoy being raped; Koss et al., 1987; Koss & Leonard 1984; Malamuth, 1984; Huehlenhard & Linton, 1987); (c) they express greater acceptance of violence towards women in general (Koss et al., 1987; Koss & Leonard 1984; Malamuth, 1986; Muehlenhard & Linton, 1987); and, (d) they tend to see relationships between men and women in antagonistic ways, characterizing them as including more "game playing" than do other men (Koss et al., 1987; Koss & Leonard 1984; Malamuth, 1986; Huehlenhard & Linton, 1987). It should be noted that all of these attitudes found to differentiate between men who report inflicting SA in dating situations from those who do not are consistent with the feminist/social control model of sexual aggression described above. In addition to the consistent findings regarding attitudinal correlates with men's SA, a number of other factors have been found to be associated with the SA variables in multiple studies. Four studies have indicated that peer group influences may be important factors in understanding the SA inflicted by men. Koss et a1. (1987) found that men who reported raping also tended to report frequent sexually oriented talks with male peers. This is 60 consistent with Kanin's (1984) conclusions, in discussing the differences he found between 71 self-identified college ‘date rapists' and a control group, that the rapists were, "shown to be products of a highly erotic-oriented peer group socialization which started during the junior and senior high school years" (p. 98). He maintains that, as these men came to college, they chose male peer groups who would reinforce and support the values learned earlier in adolescence. "Sexual conquest, as a result, becomes intimately associated with their feelings of worth" (p. 98). (Kanin's study did not report information regarding statistical significance, and so is not included in Table 1.) Two other results have been reported in multiple studies. Three of the studies have found that men who engage in SA are more sexually active than other men. Koss et al. (1987) found that men who admit rape had their first sexual experiences at an earlier age than other men. And, two other studies have found that men who report higher levels of SA also report more current sexual activity than do other men (Kanin, 1984; Malamuth, 1986). And finally, drug and alcohol usage have also repeatedly been found to be correlated with higher reports of SA among men -- heavy consumption prior to incidents of SA have been reported (Kanin, 1984; Koss et al., 1987; Muehlenhard & Linton, 1987), as well as higher levels of alcohol consumption in general (Koss et al., 1987). The results of studies pertaining to women's experiences with sexual aggression and sexual victimization are 61 summarized in Table 2. These studies reveal a different pattern of results. One of the most striking differences between the results regarding women's experiences with SA as compared to those concerning men is that there are no consistent patterns of findings which are replicated in multiple studies. While significant correlations between seven different variables and SA levels were found in two or more studies of men's experiences, there are no replicated findings in the studies of women's experiences with SA. Of course, fewer studies are included in Table 2 than in Table 1 (four as compared with seven). But three of the four studies of women's experiences also included men. And in these three studies alone, five significant results pertaining to men's experiences were replicated across multiple studies. Given the lack of replication, the results in Table 2 must be seen as tentative. Still, several are worth noting. First of all, the study conducted by Gwartney-Gibbs et al. (1987) suggests that peer group influences may be important to consider in understanding women's experiences with sexual aggression. This study, as discussed above is one of the few that conceptualizes women as being both perpetrators and victims of sexual aggression. These authors found that experiencing higher levels of SA from dating partners was associated with reports that peer groups included men who inflicted and other women who sustained sexual aggression in the context of dating. The same was true for women who reported inflicting sexual aggression on their partners. 62 These authors-take a social learning perspective on courtship violence and conclude that aggressive, mixed-sex peer groups may provide influential contexts in which sexual and other aggressive dating behavior is learned. Another interesting contrast to the findings regarding men's experiences with SA is that, for women who experience sexual victimization, few attitude differences emerged that distinguished them from other women. Two attitudinal differences were found to be significant, however. Ross (1985) found that rape victims expressed more liberal sexual values than non-victimized women. The other significant finding regarding attitudes was reported by Muehlenhard and Linton (1987) who found that women who reported higher levels of sexual victimization also reported less traditional attitudes towards women. Since neither of these findings have been replicated, caution must be exercised in interpreting them. And, as is always the case with correlational studies, it is possible that both of these attitudes may result from sexual victimization. Or, it may be the case that women with these attitudes are more likely to reveal SA experiences. A number of the other significant correlates of sexual victimization in Table 2 could be perceived as variables which define situations that place women at high risk for experiencing sexual victimization. The feminist/social control model of sexual aggression, as mentioned earlier, maintains that (because of certain unhealthy social norms) 63 the tendency toward expressing SA toward women in dating situations is relatively widespread among men in the general population. Studies of the prevalence of SA have confirmed this much. If that is true, then it would follow that the more intimate partners a woman has over the course of her dating career, the more likely it is that she will become involved with one or more men who are sexually aggressive or abusive. That is exactly what Ross (1985) found in her study of 231 college women - that the more sexual partners a woman has, the more likely she is to experience sexual victimization. Koss also reports that higher levels of sexual victimization are related to beginning sexual experiences at an earlier age, which could also result in a woman having more sexual partners, and in turn, being exposed to more risk of SA. Other variables which have been found to be significant correlates with rates of sexual victimization for women describe dating contexts that tend to increase the risk of SA experiences: Muehlenhard and Linton (1987) report that SA dates more often involve alcohol and drug consumption, are more often initiated by the man, and more often do not involve any expenses - but when they do, the man more often pays for both dating partners. All of these can also be considered variables that act to increase the likelihood of sexual victimization experiences because they increase the woman's exposure to risky situations. 64 In conclusion, the patterns of results from the research on factors associated with experiences of SA in dating situations seem to provide some support for the feminist/social control theory of sexual aggression. Studies with men indicate that sexual aggression is associated with attitudes consistent with the feminist/social control model's predictions: Men who engage in SA tend to be more traditional in their sex-role beliefs, to maintain attitudes condoning violence against women, to endorse common rape myths and to conceptualize male/female intimate relationships as being manipulative and antagonistic. The same patterns were not found among women who sustained SA. SA inflicted by women toward men has been found to be very rare. Whether or not the few women who do inflict SA maintain the same constellation of beliefs and attitudes as their male counterparts has not been tested. Women's experiences with sexual victimization seem to be associated more strongly, not with personal variables like attitudes, but with contextual variables - the presence of a sexually aggressive peer group, increased exposure to dating situations that are risky, and simply having had more sexual partners. These patterns of findings, not surprisingly, are consistent with the view of women as victims of sexual aggression and men as perpetrators. 65 Effects of sexual aggression and victimization. The empirical literature on the effects of sexual aggression and victimization is limited. It focuses almost exclusively upon the psychological repercussions of being raped, with little attention to the impact of less severe forms of SA or to issues related to rape in the specific context of dating relationships. The literature also includes very little information about the extent of physical injuries that result from experiences with SA - especially in regard to assaults by acquaintances or in dating situations. This is an unfortunate oversight, given what is known about the nature of sexual assaults in these contexts. Kanin (1984), in his study of 71 self— identified ‘date rapists' on a college campus, found that, while weapons were seldom used in sexual assaults in this context, the majority of incidents involved, "a heavy reliance on the physical overpowering of the female” (9. 101). He also reported that most of the victims ”exerted consistent physical opposition to the aggressive sexual advance”. Given that this apparently high level of physical combat occurs during SA incidents, it seems probable that injuries would not be uncommon. Belknap (1989) assessed the prevalence of physical injuries sustained during sexual assaults by acquaintances in her study of National Crime Survey respondents. These surveys are conducted annually over the phone with a random community sample, and Belknap's study focused on the 66 respondents between 1973 and 1982. Her sample included 212 women who had experienced rapes or attempted rapes by "nonrelative acquaintances", excluding assaults by family members (which included spouses) and strangers. She found that 75% of the women had experienced at least one injury as a result of the assault, and nearly half of those reported multiple injuries. Fully 40% of the sample reported requiring medical attention. The women who knew their assailants well (as opposed to those who described them as acquaintances only) reported mpg; injuries, and greater severity. As discussed previously, the methodology employed by the National Crime Surveys have been severely criticized as being unlikely to reveal the true incidence of rape and attempted rape. It is likely, further, that the incidents which are reported are the most violent, since victims identify their experiences in response to a question about "attacks". So the applicability of these injury data to the larger population of victims of severe forms of sexual assault by an acquaintance is not clear. But both the Kanin (1984) and the Belknap (1989) study do, at least, reveal the importance of assessing physical injury in future studies of sexual aggression. There are many more studies exploring the psychological impact of rape, although, as stated above, few of them examine issues specific to victims of SA in dating relationships. Two recent reviews (Koss & Burkhart, 1989; 67 HcCann, Sakheim, & Abrahamson, 1988), indicate that lowered self-esteem, depression, a variety of fear and anxiety reactions, sexual dysfunction, and interpersonal difficulties have all been found to be very common aftereffects for rape victims in a number of clinical studies, several with longitudinal designs. In addition, the following reactions have been found to be problems for rape victims in multiple studies: (a) affective constriction or blunting - especially immediately following the assault (Burgess & Holstrum, 1974; Ellis, 1983; Ross & Burkhart, 1989); (b) feelings of anger, hostility or rage (Burgess & Holstrum, 1974; McCann et al., 1988), which is noted to be a particular problem with victims of acquaintance rape (Ellis, 1983); (c) feelings of guilt and shame (Burgess & Holstrum, 1974; Ellis, 1983; McCann et al., 1988); (d) psychosomatic and other physical symptoms (Ellis, 1983; McCann et al., 1988); (e) suicidal ideation or actual attempts (HcCann et al., 1988; Warshaw, 1988); and, (f) impairment of social functioning related to work problems and restricted socializing (HcCann et al., 1988). Several authors have reported that a significant number of rape victims continue to experience these emotional and adjustment problems long after the assault has occurred (Ellis, 1983; Koss & Burkhart, 1989; Murphy et al., 1988; Nadelson, Notman, Jackson & Gornick, 1982). One follow—up study of rape victims found that over 40% of them reported serious emotional and/or adjustment problems from one to two and a half years after the assault. These included sexual 68 problems, restrictions in socializing, chronic experiences with suspiciousness and fears, and depression (Nadelson, Notman, Jackson & Gornick, 1982). This is consistent with a number of studies reviewed by Ross and Burkhart (1989) that have indicated that as many as 63% of all rape victims still report experiencing persistent adjustment difficulties for several years after the rape. In recent years victims of rape have sometimes been considered to be suffering from Post-Traumatic Stress Disorder (PTSD), a diagnostic category of the American Psychiatric Association which appears in the Diagnostic and Statistical Manual of Mental Disorders (DSM III-R, American Psychiatric Association, 1987). This diagnosis is reserved for individuals who suffer from a set of characteristic symptoms following the occurrence of a trauma which would evoke significant symptoms from almost anyone who suffered that trauma. The symptoms include (a) repetitive reexperiences of the trauma through flashback images, dreams, or thoughts, (b) denial, numbing, or avoidant responses, (c) hyperalertness, (d) physiological overreactivity, (e) irritability or outbursts of anger, (f) sleep disturbances, (g) concentration difficulties, and, (h) intensification of symptoms in proximity to events resembling the trauma. McCann et al. (1988), in discussing its application to rape victims, indicate a number of criticisms that have been made of the PTSD diagnosis. They note that a number of critics have found the diagnostic criteria to be too 69 restrictive, thus resulting in an underestimation of the prevalence of PTSD. Also, the specificity of the diagnosis, when applied to rape victims, must be questioned since it does not include reference to some of the major psychological symptoms found to be common among this population in numerous empirical studies (e.g., sexual problems, lowered self- esteem, and, feelings of guilt and shame). At the same time, many of the symptoms descriptive of PTSD are also descriptive of victims' reactions to rape. It may be that rape victims are prone to the development of PTSD and, in addition, are - subject to other common emotional reactions that are specific to the trauma of rape. Whether or not rape victims are diagnosed as suffering from PTSD, the studies discussed above make it clear that, at least among clinical populations of rape victims, the scope and seriousness of psychological reactions to experiences of sexual assault are substantial. Furthermore, they are long- lasting. What is not clear from these studies is whether the same would be true of a more representative sample of rape victims. It is known that the majority of rape victims - especially victims of acquaintance rapes - do not report their experiences to anyone, nor do they seek medical or professional help. It could be suggested that most rape victims would not be expected to experience the same degree of difficulty after the assault as do the clinical samples. There are few studies of the effects of SA that focus on more representative samples. However, two recent 70 publications concerning the results of the national, multi- campus study of college students' experiences with sexual aggression (conducted by Ross and her colleagues and described in previous sections) address this question (Koss, Dinero, Seibel, & Cox, 1988; Warshaw, 1988). Warshaw reports that 30% of the rape victims surveyed in that study contemplated suicide after the attack. Further, 31% of them reported that they sought psychotherapy afterwards and 82% indicated that the assault had permanently changed them. It will be remembered that 84% of the rape victims in this study knew their assailants, and the majority of the assaults occurred on dates. To assess the degree to which the effects of acquaintance rapes are similar to those experienced by victims of stranger rapes, Ross et al. (1988) compared the psychological impact of the rape experience on the victims of stranger vs. acquaintance rapes. These authors found no significant differences in the two groups' subsequent levels of depression, anxiety, sexual adjustment or relationship satisfaction. They also compared acquaintance rape victims' reactions across four groups distinguished on the basis of how well they knew their assailant. The results of this comparison revealed that those women who knew their assailants best tended to report stronger emotional reactions. These women also expressed greater difficulties in their subsequent ability and desire to seek out intimate relationships. 71 Thus, the results of this single large study, utilizing a representative national sample of rape victims on college campuses (including a preponderance of victims who knew their assailants), are consistent with the results of the clinical studies. They indicate that the psychological aftereffects of severe experiences of sexual victimization are profound. However, there is an obvious need for more research in this area with non-clinical samples. As the above discussion demonstrates, it is well substantiated in the literature that the most severe forms of sexual aggression place a tremendous emotional toll on victims. But the emotional effects of milder forms of SA are not well documented. In addition, very little is known about the extent of the physical injuries which are sustained during sexually aggressive interactions. Given the prevalence of SA in dating relationships, more complete information about its emotional and physical consequences is essential. One of the goals of the current study was to explore these consequences in greater detail in a non- clinical sample of college students. Non-Sexual Physical Aggression in Dating Contexts: Prevalence In recent years, an increasing interest in the phenomenon of non-sexual physical aggression in the context of dating relationships has been reflected in the growing number of published studies on the topic in professional journals. A number of different terms have been used in this literature 72 to describe the phenomenon. Some which have appeared are: "dating violence” (Sugarman & Hotaling, 1989), ”courtship aggression” (Gwartney-Gibbs & Stockard, 1989; Riggs & O'Leary, 1989), "courtship violence" (Bernard, Bernard, & Bernard, 1985; Billingham & Sack, 1986; Makepeace, 1981), "premarital abuse" (Cate, Henton, Koval, Christopher, & Lloyd, 1982), and "woman abuse in dating relationships (DeKeserdy, 1988). A range of definitions have also been employed. Some investigators have included threats of violence in the definition, while others have restricted the definition to physical acts of violence. Surprisingly, most investigators in this field have defined and measured violence in dating relationships without including reference to acts of sexual assault. This seems an artificial separation which is bound to have led to a systematic underestimation of the amount of male-to-female dating aggression. For clarity, in this paper, the term "sexual aggression” (SA) is used to refer to explicitly sexual violence and coercion, while "non-sexual, physical aggression” (NsPA), or ”other forms of physical aggression” will be used to refer to forms of physical violence which do not involve actual or attempted coerced sexual contact. When the terms "dating aggression” (DA) or "dating violence" are used, it can be assumed that both forms of physical violence are included. (More specific definitions will be given in a later section.) 73 Dozens of studies have been conducted in recent years to assess the prevalence of non-sexual physical aggression in the context of dating relationships. The results of these studies have been alarming. Sugarman and Hotaling (1989), in their recent review, reported that various studies have found that from 9% to 65% of the respondents have, at some time or another, been involved in at least one incident of this type of physical aggression as either a perpetrator or a victim. Sugarman and Hotaling (1989) point out a number of problems which arise when trying to make sense of such a wide range of rates. These include, differences in the measures and definitions employed, different sampling techniques and different populations sampled. Nevertheless, they draw the following conclusions: (a) the highest estimates have included threats of violence in the definition, (b) lower rates are found with high school student samples, as compared to college students, (c) higher involvement rates result when a greater proportion of the respondents are women, (d) more recent studies indicate higher rates of involvement in dating aggression, and, (e) regional differences seem to exist, with higher rates found in studies conducted in the south, followed by the west, and lowest rates in the midwest and east. Sugarman and Hotaling conclude that, on the average, "over one third of those surveyed report an experience of violence victimization at some point in their dating history" (p.8). In general, across all studies, the milder forms of 74 violence have been found to be the most common while more dangerous forms are more rare. But even the most severe and life threatening types of physical aggression have been found to be a part of the dating experiences of a significant portion of young people. Makepeace (1983) found that 9.3% of his sample of college students reported involvement in the most severe forms of violence, while Lane & Gwartney-Gibbs (1985) found this to be true for 10% of their sample. It is clear from this prevalence data, as well as from that pertaining to the incidence of sexual aggression, that physical aggression in dating relationships is a social problem of major proportions - and one which may have deadly consequences for a significant number of young people. Factors Associated with Experiences of Non-Sexual Physical Aggression. Introduction. Tables 3, 4, and 5 summarize some of the results of 21 studies of non-sexual physical aggression (NsPA) which have explored correlates of this phenomenon. Table 3 (pp. 75-77) presents the significant correlates with measures of men's experiences with inflicting (perpetrating an act of aggression against a partner) and sustaining (being the victim of an act of aggression from a partner) NsPA. Table 4 (pp. 78-80) presents the significant correlates with measures of women's experiences with NsPA. Table 5 (pp. 81— 82) displays the results of any analyses from the reviewed studies which did not control for gender. The results from 5 7 .o..esoo .o egos s. .o..on .e .en sososea no.3 uo.asoo oosos.eou e.ee .o egos s. .e..on an .(a os.oo...s. .0 uoon..ex.. uoeeonoou sex eeonsoee. eras os.>en recon-a no.3 uo.a3oo eoses.eou e.es .o egos s. .e..eo .~ ouo..ou .s. :2. 8.83.... .3... .535. 823...... 8.2. .3... 882.88 2 .3 88 2.8: 3:8 3.. £8: 2.2.8 38: 2.3. ...eo.nes use us.oeu noon. oo.neso.oe.o. ooes.os. s. ooso.o.> so: a .sonoooe..nu une3oo eouso.ooe e.ne.o>e. one: eeossxe <3 no.3 eeoso..oaxe e>en on: eoono .— oue..ou ..e>ox . .mou floss—V usesow mm. .sooson .eoeu . us.omoso.3 an . ee eooso..oaxo ones .3 ..o>s. .o .o>o. uooeonos. .n . use3oo oeuso.ooe us.osouoe eras .N .<3 ones us.s.eoose no.3 eo.o.osou. nausea es..3oees one. .. “mmwmwmmmmw . ae noose..ouxo ones .e ..o>s. .o .e>o. uooeosus. .n . use3oo nouso.ooe os.onouoe ones a~ ouo..ou uooo-uo..a .5. 28. 8.8.7.... .3... 88.88. 888 8:88... 83 2 "8.4.4.3. 6.8 88: 28.8 .8 e .88 .38.. so: ..oso.o.> oos we resound on..oeou >ono son: oo uoneseoo ee. oso.o.> >..eso:a .m.u Assn—v ouo..ou .omopv noon 8 .2838... 3.88.. .8... 5.... s. 8.83.... 8 88.. 8...... 888 a... 2 annals-H. 3:8 88.8.. S. o 88...... _ .ononose _ .uoonoao. oos ooe>.es< unawaqmmmmw so uooo.o>ou oosne so: Ammopv usesnon . use .eo.o>nn ouo..ou a .unesnou _ no.3 >sooso>s. e.o¢ new sou so eonooe .uonxo.xoe os..3oees. nono.n .. "mmfluquuma .o unseen: uos.neou on .usesnon I is 35393 3.5V 0.08 so: _ ones no.3 unecones use3oo ooe..e o>.o.eon uoeeoeaxo .o e.e>o. .o3o. .. "umflnqummmw eo.ooeo oo..osou ouo..ou .NQQF. sneo..o . .2... 8.828. .888... 9...... .3... 3.8288. 8.8.. 8 .38... 2 "and. 88: 328 S a 8.5... 8...... T|Il|||ullllllllllr||llllllllllu . oo.ne..e> so.eeo...< no.3 oooe.o..ou oseo...s..u o.ne..e> nooo . m oooa . sono3< _ fiaflnal. < 889...... _ C u.-..5. (eve - m. mi) n OHQMH. 76 sonoos so.» cease mo ooso..oaxo s.u.no .o >..seo noose. e s.u.no .o >..seo s. nonoos no.3 ocean oe.o one. onesosen os.oeu sueooe .o senses noose. e uneconea .esxoe .o .0953: .ono.n e Am A3 An A~ A. .<3 us.oo...s. .o uoon..en.. eeeeonos. .ooonsoeo. .ooeeno os.>en nosoeea no.3 uo.3:oo noses.eou e.eso. so >o..eseo .o Enos s. .0..on Am "A.u.osooo ma.o3...aq "<3 s. osoeo>.o>s. uoeeosos. so: no.3 uooe.ooeoe one us.3o..o. on. . .s. o us . .m.u Aono.o usesom one..ou on~ Ammo.v >n3333 .oosne u..no us.oso..o3xo no.3 uooe.ooeae ee3 <3 ms.s.eooao A. “mu.s.eomam .Aoso. .osenom. so: Aono.. .oasne u..no as.oso.303xo no.3 uooe.oooee ee3 <3 3£8.35 A. "um.oo...s. m.u uo...uox ouo..ou mo. soon a ..ene.e: .<3 us.s.eooso .o uoon..ox.. uooeeuos. no.3 uooe.ooooe one eooseno o... uooe.o. no.eon.sos .o>.oeuos ones A~ .<3 us.s.eooso .o uoon..ox.. uoeeonoou no.3 uooe.oouoe nooseno o... e.ne..eou uoonooo. ones A. "mmwmwmwmmw .<3 us.oo...s. .o uoon..ox.. uooeonos. no.3 uooe.oooae one nouseno o... ueoe.on no.eon.sos .o>.oeoos ones A~ .<3 us.oo...s. .o uoon..on.. uoeeonoou no.3 uooe.oooee eeoseno o... o.nen.oou uoosooe. one: A. "mawmmwmwmd .m.u Aeso.o .osesom so: ouo..ou so Ammo.v ooeeaoxe: .o. 8.5328 .o uoon..ox.. aoueoeos. 33.noso.oe.o. os.oeu s. <3 us.oo...s. AN .<3 os.s.eooso .o uoon..ox.. oooeonus. ouso.o.> nexuse oesne .enno> .eosonea us.oooso.3 A. "unquwummmw .<3 8.8.8... .o uoon..ox.. someones. 33.neso.oe.en us.oeu s. ooso.o.> us.s.eoeso A~ Asmo.. .oenom .<3 8.8.8... .o .35 .38: 89.8 co... .- 83.88 uoon..on.. someones. ooso.o.> .oxuse cease .en.o> .eosonea us.oeoso.3 A. "ma.ou..wu4 .o so.e.o> uo...uox ecu..ou on. .enn.o->oso.e3o .<3 us.s.eooso .o uoon..ox.. ooeeonos. u..no oe nonoe. usesoo <3 oo...s. nonooe os.oooso.3 Ao ”so.o.uue s. .<3 us.s.eooso no.3 on on oo usso. oe.e 0.03 <3 os.oo...s. no.3 ueoe.e..oo >.oseo...so.e ee.ne..e> ..< uuawuwmmmmm A.D.uc00v Ahoowv uwsozoo .o.osoo. n «Home ‘77 enosose3 us.oeu osonsso .o senses nonm.n Am 3.noso.oe.e. .o noose. nooeosu A3 oeoso>.ouo.3xo uneconos. An Aoosou..soo-o.oo .oosousonous.. so..eosoesnoes. uoaeonoou A~ uoonu..no s. ooso.o.> us.oso.nooxo A. "<3 os.s.eoose .o uoon..ox.. ono oueonos. us.3o..o. on. “mawmwmmmmw 3.neso.oe.on «a noose. .ooeonu Ae .Asoe one uooeonos. An .o senses 33.naso.oe.o. eons.os. s. ouso.o.> .o eoseo3oooe ones A~ uooo.oo.ussv oeoso>.oeo.3xo uooeonos. A. eosousom Ammo.o uoou .<. ac.ou...c. .o uoo...o... 8.. oaeo.uc. ac.:o..o. as. .uaflmmonoaq .m.o .3.3.. as...” ouc..ou no. -ao... o «soon .<3 gwcmougm *0 00»... 50:0...— Ou Yugo. 0300:. x. ms» .330. «N .<3 ocmcpnugm mo noon—dos: mumousucm Loaf—ma 3:3 comwuuwnasoo 2 ua_c_0um3m .<. .u...¢. _ *0 Gnu... 30.3.: On 0009—... s>u§u camp—8.235. .3 cm swung; a.» .<. .c..o...c. .0 uoo....... .ouuo.oc. .ac...3 .o.: co.....n..ou .n .<3 os.oo...s. , .o .8... .o.... 0. uo...o. c858; u..:oo .onao.... .oco...u..o o.oa .~ ..3.3. ..=..om. no: ..ea.. .8..3 a M .<. ac.ou...c. .0 «a... .o.... 0. nooa... «soon. >..ea. .oso. .. .mmflmuHAMn. «.3 uo...uo: ouo..oo 8.. A.... .cae.oa.m no.0e..e> so.aoona.< no.3 eooe.osnou oseo...s..w 0.3m..e> ooeo o nono3< .o.osoo. n «Home ‘78 an.ucouv co_uaogaou *0 «one; Locum; .0 (m ocpuuw.wc_ .0 noun; uozu_c an wdon new gocugoa >3 vow: uu_uuou uu_d*c0u o>.aaogooo >..ongo> yo u.o>o. goso.: ac mucosauga acoauogv ago: an ¢a_:uco_uc.og goucod Aw c0263 n.3aco_uo.og gyms» mo >u..o:c ecu cu.) copuuowa.uau an». AP ouoddou “yucca; .uoc av on: ones» :u_3 vogonaou an <5 c_auoau as: case: "umwuwmummw .mpu aosopv «sauna new «ocopv c.0uaogwg .953 2305035225; oz "a :93: . case: a .gognouu_gcu ugozou nouaumuuo o.nogo>ow oboe noosnxo <¢ gu_3 noucc_goaxo o>ag oz: «was» AP ouc..ou ..o>oa .mhu “ohopv anagum ~o~ .coucoz .ouuu ..o>c. we .o>o. vomooguc_ n~ .do>c_ *o .u>o. toaooguc_ a~ ouu..ou uoou.ooL_m .2: :2... 8 .8538". 2: 233., 3.35.. .mB .22. 83.8 33: "38 2 assets: 3:88 >2: :2... 5. 822:5 3 .38 .23.. :88 5. 2 "g 8:5 3:3: om... a 3.22:5 .LOuc0>c_ o.ox new gun xn nogaaoua an .oc_n>u-xoa oc_c_Eov co nogooo gozo. aw ”unwadmmmmm >5 voaodo>ot unana case: Amnopv uLngoa it? .8. :33: 80:8 a .22.: iota... “8 .0522 "g .6 23.2 .9838 § .222... ._u_noa noncogaxo mo o.o>o. 50:0. a~ .,u_¢on voaaogaxo «no. an “whey o.oum c0503 .L¢04.o .. co_uaog..< gu_a non-.oggou acoo.wmc._m o.na_gu> nuuo nan ouoo a non» < amwmumuuuquqmmwmxmu m.qw. .m . AIM Sm « uq0.mmm.... u0-m . -mlco q ‘3 mm .. H-.v -. - .3 e GHQMB '79 ..u.ucouv aucogon coozuon ooco.o_> uoauoc»_3 m:.>oc Ac ucmucoaaog o.nug_aou >..-_ooo *o «nuance co nogooo 5030. an moeouc_ >._an 9030. “N on. uoaooguc_ Ap u;u_3 nouo_uoaoo og- ucaou co_noo.oan<: ca c. uc_u_aog man an gocugoa zu_: coma-umnogoo an o.:n. u._;u nouco_goaxo ac.>.c .e oucog-n coozuon ouco.om> vouaocu.3 ocm>ac an oc_ucoaaog udnag_oon >..o_uoa yo nuances cc .0500. L030. n~ on. voaooguc_ a. ”gu_z nouo_uouuo «Lo ggom .cae.oowm gozunw 35:30u o>_a:naxwmdoumm>ca no: gocuos an «Locugoa oc_uon xuoouu we Longs: gouge. a «N oagaougoucm doaxoa nouco_goaxo ocm>nc AF “(a c_ acoea>do>:_ nonooguc_ gu_3 nouomoouua ago ac.:o..0w as» .mhu “aha—v «sauna c0503 oa~..oo mmN .nquV xgagax .onaoo u._gu voocomgomwo uc_>og gu_: noua_uoa¢- noun; goznvg .— .uc30v nouu.oggoo ucou_»~cu_a oz "wamc_oum:m "um_uu_.wc_ .Aosop .uaagumv who uo_..uoz c050: ouo..ou «up Amwopv oaox a d.osago: .ucso» noua.oggoo ucau_~_cu_a ca .3230. nouo.oggou acou_»_cu_u oz muflrmoumaw wu_uuw.wc~ .whu “usury .aaogum case: ouc..ou ecu .naopv cocoaoxox . ac.uu..*c. n~ ._¢ooguoo >..o:xou gu_: uca5o>do>c_ A— .<¢ oc_uu_.*c_ *0 voo;_.ox_. uoaooguc_ ua_coco_uo.ug ucmuau c_ manna doaxoa go moooguuo >..o=xou saw: acoeo>do>c_ a. "uaflmmummmw .mhu .ouo.v .aogum yo co..go> uo_.,uox cone: vooddou an. Asmapv Lassen a vgoxuoum .onn_u->ocugo3u .u_>mu.acoo docougoagouc_ .auoguu_v .aumoo.o=u>on ..ogo>o *o u.o>o. uoaooguc_ an ._oogua gaze. new unannaa .a_oou douOu uo>_uogoa 5030. as "A.u.ucouv mamawummmw no.3..go> co.umogy.< ;u_3 auuo.uggou «cauvwmc..m o.no_.a> An.uzoov ¢ manna Any-acooV «onowv c.0uuugmm _ «you can ouoo » Lozua< . 8() a_;mco_uo.og am:o_goaa a c_ no>.o>c_ o._cz Luaugum‘oco coca ago: ucwuav an unusugon ocmuou homey ac a_s¢co_uo.og c. coucogowwmu a:o_om.og an nonugaa ;u_: mount “coacog* «Lo: “N “oucan_mcou-*.ou .oucoucoaoucmv >u_.oucaeaguac_ non-oguon AF u<¢ uc.c_ouuau *o voo:..ox_. any oncogoc_ uc_:o..0m och ..o>c. o..g3 soapbon oco cog» one: oc.uou AP ”unwawwumam "umpuo_.wc. .mpu Ansopv «noun» .Acaeoz mo Logan: uouo.oa.u::v uncounum coo..ou mom “sue—v noon .95.. a 33m \uxuczoo cawzoodoanag Am “cu—3 noua.uooao use co_a¢og..< ;u_3 nouo.oggou acou_m_c._m 332...; A.U.ur.00v Aeoopv no._3 a xggoo .coeduuwm «you » Losua< Ac.ucouv e manna 81. ooo.oma Lo>o u_ osooc_ vouas_uoo .aucogoa 0.0;: o¢Ogu acos- >.0x_. mumm an nou_n:-coc cagu nou_:3 ocean >.ox.. «has as: <¢ yo cage» ogo>oa a~ ou-u_nacou 0:: «was» gay non-059:. A, uuaqadmmmmw “asap 8993 .2533 Lo>o a. «soucm vouaemuoo .aucogoa ones: omega ocean xdoxud mama a~ who Ammo—V mnn_u 3332.8 of :2: .8 828g 2 "Haas. 8:28.. 8: 3. .88: .838 8. .tctas a 8.: .uoc v.6 2.: .do>cm Louoogu anoLu .oLucou vacuums a cum: vouomooama oocodo_> doumgua uco dou_goeosa nausea nouau_uuo odnogo>ow ago: an “uncouaua .oogoa gu_g eeo Ammo". uc_uou duaaou oc_g3u augouu .mpu mo odaaaa a 305* (a su_3 Lonn0ua_gnu a. «up.n~v «so» 50* nan .amaomgoaa «aeouan n_:aco_ua.og Louma museum >..a3ma ox :8. . 3 3. 359:8 5. 833:5 as. 8.533» 3 32x... 8.2. 2.. use 3.2.8 S .38 £88.. as>uconcuaou.go>o use oucon_wcoo.».oa yo goodsv auoco>_»_caaoLuc. nouoogocw n¢ euouao.w.oa gaze. an oucodom> uLoIOu nouaumuuo odnogo>o~ ogofi AN .whu 8:93:32 :3. 5 8:30; 5:. 398285 2 2:: 83: 38: nouooguc_ <5 c. ucaao>do>c_ naugum cox upp .caso: coo..ou no~ goanaa: a .ooo 383 Rs: aasowgou egos: aoeauon amgmco_ua.ug oz» Louwa ung_w Lauoo ou >.ou.. ago: a. (3 Am .mhu a hogaOua_gsu .28. v 3 3o. 83: :93. voua_oouuo ago .;n .a:xou.coc oc_uo_.wc_ new uc_cpouaaa A. «sauna cox mm" .coaos coo._ou ~o~ .cOuco: .ouou .ms 83: 825 38: u§._§u 3:20:39. 3228. s "a 3E3: 5: .9 £25: 08:8 E 5:83.: , «Eats, maflmmmuumq mmmmqmmm. aummmquamqummmmq 4mmflmxmm .c -. -0 ml“: .m . fléfism ... .coflmmmu... ...o.m n. @839. o 9.3 mwo m.u-.xm m manna E32 .5305a mc_uou mcoso cugu manago or» guguo ocean coseou «50¢ a. oucodo_> >.co «do: .05305u can Losuo :u_: coo_5oaaou c. anago ocmuou acne: coesoo «gas a. coco.o_> 3:0 395... £853 :2. a... 3:30; .333. 50 3.352 690.5 39.3 3 ~.uuosu_z axoo. 5:059 gonna. .05. go» on...05ucou uaozu_3v «5:055 uc_uou sag» .uomggne coca .ucmuouwnagou occa- uuugu_g 05. <0 50 now-5 .ooo gov oc_d.05ucou an <5 c. ucoeu>do>c_ oc_uooguc. cu vowed»; aw one ocwaoogoou Aw guano acuuau.nagou 0:» ocean ogo>ou nae: uca coeaoo anus a. <5 A— muu_.~c_ ace _c_oum:» couzuon once covuoc_umwu oz .mp0 “20: . gonna .Acoeox new ca: 50 gunezc uoao.oa_ucav c.5aaa >u.c:asoo 505* a.o:o_>_uc. oc_uau.nazou 5n~ uco no.55ue moon uco “mucounuu ouc..oo uc_uov o~m Aowopv anagum a «you» .co_uo5:u mgucoe 005;» can» «mod 5o o5_:¢co_uc.og aces. noucoaogaoggo>o on.- cgox wan .cowuagau ugoo> oau can“ oboe 50 ago: unmgacomuodog cos: coeaau uuos “005.05 0:» oc_..au .¢o_5:mc_v coucoaooacou a~ canons». uc_uau ocean co_uogau aa5o0> 93a cosu «no. was «ac saga «Loss *0 o5.:uco.»a.og cm acoaoogy use: .— "umdawmmmmm a~wopv 05::m we. >oo-um >9 uo5o.o>oo (a van <5 50 0535.0: .acoE no. .coec: «mpv oo_55os go>uc 50 5:05a voguuae o» voggasou auconaua oooddou uo_55ge on, “owopv ..o:o: a conga .oaaoc .<5 ac_c_o»aaa o» vouadugxxducoumw_cu_u ago: modna_5o> 05am uuuflumummmw noun. Locupg gu_: unuo_uouaa aucogon coozuon ouco.o_> oc_uaoCu_: A~ Loso_z zu_: unuo_oomaa unana u._:u ocmucowguaxo AF «maflmmHJHmd “one. .aaogumv nnwopv .Apo. v av 0m. vouomuowmo 05¢ <5 acmcmnumaa uco oc_uu_.mc_ who oomwwuoz co: ~o— “c0503 «noddou «up omoa a ddacugox A.cueo: vco cos *0 «Logan: vouodua_uc=v >m:0.oo_ «none cou50 0502 05. >05» uuuaua .sucou.uc_ .oucu_5oaxo <5 50 Loan. u. uxou uaono uucaeooguuumu o>.o>c_ couwo ogos mucou_oc_ oust doauoo Am ungoza swuop> no ao>doa un_goaou oucgn oc_uou xuaauu uc_g=o Laooo “we..ov saucou_uc. u..oza page .F 5° co_u5.5oaoo 0:3 mucouaum oooddou nn~ “ooopv cocoaoxoz .<5 0505 gu_3 vane—uoamo aw oucazo coo LoCuLom‘uocu uc_>o_.on Av .<5 0503 :u_: non-moon». a. augaco_u-.05 c_ ucoEuuo>c_ guoc05ua oc_>a; an .<5 0503 ;u_3 vow-wooaua u_ guacoa 5.:mco.uo.05 «N .<5 c. auuoogucu zu.z non-woomao u_uoau uc_uu.uouoc an :uocoua_ugoa= ocpma A- .whu Anhopv .mnauum o.namgo> co: 3. mo. Lawn mummmxmm .suo_.mcou we ado>od aoogo>o o>ono gu.3 mamznco_uu.05 «30550»: c_ ..< .co: .0 .coeo: oun..ou ~op .oooFV once a .o>og .nxo.4 o no Logua< an.ucoov m manna 83 one study may appear in all three tables, where appropriate. All but two of the studies depicted in these tables focused exclusively on college students: The exceptions are a study by Henton, Cate, Koval, Lloyd & Christopher (1983) which examined NsPA among high school students, and another (Stets & Straus, 1989) which included data, not only from college students, but from representative community samples of married and cohabiting individuals as well. (Both of these studies, coincidentally, appear only in Table 5.) A striking similarity across the studies in these tables is that so many of them have utilized some portion of Straus' (1979) Conflict Tactic Scale (CTS), or a modification of it, to measure NsPA. The CTS is presented in Table 6. It consists of three scales: Reasoning (items 1 - 3), Verbal Aggression (items 4 - 6 and 8 - 10), and Violence (items 11 - 19). The Violence Scale, which measures, "the use of physical force against another person as a means of resolving...conflict” (Straus, 1979, p. 77) is the one that directly measures acts of non-sexual physical aggression. In most studies this was the only scale utilized, but some investigators used all three (Billingham, 1986; Billingham and Sack, (1986). Others used modified versions which included items that assessed threats of violence, or other forms of verbal aggression in addition (Bernard, Bernard & Bernard, 1985; Gwartney-Gibbs & Stockard, 1989; Lane & Gwartney-Gibbs, 1985; Marshall 5 Rose, 1988). 84 The findings presented in Tables 3 through 5, when possible, include only those that are directly relevant to non-sexual physical aggression (i.e., for the most part, only results from the Violence Scale are presented). However, those studies which included measures which assessed some forms of verbal aggression along with acts of NsPA in the same scale are also listed. It is noted where the measures incorporated other than acts of NsPA. Only three of the studies included in Tables 3 through 5 used measures of NsPA that were not derived directly from the CTS (Bernard, Bernard & Bernard, 1985; Hakepeace, 1989; Rouse, Breen & Howell, 1988). As previously mentioned, this widespread reliance on the CTS is problematic, since that measure has been criticized on a number of counts, which will be delineated below. One thing that is striking from a perusal of the results of these studies of non-sexual dating aggression is the wide range of factors that investigators have addressed. Although it is undoubtedly useful to examine a range of factors that might be related to a phenomenon of interest, it is also probable that this range reflects the lack of theoretical focus in the field, which was discussed earlier. A review of these tables reveals several findings which have been replicated across numerous studies. One thing that has been found repeatedly is that inflicting and sustaining NsPA are significantly correlated (Burke et al., 1988; Cate, Henton, Koval, Christopher, & Lloyd, 1982; Firestein, 1989; 85 Table 6 Items fromfiTh§_gonflict Tactics Scales (Gelles & Straus, 1988; Straus, 1979). 1. Discussed an issue calmly 2. Got information to back up your/his/her side of things 3. Brought in or tried to bring in someone to help settle things 4. Insulted him or swore at him/her/you 5. Sulked or refused to talk about an issue 6. Stamped out of the room or house or yard 7. Cried 8. Did or said something to spite him/her/you 9. Threatened to hit or throw something at him/her/you 10. Threw or smashed or hit or kicked something 11. Threw something at him/her/you 12. Pushed, grabbed, or shoved him/her/you 13. Slapped him/her/you 14. Kicked, bit or hit him/her/you with a fist 15. Hit or tried to hit him/her/you with something 16. Beat him/her/you up 17. Choked him/her/you (added in version published in 1988) 18. Threatened him/her/you with a knife or gun 19. Used a knife or fired a gun Note. From "Measuring Intrafamily Conflict and Violence: The Conflict Tactics (CT) Scales” by Murray A. Straus, 1979. In The Journal of Marriage and The Family (pp. 87-88), February, 1979. Copyright, 1988 by Richard J. Gelles and Murray A. Straus. 86 Gwartney-Gibbs, Stockard & Bohmer, 1987; Benton, Cate, Koval, Lloyd & Christopher, 1983; Marshall & Rose, 1988; Sigelman, Berry & Wiles, 1984). The magnitude of this correlation seems to be quite consistent. In those studies where the bivariate correlations were reported, they range from .55 (Burke, et al., 1988) to .64 (Cate et al., 1982). Another, related finding in a number of studies is that the majority of respondents who have reported the occurrence of non-sexual violence in their dating relationships have consistently indicated that both partners have inflicted violent actions against their partner rather than just one partner doing so (Billingham & Sack, 1986; Cate et al., 1982; Deal & Vampler, 1986; Benton, et al., 1983; Sigelman et al., 1984; Stets & Straus, 1989). Some have suggested that this demonstrates that NsPA is best understood as a mutual problem in dating relationships, most likely to be perpetrated by both dating partners, rather than one that would be more accurately characterized as a pattern of violent interactions with distinct, stable, gender-related assailant and victim roles, as the feminist/social control model predicts. But, of course, unless the measures of NsPA that one uses are capable of distinguishing between self-defensive and initiated acts and can assess qualitative differences between different acts (e.g., in terms of injuries sustained), then information about the concurrence of two partners' aggressive behaviors is fairly meaningless in terms of sorting out this issue of 87 gender-related assailant/victim roles. The CTS, as mentioned previously, does not allow for these distinctions. Some patterns of results from these studies which are noteworthy are those pertaining to the following areas: (a) relationship variables that are associated with increased levels of NsPA, (b) the role of childhood experiences with violence, (c) the impact of attitudes about violence in interpersonal relationships, and, (6) issues related to sex- roles, gender identities, and beliefs about control in relationships. These issues will be addressed in separate sections below. Relationship characteristics. Numerous studies have indicated several relationship characteristics which are related to increased rates of non-sexual physical aggression. Five studies have found a positive relationship between rates of NsPA and the length of time that the couple has been involved (Arias, Samios, & O'Leary, 1987; Firestein, 1989; Lloyd, Koval & Cate, 1989; Rouse, Breen & Howell, 1988; Stets & Pirog-Good, 1987). A number of other studies have indicated the importance of other, related measures of relationship involvement in increasing the likelihood of the occurrence of NsPA: (a) a combined measure of the number of dates and length of relationship (Burke et al., 1988); (b) more frequent dates (Stets & Pirog—Good, 1987); (c) fewer current dating partners (Stets & Pirog-Good, 1987); (d) increased relationship commitment (Billingham, 1986); (e) 88 increased rated ”seriousness” of the relationship (Cate et al., 1982; Henton et al., 1983; Hakepeace, 1989), and, (f) higher levels of stated investment in the relationship (Lloyd et al., 1989). In regard to the timing of occurrences of dating aggression in the different stages of dating relationships, Makepeace (1989) found some interesting trends. His study focused on students' reports of the "worst incidents" of dating aggression (including both sexual and non-sexual physical aggression). He found that, while the frequency of ”worst incidents” was highest among steady and then casual daters, the most serious effects (emotional and physical) were more common in ”first date" incidents and among couples who were living together. Others have similarly found that couples living together were at higher risk of experiencing more, and more serious NsPA (Lane & Gwartney-Gibbs, 1985; Sigelman et al., 1984; Stets & Straus, 1989). In contrast to those in the "first date" and "living together” groups, Hakepeace found that serious effects were less frequently reported during the casual or steady dating phases and reported with moderate frequency during engagement. He also reported that, while first-date incidents were usually preceded by disagreements about sex and alcohol or drug usage, the incidents which occurred at later stages of the dating relationship were more likely to be triggered by disagreements about jealousy. 89 These findings led Hakepeace (1989) to suggest two different types of dating aggression: "predatory" and ”relational". He described the predatory type as, "particularly intense violence...(which is)...emotionally and physically dangerous and is frequently coupled with sexual assault, often motivated precisely by sexual exploitation" (p. 103). In contrast, Makepeace maintains that relational courtship violence increases in frequency over the course of time as issues arise for the couple concerning the status of the relationship and how and with whom to socialize. Hakepeace suggests that the violence at this stage is less severe and dangerous. But, as the couple's commitment to the relationship increases and the partners become more and more exclusively involved, the conflicts may become more intense and dangerous, since the likelihood of walking away is diminished and the sources of conflict may be more important to the two partners. Henton et al. (1983) noted a bimodal frequency distribution in their data that is consistent with Hakepeace's typology. They found that incidents of NsPA were most common during the serious dating phase, but reported also that a sizeable proportion of their sample (23.1%) experienced NsPA during the early stages of dating. Rouse et al. (1988) reported a similar bimodal distribution of "consequences" among their sample. They found that serious consequences were most likely during relationships of two years or more duration, but they also appeared, with greater 90 frequency than expected, among relationships of less than three months duration. This finding is also consistent with Hakepeace's notion of two distinct types of dating violence. Hakepeace suggests that the predatory type of courtship violence is largely overlooked in the literature and, further, he suggests that, since it is of such a severe and dangerous nature, it requires much more attention, both in terms of research and resources directed toward intervention efforts on campuses. To what extent is the notion of two distinct types of dating violence consistent with or at odds with the feminist/social control model of dating aggression? That perspective maintains that all forms of violence in intimate relationships between men and women are caused, fundamentally, by common factors - namely, widely accepted norms of male dominance in their relationships with women and norms condoning male aggression as an appropriate means of achieving that end. It is consistent with this view to suggest that, while the overriding goal of dominance may remain constant among men who utilize physical force and violence in intimate relationships, the specific goals (e.g. sexual intercourse, or controlling his partner's social contacts with others) may vary across the different stages of dating relationships. In other words, the feminist/social control model would predict that, regardless of differences in other aspects of the two types of dating violence (the relative role of sexuality, different consequences, etc.), 91 the men who perpetrated both types would be expected to maintain relatively stronger expectations of male control in relationships, and more favorable attitudes toward physical aggression to achieve this than would non-violent men. Childhood experiences with violence. A number of studies have found that experiences with violence in families of origin are related to the incidence of non-sexual physical aggression. This relationship has been found somewhat more frequently among men than women, though. Four studies found that, for men, experiencing abuse, or "harsh punishment” (DeMaris, 1987) as a child was related to later experiences with NsPA with dating partners (Burke et al., 1988; DeMaris, 1987; Marshall and Rose, 1988; Murphy, 1988). In addition, three studies found that having witnessed physical violence between parents was a significant factor in accounting for men's subsequent rates of NsPA in dating relationships (Burke et al., 1988; DeMaris, 1987; Gwartney-Gibbs et al., 1987). This was found to be related more often to reports of sustained acts of aggression (Burke et al., 1988; DeMaris, 1987; Gwartney-Gibbs et al., 1987) than to reports of inflicted aggressive actions (Gwartney-Gibbs et al., 1987), however. Women's childhood experiences with violence have also been found to be related to subsequent experiences with NsPA with dating partners in three studies. .Marshall and Rose (1988) and Sigelman, Berry & Wiles (1984) found that women's past 92 experiences as victims of child abuse were related to subsequent experiences of victimization from NsPA in dating relationships. Sigelman et al. (1984) found a significant relationship with inflicting violence upon dating partners as well. Murphy (1988) and Sigelman et al. (1984) also found that witnessing violence between parents was related to experiences with NsPA in dating relationships for women. These findings, taken together seem to provide some support for the notion that the risk of experiencing non- 'sexual violence in dating relationships is increased for individuals who have childhood histories of experiencing or witnessing domestic violence in the home. This, of course, suggests that the aggressive interaction patterns which are reflected in abusive dating relationships can be, to some extent, learned from early experiences within the family of origin. This social learning perspective is consistent with both the feminist/social control model of dating aggression and with more gender-neutral models, like that put forth by Riggs and O'Leary (1989). Attitudes toward violence. Five studies in Tables 3 through 5 examined the relationship between attitudes toward violence in intimate relationships and experiences with NsPA (Burke et al., 1988; Cate et al., 1982; Deal & Wampler, 1986; Henton et al., 1983; Stets & Pirog-Good, 1987). All of the studies which explored this relationship while controlling for gender found significant positive relationships between 93 favorable attitudes toward violence and NsPA experiences in dating relationships for men (Burke et al., 1988; Cate et al., 1982; Stets & Pirog-Good, 1987), but only Cate et al. (1982) found such a relationship for women. Deal and Wampler (1986) and Henton et al. (1983), unfortunately, did not control for gender in their analyses, but they also found a positive association between favorable attitudes toward violence in intimate relationships and involvement in non- sexual physical aggression in their samples. This relationship, especially the gender differences, require further study, but the results so far led Sugarman and Hotaling (1989) in their review of the literature, to conclude that "the male's attitude regarding violence is more predictive of dating violence involvement than the female's attitude” (p. 14). Of course, since none of these studies involved longitudinal designs, no conclusions can be drawn about the direction of causality in this relationship. It is possible that men who are more accepting of violence in intimate relationships are more likely to inflict it, but it is also possible that the accepting attitudes follow from the engagement in violent activities (Sugarman & Hotaling, 1989). Longitudinal designs would be required to distinguish between these possibilities. The feminist/social control model predicts that male acceptance of violence as a means to achieve control in relationships with women is central to the maintenance of violence in intimate contexts. On the whole, the results 94 from these studies are consistent with this perspective, even in respect to the gender differences which indicate that acceptance of violence is a predictor of men's involvement in NsPA more than of women's. Since the feminist/social control model views men as the instigators of dating violence, their attitudes would be expected to be more predictive of the occurrence of violence in the relationship. Women, seen as the victims of violence, would be expected to experience violence contingent upon their partner's expression of violent behaviors independent of their own attitudes. Of course, even from this perspective, women's attitudes toward violence might be expected to become more positive as a result of having experienced violence from a dating partner - as mentioned above. This attitude shift could serve to reduce the cognitive dissonance which might result from being involved in a relationship with a violent man. In other words, it may serve as a way to cope with the violence. Further, such attitude shifts might be expected to occur more often, or more strongly when the dating partners are more seriously involved, in contrast to more casual relationships, where leaving the relationship might be an easier caping choice. Unfortunately, none of the studies reviewed here examined gender differences in the relationship between attitudes toward violence and the stage of dating among those who had experiences with dating violence. 95 Sex-roles, gender identities. and beliefs about control. Six studies in Tables 3 through 5 examined variables related to sex-role attitudes, gender identities or beliefs about control in dating relationships (Bernard et al., 1985; Burke et al, 1988; Deal & Wampler, 1986; DeMaris, 1987; Sigelman et al., 1984; Stets & Pirog-Good, 1987). Sigelman et a1. (1984) and Deal and Wampler (1986) examined the relationship between sex-role attitudes (or stereotypes about proper role behavior for men and women) and non-sexual physical aggression in dating relationships. The clinical literature in the domestic violence field has maintained that men who are abusive toward their wives tend to maintain extremely traditional, rigid sex-role attitudes. walker (1984, 1985) found in her study of battered and formerly battered women that they rated their violent partners' attitudes toward women as "very traditional", while they rated non-violent partners they had been involved with as significantly less so. But there is little evidence from violent men themselves to substantiate this claim. To explore sex-role attitudes, Sigelman et al. utilized the short version of the Attitudes Toward Women Scale, while Deal and Wampler used the Osmond-Martin Sex Role Attitude Scale. Both measures indicate how traditional or non- traditional the respondents' standards are for appropriate behavior for women. In addition, the measure used by Deal and Wampler examines the traditionalism of standards for men's behavior. Deal and Wampler, who did not conduct 96 separate analyses for males and females, found no relationship between this measure and NsPA among their sample of college students. However, Sigelman et al. found a significant but small correlation between men's attitudes toward women and their rates of inflicting NsPA upon a dating partner. These authors went on to explore the possibility that the discrepancy between dating partners' sex-role attitudes would be more strongly related to the incidence of NsPA than would the men's scores by themselves. Respondents had rated their own and their partner's sex-role attitudes on a 10-point scale. From these they calculated a relationship discrepancy score for each respondent. For men, they found no significant differences in the rates of NsPA between five groups defined on the basis of the size and direction of discrepancy between the two partners' sex-role attitudes. However, for women a significant curvilinear relationship emerged wherein they were much more likely to report having inflicted violence against their partners if they rated them either much more or much less traditional in their sex role attitudes than they rated themselves. These authors concluded from these results that, for women anyway, it may be more important to attend to discrepancies between dating partners in exploring the role that sex-role attitudes play in dating aggression. From the perspective of the feminist/social control model of dating aggression one might predict that traditional 97 attitudes toward women would have a direct effect upon the likelihood of male-inflicted violence in intimate relationships. Yet, as described in the previous theoretical discussion, this model would predict that the man's need to forcefully assert control and dominance (and, therefore, to engage in violence as a means of doing so) would only arise if and when he was not getting what he wanted from his partner without such a forceful assertion. This would be most likely to occur only under conditions of disagreement or conflict. If his partner's sex-role attitudes are consistent with the man’s, then one might expect fewer disagreements between them about what is appropriate behavior for her, and hence, lower levels of dating violence. This is consistent with Sigelman et a1.'s (1984) finding that what is important is a measure of the discrepancy between the two partners' sex-role attitudes. It is difficult, from the perspective of the feminist/social control model, to account for the fact that Sigelman et a1. (1984) found this effect only in the relationship between women's ratings of sex-role attitudes and their reports of inflicted violence against their partners. The feminist/social control model would expect this same relationship to be evident in the men’s data. But it is interesting to note that a direct relationship between individual attitudes and inflicting NsPA is only obtained from the men's data, while the significant association between attitudes and action for women requires a measure 98 which incorporates an estimation of her partner's attitudes. This difference is consistent with the feminist/social control model's fundamental perspective of men as actors and women as reactors in the experience of dating aggression. Three other studies explored the role of gender identities (in contrast to sex-role attitudes‘) in non-sexual physical aggression in dating relationships (Bernard et al., 1985; Burke et al., 1988; Stets & Pirog-Good, 1987). The measures used in these studies were designed to assess the extent to which respondents' descriptions of themselves fit with traditionally masculine or feminine characteristics. Bernard et al. used the Bem Sex Role Inventory (BSRI): Burke et al. developed a 9-nine item semantic differential measure which they intended to distinguish between masculine and feminine identities (they conceptualized these as bi-polar opposites); and, Stets and Pirog-Good used the Masculinity and Femininity Scales of the Personal Attributes Questionnaire. The results of these studies present a mixed and contradictory picture, due in large part, it will be suggested, to measurement issues. Bernard et al. (1985) found that for men, a more clearly masculine identity, as measured on the BSRI, was associated with inflicting dating aggression. In contrast, for women, those who reported sustaining more dating aggression had BSRI ‘ The measures of sex-role attitudes ask respondents about how men and women typically should or do act, while the measures of gender identity ask respondents to describe themselves on dimensions that have been previously determined to be more characteristic of men or women. 99 scores which indicated they were less clearly sex-typed as feminine than were the women who had not had experiences with abuse in their dating relationships. In discussing these results, Bernard et al. proposed that males who were more traditionally masculine were likely to be more aggressive, since this is a traditionally masculine characteristic. When a man with this combination of traits "perceives that his status is threatened by his partner” (p. 575), he is more likely to respond aggressively. These authors maintain that this constellation of events is more likely to occur, "when his partner is not traditionally sex-typed feminine” (p. 575). These results and conclusions obviously are consistent with the predictions of the social control model. Burke et al., (1988) developed a scale of gender identity specifically for this study. They identified nine items from the Personal Attributes Questionnaire which were shown to maximally discriminate between college men's and women's descriptions of themselves. The resulting scale was then administered along with measures of SA and NsPA. In direct contrast to Bernard et al. (1985) these authors report that, for both men and women, a more feminine gender identity was associated with both inflicting and sustaining more NsPA. They report finding the same relationships between sexual abuse and gender identity. Their conclusion is that abusive dating behaviors must be viewed as, "consistent with the acting out of a feminine gender identity" (p. 87). They suggest that, "this relationship may be the result of two 100 components of a more feminine gender identity: an emotional excitability, which may result in physical outbursts, and an orientation to the dating relationship, which may lead the respondent to seek greater sexual involvement than desired by the dating partner" (p. 88). A closer examination of the items in Burke et a1.'s (1988) measure, however, calls into question the meaning of these results. The feminine end of the continuum on six of the nine items in the scale are: "very emotional", "very passive”, ”feels very inferior”, ”very excitable in a major crisis", ”feelings easily hurt", and, "cries very easily" (p. 80). One might expect a number of these characteristics to signify conditions of emotional distress. If that is true, then it may well be that the relationship which is being inadvertently measured in this study is a link between participation in incidents of dating aggression and subsequent feelings of emotional distress. Given the nature of the items in the scale, (not to mention the correlational nature of the study) it seems unwarranted to conclude, as Burke et al. do, that feminine gender identity plays a causative role in dating violence. Similar problems may plague the measure of gender identity utilized by Stets and Pirog-Good (1987). These authors utilize the Femininity and Masculinity Scales of the PAC, as mentioned above. They find that, for men, both inflicting and sustaining NsPA are related to higher scores on the Femininity Scale ("expressiveness") and, in addition, 101 sustaining NsPA is related to decreased scores on the Masculinity Scale ("instrumentality"). For women, also, they find that sustaining NsPA is related to decreased scores on the Masculinity Scale. Stets and Pirog-Good (1987), emphasize the expressive and instrumental qualities that these two scales are intended to measure. In discussing the relationship between the Masculinity Scale and sustaining NsPA they suggest that, "possessing instrumental skills (e.g., independence, self- confidence and holding up well under pressure) decreases the likelihood for a man and woman to receive violence. Apparently, these individuals are less likely to be involved with another who wants to control them if that control takes the form of violence" (p. 245). But, again, the direction of causality must be questioned. Spence, Helmreich,and Holahan (1979) reported that the Masculinity and the Femininity 'Scales of the PAC correlated significantly with self esteem for both men and women. In addition, the Masculinity Scale was correlated with a measure of ”neuroticism” which included items tapping "depression,...satisfaction with social life, general life satisfaction, and voluntarily seeking professional help for psychological problems" (r = -.53 for men and -.41 for women). Here too, then, it seems just as plausible that the observed relationships between these two scales from the PAQ and experiences with dating aggression are reflecting the emotional distress that results from those experiences. 102 It seems clear, then, that the research into the relationship between NsPA and gender identity is made difficult by the fact that measures of gender identity are, in at least two cases out of these three, likely to be measuring emotional states that would be expected to result from traumatic interpersonal experiences. Additional work in this area is necessary in order to assess the role that gender identity plays in the phenomenon of dating aggression. However, progress in this endeavor will require measures which are not confounded with symptoms of emotional distress, a challenge which may be difficult to meet. Another approach to looking at issues that are central to gender role patterns is to move to a less general level of examination to explore more specific attitudes and behaviors which are thought to be derived from gender identities or sex-role attitudes. Studies which explore the beliefs about and actual patterns of dominance and relative power in decision-making in abusive dating relationships are examples of this approach. . One such study is that conducted by DeMaris (1987). He attempted to assess the extent to which patriarchal norms were associated with experiences of dating aggression by examining the role played by, (a) the balance of resources between dating partners (such as the amount of weekly allowance or income, scholastic average, parental income and education, and ratings of intelligence, attractiveness, self esteem, etc.) and, (b) the respondents' beliefs about how 103 much "control” over the partner men and women should have in a relationship. (Responses here ranged from "the woman should have a lot of control over the man" to, "the man should have a lot of control over the woman".) Consistent with the predictions of the feminist/social control model, he hypothesized an interaction effect for beliefs about control and the balance of resources on violence. He expected that when males endorsed beliefs in male control of relationships, they would use violence to gain control if they did not have other resources available. He found an interaction, but in some unexpected directions. For males, as expected, he found that the lowest probability of inflicting violence resulted when the respondent believed in equal control and the partners had equal resources (Group 1). This replicates similar findings in the marital abuse field (Straus, Gelles & Steinmetz, 1980). However, contrary to expectations, DeMaris found that the probability of inflicted violence was also low for the group of men who believed in male control whose partners had greater resources (Group 2). In line with the predictions of the social control model again, he found that males were most likely to report inflicting violence when they believed in male control and their partners had equal resources (Group 3). Violence was also likely when the men endorsed beliefs in equal control of the relationship and their partners had more resources (Group 4). 104 DeMaris simply concludes that, since this pattern of results makes little theoretical sense, his hypotheses were not supported. However, it seems important to note that, of the four significant differences found for the groups of men defined on the basis of control beliefs and resource distribution, three of them were in the direction predicted by feminist/social control theory (those for Groups 1, 3, and 4), if you consider that groups 3 and 4 both define males who want more power than the resource distribution in the relationship affords them. The group that does not fit the predictions is Group 2, those men who want control, but whose partners have more resources. The social control model would predict high levels of violence for this group. It is hard to make sense of the low violence rates for these men. It is possible that the relationship power distribution is so far afield from their picture of an ideal relationship that they are less invested and involved with their dating partners than are the other groups, which might account for the lower levels of violence here. However, this possibility was not addressed by DeMaris. DeMaris (1987) did not find any similar interaction effects related to women's experiences with either inflicting or sustaining dating violence. Nor did he find any main effects for the balance of resources on either men's or women's experiences of violence. However, women who endorsed a belief in female control of relationships, independent of the resource distribution in 105 the relationship, were more likely to report inflicting violence upon their partners than were other women. Since DeMaris found only 2.4% of the men responded that they believed the woman in a relationship should have any control at all over the man, it is probable that the majority of the women who believed that they should have some control over the man in a relationship were involved with men who did not share this belief. It seems likely that such a relationship would involve frequent conflicts over control. The social control model would predict that these women challenge their partners' sense of control in the relationship and the men respond to this challenge with violence. The women's reported inflicted violence is likely then, from this perspective, to be self-defensive. If this were true, one would expect a significant relationship between believing in female control and women's experiences of sustaining violence, as well. DeMaris reported that, although this relationship was of similar magnitude to that with inflicting violence, it was slightly smaller and not quite significant. An alternative explanation of this result is that women who want to exercise some control over their partners use violence in their attempts to achieve it. This, of course, is contradictory to the perspective taken by social control theorists. Information concerning the context of the violent encounters and a measure of NsPA which allowed for a distinction between self-defensive and initiated violence 106 would be required to explore these alternative hypotheses further. Another study which examines the relationship between NsPA and the power distribution in relationships is that conducted by Sigelman et al. (1984). They asked college men and women to use a five-point scale to indicate the existing ”dominance" pattern in the relationship from "almost completely dominated by the man", through equal distribution of power to "almost completely dominated by the woman”. They found that both inflicting and sustaining NsPA were more common experiences in power-imbalanced relationships and were least likely in equal relationships. These curvilinear patterns were found in both men's and women's reports of experiences with NsPA, but were less pronounced for the men. The relationship was only significant for the distributions of women's reports of inflicting and sustaining non-sexual violence. Consistent with DeMaris' finding pertaining to women who endorsed a belief in female control of relationships, Sigelman et al. (1984) found that the group of women who reported the highest rates of inflicting violence against a partner were those who described their relationship as "almost completely dominated by the woman". Fully 83.3% of these women reported inflicting NsPA upon their partners. Unfortunately, in neither case do the authors indicate how many women fell into these categories, so it is difficult to tell how prevalent a pattern this is. 107 Again, it is hard to make sense of these findings. The gender differences, in particular, are contrary to the predictions of the social control theory, and neither are they consistent with the predictions of any other theoretical model. What seems to be clear from both the DeMaris (1987) and the Sigelman et al. (1984) studies is that, in relationships where power is shared equally, violence is unlikely. These studies also indicate that imbalances in power and resource distributions, as well as dating partners' beliefs about control, are important variables to consider in achieving an understanding of non-sexual physical aggression in dating relationships. This fits with Stets and Straus' (1989) observation that control and autonomy issues are central to the conflicts that fuel dating violence. They suggest that it is the increased conflict about these issues that may account for the increased risk of violence between cohabiting partners, since issues of control and autonomy are likely to be very pronounced in these, highly ambiguous relationships. Dobash and Dobash (1979) also assert that control and autonomy issues are crucial determinants of violent interactions between intimate partners. One contrast between these two positions is that, in line with the social control model, Dobash and Dobash emphasize gender differences stemming from the impact of the patriarchal norms which shape conflicts over power and control between men and women in intimate relationships. Stets and Straus', on the other 108 hand, do not differentiate between the ways these issues of control and autonomy might be experienced by men and by women. The results of the DeMaris (1987) and the Sigelman et al. (1984) studies suggest that these issues do play different roles in men's and women's experiences - but in some ways which seem contrary to the expectation of the social control theorists. However, as stated before, utilizing a measure of NsPA which allows for a distinction between self-defensive and initiated violence and also for more attention to the context of violent interactions between dating partners may help to to sort out the meaning of these differences. Effects of Non-Sexual Physical Aggression Sugarman and Hotaling (1989), in their recent review of the literature point out that few authors have given the emotional and physical consequences of experiences of non— sexual physical aggression enough emphasis. They maintain that, in general, the choice has been to focus upon quantitative comparisons of different acts of violence (i.e. comparing rates of violent actions) rather than on qualitative differences in the experiences (i.e. comparing consequences). Only Makepeace (1986) and Firestein (1989) have conducted studies focused specifically upon the physical and emotional consequences of dating aggression. Makepeace (1986) found that, among those students who had experienced NsPA, injuries were very common. More than half 109 of the women who had such experiences reported some form of physical injury, while 18% of the men reported sustaining injury. Rouse, Breen & Howell's (1988) data also indicates that physical injuries are common. Among their sample of college students, 29% of the women and 25% of the men who reported sustaining NsPA experienced "visible injuries" as a result. These figures contradict the conclusion, drawn by some, that dating violence is overwhelmingly mild in form (Sigelman et al., 1984). Generally such conclusions are based on the frequency distributions across the different types of aggressive actions experienced in dating relationships (usually derived from data from the Conflict Tactics Scale). However, the picture which emerges when one considers the injury data is very different. Several authors in the domestic violence field have indicated that this is one of the major weaknesses of the Conflict Tactic Scale, which is focused exclusively upon the frequency question (Browning & Dutton, 1986; Dobash & Dobash, 1988; Walker, 1989; Yllo, 1988). Further, when proponents of that measure do address issues other than the relative frequencies of different acts, they usually make arbitrary distinctions between "mild" and "serious" forms of violence without reference to respondents' actual experiences of injury. The emotional consequences of experiences with non-sexual physical aggression have also been overlooked in the 110 literature on dating violence. Studies of battered women indicate that violence in the context of marital and cohabiting relationships has profound emotional effects upon its victims (McCann, Sakheim & Abrahamson, 1988; Walker 1984, 1985). Walker (1984, 1985) advocated that a special subcategory of Post Traumatic Stress Disorder be created which would describe the specific reactions that battered women demonstrate in reaction to experiences of prolonged, severe violence in intimate relationships. She labeled this the "battered woman's syndrome" and suggested the following common signs: ”anxiety, fear, depression, shock, anger,...guilt, humiliation, confused thinking, intrusive memories, uncontrolled re-experiencing of traumatic events, rigidity, lack of trust, suspiciousness, hypervigilance and increased startle responses to cues of possible violence" (1985, p. 161). She maintains that disturbances in social functioning and in the capacity for intimacy, as well as mood swings are also common features of battered women's psychological reactions to their victimization. The emotional reactions to experiences of dating violence among a non-clinical sample of college students would not be expected to be as extreme as the syndrome that Walker describes, which is based upon the reactions of women who have experienced the most severe forms of violence. Yet there is evidence which suggests that experiencing serious emotional reactions to incidents of dating violence may be common. Makepeace (1986) found that 31.2% of the women who 111 reported incidents of NsPA said that they experienced "major emotional trauma” as a result. Only 12.2% reported experiencing no emotional reactions. For men who reported incidents of NsPA almost one in seven (14.9%) experienced "major emotional trauma" following the event while only a quarter (27.6%) reported no emotional aftereffects. Firestein (1987) studied the experiences of women who had been victimized in dating relationships. She reported that, in comparison with those who did not have experiences with violence, the victimized group reported significantly more depression. In addition, she also found significant correlations between women's reports of violence and their (a) overall emotional distress (as measured by the Symptom Check List-90), and, (b) interpersonal sensitivity (as measured by a scale on the same measure). Sugarman and Hotaling (1989) also report that four studies which have explored the relationship between dating violence and self esteem have found that victims consistently demonstrate lower self-esteem than do those who have not had these experiences. This same relationship has also been reported for aggressors, but not as consistently. It is clear from these few studies which have explored the emotional and physical consequences of non-sexual physical aggression in dating relationships that they are substantial and widespread. Further, Makepeace (1986) has indicated that assessing the consequences of dating violence is essential to understanding gender differences in these experiences, a 112 position which has also been put forth by a number of authors from the social control perspective in the domestic violence field (Berk, Berk, Loseke, & Rauma, 1983; Saunders, 1988; walker, 1989; Yllo, 1988). In order to achieve more than a superficial understanding of dating violence, it seems clear that future investigations must follow Sugarman and Hotaling's (1989) advice and move beyond the limited perspective of quantifying and comparing rates of different violent actions. Instead, the task calls for a more contextual approach. Obviously, a longitudinal design would be necessary to conduct an adequate assessment of the long- term psychological consequences of involvement in non—sexual physical aggression in dating relationships. But even in correlational group comparison designs, more consistent attention to the effects of dating violence would be an important step toward the expanded perspective that is required for a fuller understanding of the phenomenon. One of the goals of the present study was to address the physical and emotional consequences of non-sexual physical aggression in dating relationships among a college population. Gender Differences. As mentioned earlier, many of the authors who have been working in the field of non-sexual dating violence have assumed a gender-neutral perspective in theorizing about the causes of this phenomenon. This has been due, in part, to the over reliance upon the Conflict Tactics Scale, which 113 focuses exclusively upon a limited range of aggressive behaviors and excludes from consideration many of the contextual factors that are thought to be important in assessing gender differences in the experiences of dating violence. Studies of NsPA in dating relationships which have relied upon the CTS have, for the most part, indicated that males and females report about equal rates of victimization in dating relationships (Sugarman & Hotaling, 1989). Sugarman and Hotaling, in reviewing the results of over 40 studies of dating violence concluded that, although the differences are slight, women have tended to report slightly more often that they have been the victim of acts of violence. However, the gender differences which have emerged in rates of inflicting violence against a partner have been more substantial. Women have consistently reported higher rates of inflicting aggression against their dating partners than have men. In summarizing these differences found in the studies they reviewed, Sugarman and Hotaling reported that, "Almost four out of every ten women and almost one-third of the males are reported to have been violent at some point in their dating careers" (p. 8). In the field of marital violence, studies utilizing the CTS have resulted in similar findings. One national study based upon a large random community sample, conducted by Gelles and Straus and their associates created considerable controversy. The results of that study indicated that 114 female-to-male violence in intimate relationships was as prevalent as male-to-female violence (Straus, Gelles & Steinmetz, 1980). The report of this finding set off a heated debate about its meaning and validity. Reactions can be classified roughly into two major "camps". Many subsequent authors took the position that intimate violence is often "reciprocal” in nature, with women inflicting as much violence as do males (Shupe, Stacey & Hazlewood, 1987). Some of the people who took this perspective, (like Shupe et al.) see relationship violence from a ”systems perspective" which views male-initiated and female-initiated violence as basically similar in terms of the underlying causes of each. This perspective corresponds to the gender-undifferentiated models and theoretical stances discussed earlier, like Riggs and O'Leary's (1989), which have dominated the field of dating violence. In the other ”camp", the feminist/social control theorists have questioned the basic validity of Gelles and Straus' findings of similar rates of male and female-inflicted violence (Saunders, 1988; Walker, 1989; Yllo, 1988). As discussed previously, they see women's violence as being solely self-defensive in nature and have suggested that the research methods employed by Gelles and Straus (including, specifically the Conflict Tactics Scale) fail to adequately measure the phenomenon of domestic violence. No one in the dating violence literature has taken a position similar to that taken by these social control theorists in the domestic 115 violence field. Instead, the findings of equal rates of violence that result from the utilization of the CTS have most often been accepted at face value. And yet, there is evidence that calls into question the gender-neutral perspective that is prevalent in this field. Sugarman and Hotaling (1989) reviewed some of this evidence. They pointed out that two studies which assessed the gender differences in the severity of inflicted violence indicated that males were more likely to utilize more serious forms of violence (Makepeace, 1983; Lane & Gwartney-Gibbs, 1985), though one study which conducted this comparison did not find significant differences (Arias et al., 1987). However, in reviewing the research which focuses on the conseguences of dating violence, Sugarman and Hotaling concluded that the results are unequivocal: women were victimized much more severely than are men. They sustained more frequent and more serious injuries and consequences than did men, as reported in the section above. Makepeace's (1986) study had the specific goal of exploring gender differences in experiencing dating violence. In his sample of college students, the women who sustained violence reported three times as much mild injury, two times as much moderate injury and all of the reported serious injuries as compared to men. Women also reported more emotional trauma. In discussing his results, Makepeace addressed the possibility that men may systematically underreport their own violence. 116 This is a problem that has been discussed by a number of authors in the clinical literature on domestic violence. In reporting on their behavior, violent men have consistently been observed to minimize and deny their violent actions toward their partners, even in the face of irrefutable proof to the contrary (Ganley, 1981; Ptacek, 1988b; Sonkin, Martin & walker, 1985). In fact, a number of studies have been conducted which have empirically demonstrated this tendency of male batterers to underestimate and deny their violence (e.g., Browning & Dutton, 1986; Jouriles & O'Leary, 1985). But, if physically aggressive men underreport their violent behaviors, isn't it likely that physically aggressive women would do the same? Perhaps not. If men's violent behaviors inflict more physical injury than do women's, then, it could be argued, these behaviors may be more unacceptable and associated with stronger feelings of guilt or shame. If so, there might be a stronger tendency for men than for women to deny and minimize their actions defensively. In addition, walker (1989) and others in the domestic violence field have suggested that battered women tend to underreport their experiences of victimization and to overreport their own aggression. Such a pattern could be a defensive strategy, as well, which could help the battered woman to decrease her feelings of helplessness and to cope, emotionally, with the experience of victimization. Another gender difference that has been observed in the dating violence literature was reported by Lane and Gwartney— 117 Gibbs (1985) and by Makepeace (1986). In both studies, it was found that men reported having had experiences of non— sexual dating violence with more partners than did women. Lane and Gwartney-Gibbs concluded that, for the women in their sample, dating violence tended to be a one-time occurrence, whereas, for the men it was more likely to be a habit in their relationships with women. Additional important differences have been found in the motives that men and women have reported for their violent behaviors. Makepeace (1986) provided respondents in his study with a checklist of possible motives for their actions. He found that women described their behaviors as motivated by self defense significantly more often than did men (35.6.% vs 18.1%). Women also reported that they intended to "injure or harm" their partner more than men did (8.3% vs. 2.4%). In exploring this latter difference, Makepeace examined the possibility that the women who intended to do harm were motivated by self-defense as well. He found that, in fact, 69.6% of the women who listed this reason also listed self- defense. On the other hand, men reported significantly more often that their violent behaviors were intended to intimidate their partners (21.3% vs 6.8%). The direction of these differences in motivations is consistent with the predictions of the social control model, although that model maintains that self defense is the sole motive for women's violence in intimate relationships, whereas Makepeace found that only 35.6% of the women who inflicted violence reported 118 that they did so to defend themselves. Still, that was the most frequent motive given by the women. Additional evidence concerning the existence of important gender differences in the experiences of NsPA is apparent from the results of the studies summarized in Tables 3 and 4. Of the twelve studies that appear in both tables, all but 2 (Billingham & Sack, 1986; Cate et al., 1982) found differences between men and women in the patterns of correlation between the independent variables of interest and the measures of inflicted and sustained NsPA in dating relationships. It is clear from the above discussion that there are profound gender differences in the experience of non-sexual physical aggression in dating relationships. It is important to try to achieve a fuller understanding of the dynamics underlying these differences. Gender-undifferentiated models, such as Riggs' and O'Leary's (1989) clearly cannot provide an accurate understanding of the phenomenon of dating violence. It also seems likely that the CTS is not adequate to the task of assessing these differences. Understanding the different experiences that men and women have with dating aggression clearly requires moving beyond the simple comparison of rates of male and female—inflicted violence, since not only can such an approach never overcome the inherent problem of differential inclinations to report violent actions among males as compared to females, but it also simply does not address the crucial issues. 119 One of the goals of the present study was to explore differences in men's and women's experiences of dating violence by comparing: (a) the emotional and physical consequences of these experiences; (b) the role of attitudes toward violence and toward the relationships between men and women; and, (c) the importance of the balance of resources between the dating partners, as well as the existing and ideal balance of power in decision-making. Criticisms of the Conflict Tactics Scale (CTS) As discussed previously, the CTS has been used almost exclusively in the study of dating violence as the dependent measure. And yet, it has been criticized extensively in the domestic violence literature as an inadequate measure of intimate violence. walker (1989) maintained that the CTS presents a number of serious problems. First of all, because of its exclusive reliance upon the tally of reported frequencies of different violent behaviors, it is especially susceptible to bias stemming from the differences in male and female response patterns, discussed above, wherein males who inflict violence consistently underreport their violent actions, whereas women may tend to do the same in reporting their experiences of victimization. Secondly, the CTS doesn't allow for a contextual examination of incidents of violence. For example, it equates all forms of "hitting" -- not differentiating between a slap which does not cause any pain 120 from a blow which dislocates a jaw and knocks the victim down. Further, as discussed previously, it doesn't differentiate between actions taken for self-defense and those initiated for other reasons. Overall scores and prevalence rates based on the total score of the CTS compound these problems so that a person who once pushed their partner away with light force in self-defense and once threw a ring at them in an argument would achieve the same total score on the CTS as someone who beat their partner up twice and inflicted serious injury. Others have noted that, in addition, the CTS does not account for sexual assault at all - a fact which is bound to result in a picture which seriously underestimates the violence sustained by women in intimate relationships. As noted above, sexual assault is an extremely common form of intimate violence and is perpetrated mostly by men against women. Another, related oversight of the CTS is that it does not include any measure of physical restraint - being held down, having one's movement inhibited by one's partner. Kanin (1984) reports that this form of physical aggression is pp; most common form employed by the date rapists he studied during their assaults. It seems very likely that this form of violence is utilized frequently in other than sexually assaultive contexts also and that it would be more frequently employed by men than by women, due to their advantages in physical size and strength. As in the case of sexually 121 assaultive behaviors, omitting this type of physical violence from the measure probably serves to underestimate the amount of violence sustained by women and initiated by men. It seems clear that, in order to achieve a valid and accurate picture of violence in intimate relationships, these limitations of the CTS must be remedied. But, in the meantime, the CTS continues to be widely used to describe (or misrepresent, as the case may be) the violence in dating relationships. The Relationship Between Sexual Aggression and Non-Sexual Physical Aggression As described in the review of the sexual aggression literature previously, it is known that sexual aggression -- even its most severe forms -- occurs frequently in dating relationships. The evidence clearly indicates that dating relationships are the locus of a tremendous number of sexual assaults (Kanin, 1984; Koss, 1985; Muehlenhard & Linton, 1987; Warshaw, 1988). In addition, the evidence suggests that sexual interactions may be the impetus for many other forms of physical aggression (Lane & Gwartney-Gibbs, 1985; Kanin, 1984). Makepeace (1981) reported that anger over sexual refusal was given as the third most frequent type of disagreement that resulted in an incident of NsPA. And yet, most investigators who address the issue of dating violence do so without including measures of sexual aggression. One notable exception is that Gwartney-Gibbs and 122 her colleagues have addressed the issue of sexual aggression in their work and have advocated that it be included in the study of dating violence consistently (Gwartney—Gibbs & Stockard, 1989; Gwartney-Gibbs, Stockard & Bohmer, 1987; Lane & Gwartney-Gibbs, 1985). However, the majority of studies of "dating violence” reviewed here have pg; addressed the issue of sexual aggression. Furthermore, many of those who have included it, have measured it with one or two items, focusing only on the most extreme experiences of sexual aggression (Makepeace, 1986, 1989; Murphy, 1988) or asking such a general question, that it is not clear what was being assessed or reported (Bernard et al., 1985; Rouse et al., 1988). One sign of progress in the field, however, is that many more of the studies of "dating violence" published in the last three years have included a measure of sexual aggression than did those published earlier. The discussions in the literature about the relationship between sexual aggression and non-sexual physical aggression are even fewer. One exception is Makepeace (1989), who, in exploring the relationship between the consequences of dating violence and the stage of development of the relationship, found that, although there was substantial overlap, sexual violence seemed to be most prevalent and to account for most of the serious physical and emotional consequences that result from the dating aggression which occurred on first dates and during the casual dating phase of relationship. In contrast, non—sexual violence occurred more at later stages 123 of relationships and was most dangerous and severe at very committed stages. As discussed above, this bimodal pattern led Makepeace to suggest two different types of courtship violence: predatory and relational. He pointed out that the first type, which is primarily sexual violence, has been largely ignored in the literature on dating aggression. It is important that measures of sexual aggression be incorporated into any study of "dating violence" in order to get a complete picture of this phenomenon. It is known that much of the "violence" in dating relationships 1; sexual in nature. To define "dating violence” as distinct from the sexual aggression which is known to occur with such frequency in this context is to seriously misrepresent the phenomenon. One of the goals of the current study was to explore the relationship between sexual aggression and non-sexual physical aggression. This included a direct assessment of their association, but also involved a comparison of the patterns of association between each and the other variables of interest (e.g., physical and emotional consequences, attitudes about violence and about relationships between men and women, and the balance of resources and decision making power in relationships). The social control model predicts that these two forms of violence would be different expressions of the same phenomenon of male violence against women, both arising from common causes. If so, one would expect that the measures of acceptance of the norms of 124 violence and male dominance would be associated with both types of violence. Egycholoqlcal Abggg. Psychological abuse has long been acknowledged by practitioners and theorists in the domestic violence field as an important component of the patterns of abuse in violent relationships (Ganley, 1981; NiCarthy, 1982; Okun, 1986; Sonkin, Martin & walker, 1985; Walker, 1979). Okun maintained that, in clinical observations of battering relationships among cohabitating and married individuals, psychological abuse always plays a part in the pattern of violence that batterers direct at their partners. A number of authors have suggested that psychological abuse, in fact, accounts for many of the debilitating psychological effects of abusive relationships that women experience (Okun, 1986; Sonkin, Martin & walker, 1985; Walker, 1984). Walker (1984) went so far as to suggest that psychological abuse may be more damaging and painful in the long run than physical violence. The domestic violence theorists see the psychological abuse of women as one method, like physical violence, by which violent men attempt to achieve their goal of dominating and controlling their partners (Adams, 1988; Okun, 1986; Sonkin, Martin & Walker, 1985). There seems to be some agreement also as to how this psychological abuse functions. Ganley (1981), Adams (1988) and Okun (1986) have all 125 maintained that psychologically abusive statements and actions directed at a person over time, tend to undermine the self-concept and identity of the victim, thereby diminishing her ability and inclination toward self-determination, and making her more vulnerable to her partner's attempts at control. However, in spite of the consensus in the clinical domestic violence literature that psychological abuse plays an important role in violent relationships, little research has been conducted to substantiate this claim (Okun, 1986). walker (1984) has suggested that the difficulty in quantifying and measuring psychological abuse is the major impediment to research in this area. But the first obstacle which needs to be overcome in order to quantify psychological abuse is that of achieving a definition. Sonkin et al. (1985) in their discussion of domestic violence, distinguished between four different kinds of violence: physical, sexual, property, and psychological. They asserted that all four are "ways of controlling, dominating and intimidating another person" (p. 38). (For the purposes of the present study, in order to achieve a comprehensive view of the entire range of violent and abusive behaviors that occur in intimate relationships, both property and psychological violence will be considered to be components of psychological abuse.) Property violence includes any physical violence that is exhibited towards objects or non- human life forms in the victim's presence or environment. It 126 includes breaking property, pounding on the table, putting a fist or foot through the wall, breaking down the door, throwing food, and breaking dishes. Psychological violence includes, "explicit threats of violence, implicit threats of violence, extreme controlling types of behavior, pathological jealousy, mental degradation and isolating behavior" (p. 38). walker (1984) suggested that the definition of psychological torture which has been developed by Amnesty International provides an adequate objective definition of psychological abuse in violent intimate relationships. She described this definition: It includes eight areas of abuse. They are: (1) Isolation of the victims; (2) Induced debility producing exhaustion such as limited food or interrupted sleep patterns; (3) Monopolization of perception including obsessiveness and possessiveness;-(4) Threats such as death of self, death of family and friends, sham executions, and other vague threats; (5) Degradation including humiliation, denial of victim's powers, and verbal name calling; (6) Drug or alcohol administration; (7) Altered states of consciousness produced through hypnotic states; and, (8) Occasional indulgences which, when they occur at random and variable times, keep hope alive that the torture will cease (pp. 27-28). Not surprisingly, given the difficulties entailed in defining and measuring psychological abuse, the research on dating aggression has largely ignored this form of abuse. 127 Sugarman and Hotaling (1989), in their review of the dating violence literature state that, "although psychological strategies are the primary means of controlling another person, little work has focused on operationalizing this construct within the dating context" (p. 5). Thus, very little is known about the extent to which psychological abuse occurs in the context of dating relationships, or about how it is related to dating violence. One of the goals of this study was to develop a definition and measure of psychological abuse and to examine the role that this form of abuse plays in dating aggression. This involved exploring its relationship to sexual and non-sexual physical aggression as well as to indices of psychological well-being. The social control model would conceive of psychological abuse as another method, like sexual and non- sexual physical abuse, utilized by men who strive for dominance over women to achieve their goals of dominance and control. This perspective would expect that psychological abuse would be associated with sexual and non-sexual physical aggression, and that it would be related to the measures of attitudes about dominance and violence in similar ways. These issues were explored. Research Goals and Hypotheses This study examined the phenomenon of aggression in heterosexual dating relationships among a college population. 128 It expanded upon the previous work in this field in a number of ways. The social control or feminist perspective on violence and aggression in intimate relationships has not been very influential in the dating violence literature up to this point. The prevailing perspectives assume a gender- neutral stance in their analyses at worst, or, at best, they discuss sex differences in ways that minimize and obscure their importance. The few studies which have addressed variables central to this perspective have yielded mixed results, and yet there is ample evidence of substantial gender differences in the experiences of dating violence. Further, the differences that have been found are often in the directions predicted by the social control model. The major purpose of this study was to assess the relevance of this model to understanding the aggression and violence which occur in the context of heterosexual dating relationships. All of the hypotheses and research questions addressed pertained directly to this question. These fell into the following three main goal areas: (a) achieving a comprehensive view of the phenomenon of dating aggression, (b) assessing the impact of sex-role stereotypes, attitudes towards violence and aggression in relationships, resource distributions, and patterns of dominance upon the prevalence of dating aggression, and, (c) exploring gender differences in the experiences of dating aggression. This last goal involved, among other things, evaluating the effectiveness of the CTS in assessing gender differences in the experiences of 129 non-sexual dating aggression. Its performance will be compared to that of another measurement strategy developed for this study, which was derived from the CTS and was intended to be more sensitive to the qualitative aspects of the experience of NsPA. Each of these goal areas are addressed separately below and the hypotheses and research questions relevant to each area is presented. AgpigvinqggiComprehensive View of the Phenomenon of Dgpipg Aggression In this study, dating aggression was defined more comprehensively than in previous investigations. It was considered to include three components: sexual aggression, non-sexual physical aggression and psychological abuse. The social control model views all three forms of dating aggression as expressions of the same underlying social issues: the prevalence and widespread acceptance of dysfunctional cultural norms which support male dominance in relationships with women and which condone aggressive and violent means to achieve this dominance. If this is true, it is important to consider the whole range of aggressive, controlling behaviors which occur in intimate relationships between men and women to assess the full extent of the social problem. It is easier to avoid an issue when it is never acknowledged in its totality. That is inevitable when the definitions of the scope of the problem preclude perceiving 130 the common issues and patterns which run through the whole range of its different expressions. To reiterate, for the purposes of this study dating aggression (DA) was defined as the combination of (a) sexual aggression (SA), (b) non-sexual physical aggression (NsPA), and (c) psychological abuse (PsyA). Sexual aggression (SA) is defined as any attempt via coercion, threat of physical force or actual physical force to engage in any sexual act with another person against that person's will. Non-sexual physical aggression (NsPA) is defined as any act of physical aggression directed against another person which has the clear potential for causing physical injury or harm to that person, regardless of whether any injury is inflicted. Both SA and NsPA are assumed to vary in frequency as well as across a continuum from milder to more and more violent and potentially endangering acts. Psychological abuse (PsyA) is defined as any aggressive act (excluding NsPA acts) that has the clear potential for intimidating, controlling, demeaning or socially isolating another person (where "controlling" is defined as forcing another person to act against their will in cases where they clearly have the right to self—determinationa ). PsyA includes violence against property, threats of violence and verbal aggression. PsyA is also assumed to vary along a 3 This restriction would exclude, for example, attempts to keep another from stealing, or actions directed towards minors that are legitimate controls of their activities. 131 continuum from mild to severe, but the difference between levels of severity hinge more upon the frequency and chronicity of the psychologically abusive actions rather than upon a hierarchy of more and more abusive acts. The study of psychological abuse has been overlooked in the dating aggression field. Even in the domestic violence literature, a systematic assessment of the role of psychological abuse has been absent, due in large part to problems of definition and measurement. In the effort to achieve a more comprehensive definition of dating aggression which includes psychological and sexual abuse, one of the goals of the current study was the development and testing of a new measure of psychological abuse, The Inventory of Psychological Abuse“. This measure was developed for use in this study and a concurrent study of follow-up services for battered women leaving a shelter. Hypotheses, research questions and analyses. Some of the questions and goals of this study were best addressed by specific hypotheses which were tested, but others, because of their exploratory nature, were best addressed by more general research questions which did not contain such specific predictions about the relationships which were expected. ‘ The inventory of Psychological Abuse was developed jointly by the author and Cris Sullivan, Ph.D., Psychology Department, Michigan State University, East Lansing, Michigan. Dr. Sullivan is the director of The Community Advocacy Project, a program designed to provide follow-up services to women leaving a battered women's shelter. 132 The following hypotheses and research questions were addressed to explore the relationships between SA, NsPA, and PsyA. Hypothesis 1: For, males and females, inflicting SA, inflicting NsPA, and inflicting PsyA were expected to be significantly correlated with one another, as were sustaining SA, sustaining NsPA, and sustaining PsyA. Additional research quegtjons. In addition to this main hypothesis in this area a number of other comparisons and analyses were conducted. 1. The patterns of relationship between inflicting each of these three types of dating aggression (dependent variables) and the independent variables predicted by the social control model to be related to the phenomenon of dating aggression were explored. These variables are: sex—role stereotyping, acceptance of interpersonal violence, aggressive/adversarial attitudes towards male/female relationships, patterns of resource distribution in the relationship, and the goals and actual patterns of dominance in decision making. Each of these variables and the social control model's predictions specific to each are discussed more fully in the following section. 2. The self-reported emotional consequences of inflicting and sustaining SA, NsPA, and PsyA were compared. Respondents were asked to rate the seriousness of the emotional effects of their experiences after the measure of each of the three forms of DA. Analyses were conducted to compare the 133 differences between the emotional consequences reported by the respondents to result from their experiences with each type of DA. 3. The physical consequences of sustaining SA and NsPA were compared. Respondents reported on injuries that they sustained as a result of their experiences of both types of DA. Analyses were conducted to compare the physical consequences reported by the respondents in their experiences with each type of DA. 4. The relationships between inflicting and sustaining each of the three forms of DA and measures of psychological symptoms were explored and compared. It was expected that victims (i.e., those who report sustaining abuse) of all three types of DA would report significantly more problems with depression, anxiety, interpersonal sensitivity, somatization, hostility, and a global index of psychological symptomatology, all of which are scales included in the Brief Symptom Inventory (Derogatis & Spencer, 1982). The psychological symptoms exhibited by those who inflict DA were also examined. This work was exploratory, since little attention has been given to the psychological symptoms experienced by those who inflict DA in a non-clinical sample. Also, little previous work has focused on the psychological correlates of men's experiences of sustaining DA. 134 Assessing ghe Impgct of Attitudes and Patterns of Dominance. As discussed at length previously, the feminist/social control model suggests that a number of specific variables are related to the occurrence of all three forms of dating aggression. The present study attempted to assess the relevance of several of these to the rates of violence and aggression in the dating relationships of college students. The following variables were examined: sex-role stereotypes, attitudes towards violence and aggression in relationships, resource distributions, and patterns of dominance. Sex-role stereotyping - Discussion and hypotheses. Each respondent answered questions designed to measure their sex- role stereotyping attitudes and they also answered the same questions as they thought their partner would. The feminist/social control model predicts that men who are aggressive and violent in intimate relationships engage in these behaviors to achieve dominance and control and that they tend to maintain traditional views of the relationships between men and women, and of gender-roles in general. The strongest determinant of DA, however, was expected to be the discrepancy between the two partners' levels of sex-role stereotyping, since, if they agreed that the men should be dominant and women submissive, then fewer conflicts about control would arise, and the likelihood of violence would not be as high. Thus, the discrepancy scores were expected to be associated with the occurrence of all three forms of DA, with the highest rates being reported for the couples where the 135 male was described as being more traditional in his sex-role stereotypes than was the woman. Sigelman et al., (1984) found that the rates of DA were higher in both directions of discrepancy, not just when the males' attitudes were more traditional than the females. The feminist/social control theorist might account for this by suggesting that in both cases, the woman's behavior may not be meeting the man's expectations and, in such cases, he may resort to aggressive means of control even when he endorses less traditional sex— role attitudes than does she. Even so, it seems likely that the model would expect all forms of DA to be somewhat less likely the less traditional were the male's sex-role stereotypes, regardless of the discrepancy score. Hypothesis 2: For males, it was expected there would be a small but significant association between sex—role stereotypes and inflicting all forms of DA, such that the more traditional his stereotypes, the more experiences with inflicting DA he would report. For females, it was not expected that these would be associated. Hypothesis 2a: For males and females, the magnitude of the sex-role stereotypes discrepancy score (regardless of the direction of the difference) was expected to be associated with higher rates of inflicting and sustaining all three forms of DA. Rates were expected to be highest, though, for the discrepancy scores which reflected that the male's stereotypes were more traditional than the female's. Acceptance of aggression — Discussion and hypotheses. The feminist/social control model further predicts that patriarchal social norms result in attitudes which condone male-aggression and violence as acceptable means of achieving control in their intimate relationships. Thus, attitudes 136 which reflect an acceptance of violence in interpersonal relationships and an adversarial/aggressive view of male- female relationships in general, would be expected to be associated with higher rates of inflicting all three forms of DA among men. However, for women, the expectations are different. The model views women as the victims of all forms of DA, and their aggressive behaviors are seen as self-defensive. So women's DA actions would not be expected to be a reflection of their own attitudes, but rather a reflection of their partner's behavior. However, for women who are more deeply involved in committed relationships with men who exhibit aggressive behaviors (as compared to those who are in casual or first-date relationships), it may well be that they would come to accept interpersonal violence and to see aggression between men and women as normative. This could serve as a way of coping with the aggression that they experience in their own relationships. Thus, it was expected that a correlation between women's attitudes of acceptance of violence and aggression in male/female relationships and their involvement in DA, if demonstrated, would be accounted for by those women who experienced DA in the more committed and involved stages of dating relationships. Hypothesis 3: Men's degree of acceptance of violence in interpersonal relationships and of an adversarial/aggressive perspective about male-female relationships in general, were expected to be correlated with higher rates of inflicting all three forms of DA. 137 Hypothesis 3a: Women's scores on these attitude measures were expected to be unrelated to their rates of inflicting violence at low levels of involvement in their relationships, but to be positively associated at high levels of relationship involvement. Eggources and dominance in decision-making - Discussion and hypotheses. The feminist/social control model maintains that a dating couple's distribution of resources and the extent to which the male endorses the goal of male dominance are important determinants of DA. In this study, dominance goals were measured with a dating decision-making questionnaire that assessed how the respondent thought a variety of decisions ideally should be made by a dating couple. Responses range from, "the man should always decide” to ”the woman should always decide". This was used to assess each respondent's ideal power distribution. The same questionnaire was used to measure the actual power distribution in the respondent's dating relationship and a measure of the discrepancy between the ideal and actual distributions. Resources were defined to include a combination of age, educational status, GPA, parental income, personal monthly disposable income, and ratings of physical health, self-confidence, physical attractiveness, and social activity. Respondents rated themselves and their dating partners on each of these and discrepancy scores for the two partners were computed. The feminist/social control model suggests that men who maintain a goal of dominance in their intimate relationships 138 with women, and who do not have legitimate resources to afford them this position, resort to violence to assert their power and control. It also maintains that, in relationships which are not terminated at early stages because of the aggression, all three forms of DA, in fact, would result in the man having more power in the relationship derived from the influence of intimidation. For women, the model predicts a different pattern of responses. Women who report having equal or more resources than their dating partners, and who describe their dating partners as having expectations of dominance for the relationship (assessed in a separate item) would be expected to report more of all three forms of DA (both inflicting and sustaining, since this model maintains that, for women they both reflect self-defensive actions). Women who report sustaining DA in their relationships would be expected to describe them as being more male dominant (because the abuse, through intimidation, lends their partners more power) and, as a result, to express stronger desires for increased decision making power for themselves. So, for the women, as well as the men, it was expected that higher levels of all forms of DA would be associated with higher discrepancy scores on the decision making questionnaire between the level of power desired and the level of reported actual power over decisions. The important distinction here, though is that women's goals were expected to describe equal or slightly male dominant ideals; and men's, to describe more clearly 139 male dominant ideals. In other words, though the discrepancy scores may be of the same magnitude, the men were expected to maintain goals that describe a dominant-submissive relationship, while the women were not (though they too will report discrepancies between the amount of influence they want and what they have). These predictions are expressed in the following hypotheses: Hypothesis 4a: Males who expressed dominance goals for male-female relationships and who reported having fewer resources than their partners were expected to be more likely to report inflicting all three forms of DA. Hypothesis 4b: Women who reported having equal or more resources than their dating partners and who described them as having expectations of dominance were expected to report both sustaining and inflicting more of all three forms of DA. Hypothegis 4c: For both women and men who reported relationships at moderate and high levels of commitment, their experiences with all three forms of DA were expected to be associated with more male dominant actual power distributions in their relationships. Hypothesis 4d: Among both women and men, reports of experiences with DA were expected to be associated with greater discrepancies between the amount of influence they actually experienced and their reported ideals (i.e. they would report having less power than they wanted). Hypothesis 4e: Among those who reported experiences with inflicting DA, the males were expected to report male dominance ideals, while the women were expected to report ideals of equality in decision making. 140 Assessing the overall impact of attitudes, resources and dominance - research questions. In order to assess the combined impact of attitudes, resources and relationship dominance patterns, predicted to be important correlates of dating violence by the social control model, the following predictions were explored. 1. Discriminating between men who inflicted SA, NsPA, and PsyA and those who did not. It was expected that the male respondents' scores on the variables (specified below) which the feminist/social control model links to male violence and abuse against women would discriminate well between those males who reported experiences with inflicting the three types of DA and those who did not. The variables which the social control model sees predictive of higher levels of all three forms of male aggression toward dating partners are: (a) lower absolute levels of personal resources (e.g. available spending money, academic performance, popularity, physical health), (b) fewer resources in comparison with partner's, (c) endorsement of the ideal of male dominance in relationship decision making, (d) higher levels of reported actual decision-making power in one's relationship, (e) reportedly having less decision- making power in one's relationship than one's ideal, (f) maintaining traditional sex-role stereotypes, (g) reported discrepancies between one's own and one's partner's sex-role stereotypes, (h) endorsement of attitudes characterizing 141 male/female relationships as manipulative and antagonistic, and, (i) acceptance of violence toward women. In line with the feminist/social control model's predictions, these variables were expected to differentiate between those men who reported inflicting SA and NsPA against a dating partner, and those who did not. They were also expected to differentiate between three groups of men who reported engaging in low, medium and high levels of PsyA. 2. Discriminating between women who inflicted SA, NsPA, and PsyA and those who did not. As discussed before, the feminist/social control model views female-inflicted aggression as primarily self-defensive in nature. Therefore, it was predicted that women's aggression would be significantly related to the following variables which would describe high risk relationship contexts for experiencing sustained dating aggression: (a) reporting that partners had fewer resources than themselves, (b) rating partners as subscribing to more traditional sex-role stereotypes, (c) larger discrepancies between their own and their partner's sex-role stereotypes, (d) rating partners as subscribing to more male—dominant decision making ideals, and, (e) reports of lower levels of actual decision-making power in the relationship. 142 Gender Differences - Evaluatinqathe Performance of the CTS The social control model, as described previously, makes very different predictions about male and female inflicted DA, and the results of addressing the research questions and hypotheses listed above were expected to delineate many of the gender differences in these college students' experiences with the three different forms of dating aggression. In addition to these issues, however, the following predictions were tested: Hypothesis 5: More women were expected to sustain more serious forms of sexual aggression and physical battery than men, to experience more psychological abuse, and to experience more injuries and more negative psychological effects associated with these DA experiences than did men. In addition, women were expected to rate the incidents of NsPA victimization that they experienced as more dangerous and to indicate more frequently than men that their acts of inflicted NsPA abuse were self-defensive. As discussed earlier, social control theorists have suggested that, in the domestic violence field, reliance on the CTS as a measure of violence has been largely responsible for the underestimation and lack of attention to existing gender differences. The CTS is seen as too insensitive to the qualitative differences in the experience of NsPA. It equates a mild push (as, for instance when one partner pushes the other away in a disagreement to keep that person from putting an arm around them) with shoving someone down the stairs: It equates throwing a book of matches with throwing a brick. And yet, clearly these are not experienced as equally violent events by the participants. Also, the CTS 143 ignores the issue of what the physical effects of the violent action were to the victim, although these effects would obviously be very strong determinants of the meaning that the partners give to the event and the impact that it has on the subsequent relationship. These issues are particularly important as investigations move beyond questions of incidence to explore the interpersonal causes and consequences of violent actions. It seems likely that a measure of non-sexual physical aggression which incorporated these qualitative differences would be better suited to the task of assessing the meaning and impact of NsPA in these interpersonal contexts. What is required is an examination of the effect that the consideration of these variables might have on the resulting picture of the phenomenon of non-sexual physical aggression, particularly as it reflects gender differences. One of the goals of the present study, was to compare the performance of the CTS with that of another strategy for measuring NsPA which included measures of perceived dangerousness and sustained injuries in the overall violence score. The new measure was derived from the CTS. It included all of its current items, but in desegregated versions (e.g. "kicked, bit, or hit you with a fist” was made into three separate items). In addition, new items were included to assess the "wrestling-type" behaviors and physical restraint forms of NsPA which the CTS overlooks. In the new measure, for each violent act, the respondent was asked the number of 144 times this occurred, the perceived dangerousness of the most serious occurrence, and whether or not any injury did, in fact, occur. The perceived dangerousness was rated on a 6- point scale from "could have caused no physical harm or pain”, through "could have caused lingering pain and minor injuries" (which are specified), to "could have been life threatening". Legally (and logically) it is true that a violent act may be defined as such without reference to the victim's (or perpetrator's) perception of its dangerousness and without reference to whether or not any injuries occurred as a result of it. However, in research on violence in intimate relationships judgements about the "violence” of an action must be made in the absence of being able to actually witness the specifics of an action to determine its force and potential for inflicting harm. It is those specifics that determine the dangerousness of an action. And the concept of violence is intricately bound up with those of danger and threat of injury. It is not that, in the absence of any perceived potential for inflicting harm, or of any actual injuries, a violent act has not occurred. But without assessing these qualities as a component of the act, the differences between trivial acts of violence and life threatening ones are obscured. And, in trying to understand the impact of violence on intimate relationships and upon psychological functioning, these differences were judged to be crucial. 145 Certainly, it was expected that it would be problematic to ask the participants involved to judge the dangerousness of an action. It was expect that perpetrators would usually judge the dangerousness of an act to be much less than the victim. But, even so, differences between dangerousness ratings of different actions within the perpetrator's report were expected to be meaningful, and some measure of these qualitative differences between acts, however inaccurate, was felt to be preferable to overlooking the differences entirely. The ultimate test of this strategy of measuring NsPA, however, would be in comparing its ability to account for observed differences in measures of the theorized ‘outcomes' of NsPA (e.g., reported emotional distress, psychological symptoms, and relationship power distributions) with that of the CTS measure. This comparison was made in the current study and the predictions are stated in the following hypothesis. Hypothesis 6: It was predicted that the NsPA scores which resulted from the qualitative scoring strategy, as compared with the CTS-based scoring strategy, would be more highly correlated with the measures of reported emotional distress, and with scores on the depression, anxiety, interpersonal sensitivity, somatization, hostility, and global severity scales of the Brief Symptom Inventory, as well as with the measures of decision-making power distributions in relationships. In addition to testing this hypothesis, the gender differences in the reports of NsPA experiences which resulted from the new measure were explored. This focused on the 145 Certainly, it was expected that it would be problematic to ask the participants involved to judge the dangerousness of an action. It was expect that perpetrators would usually judge the dangerousness of an act to be much less than the victim. But, even so, differences between dangerousness ratings of different actions within the perpetrator's report were expected to be meaningful, and some measure of these qualitative differences between acts, however inaccurate, was felt to be preferable to overlooking the differences entirely. The ultimate test of this strategy of measuring NsPA, however, would be in comparing its ability to account for observed differences in measures of the theorized ‘outcomes' of NsPA (e.g., reported emotional distress, psychological symptoms, and relationship power distributions) with that of the CTS measure. This comparison was made in the current study and the predictions are stated in the following hypothesis. Hypothesis 6: It was predicted that the NsPA scores which resulted from the qualitative scoring strategy, as compared with the CTS-based scoring strategy, would be more highly correlated with the measures of reported emotional distress, and with scores on the depression, anxiety, interpersonal sensitivity, somatization, hostility, and global severity scales of the Brief Symptom Inventory, as well as with the measures of decision-making power distributions in relationships. In addition to testing this hypothesis, the gender differences in the reports of NsPA experiences which resulted from the new measure were explored. This focused on the 146 differences in reports of dangerousness and injuries and on the added items assessing physical restraint and "wrestling" types of physical aggression. As an additional tactic for exploring gender differences, a set of open-ended questions were included in the measure of NsPA to assess the contextual features of aggressive interactions which are missed by more traditional quantitative measures. A number of feminist/social control theorists have suggested that more qualitative data collection procedures are required to achieve a valid picture of the experiences of intimate violence (Dobash & Dobash, 1983; Yllo, 1988). They have suggested that the reliance on so-called "objective" measures in general (not just the CTS) has been responsible for the underestimation of gender differences in the experiences of aggression in intimate contexts. Yllo argues for a combination of qualitative and quantitative measures, suggesting that the inclusion of qualitative ones allows the participants "to bring in information and analyses that they think is helpful in making sense of what happens" (p. 34). This view acknowledges the expertise of the participants and suggests that, by assuming that perspective, one can avoid some of the blind spots created by one's own values, predictions and world view. In the current study, the inclusion of open-ended questions was risky because of the survey format. It could not be guaranteed that a sufficient number of respondents would provide enough written information to augment the 147 understanding of their experiences. And of course the absence of an opportunity to probe further, which is available in face-to-face interviews, also severely limited the quantity and quality of the information which could result from open-ended questions. However, the benefits which might have been gained (especially in exploring answers to questions which might arise from unexpected results and in generating further hypotheses) led to the decision to include these questions. It was hoped that information from these questions would be helpful in exploring gender differences in the experiences of DA further. In the following chapter, the specific procedures and measures used in this study will be described. 149 times. Subjects and Procedgres The subjects for this study were junior and senior undergraduate students, attending The University of California at Berkeley during the Fall Semester of 1989. There were a number of reasons to focus upon college students, in particular, in studying dating violence. First of all, college students are easily accessible for study and can be expected to include a relatively large number of single members who are involved in dating relationships. Secondly, studies of sexual aggression have indicated that the developmental period which includes the college years is a time of very high risk for experiences of violence in intimate relationships (Koss et al., 1987; Stets & Straus, 1989; warshaw, 1988). A mail-out, mail-back survey strategy was employed for this study. A computer generated mailing list was used which consisted of 200 women and 200 men randomly sampled from each of the junior and senior classes enrolled in the University during the Fall Semester (800 potential respondents in all). Foreign and immigrant students were excluded from initial mailing lists, if their Visa information was included in their student records, since cultural norms pertaining to gender roles were of central interest and they are undoubtedly culturally specific to some degree (3.8% of undergraduates at this university fall into this category). 150 Juniors and seniors were targeted for study because the incidence of dating aggression has been found to be positively associated with the length of a relationship, the depth of commitment, and the level of intimacy between dating partners (Carlson, 1987; Cate, Henton, Koval, Christopher, & Lloyd, 1982; Henton, Cate, Koval, Lloyd, & Christopher, 1983; Roscoe 5 Benaske, 1985; Stets & Pirog-Goode, 1987), as well as with age (Henton, et al., 1983; Stets & Pirog-Goode, 1987). So, it was expected that dating aggression would be more common among the upperclass students. Students were asked about incidents of aggression that they had experienced in a dating relationship during their college years (or during the past four years, if the respondent was over the age of 24). A series of screening questions guided respondents to focus on the one relationship during that time in which they had experienced the most, or most serious, physical (sexual and/or non-sexual) abuse. Focusing on more than just current relationships was expected to further increase the frequency with which incidents of dating aggression would be reported, while restricting the focus to the college years was important to ensure that the respondents were describing relationships that were not so far in the past that their characteristics had been forgotten. Although a mail-out, mail-back survey strategy risked the possibility that the resulting sample would differ from the population in unknown ways, this was chosen as the best 151 method of study for a number of reasons. First of all, it is important to assess the factors associated with and consequences of intimate violence in non-clinical samples. This method was chosen over in-person interviews because the relatively low incidence of sexual aggression and non-sexual physical assault in the general population requires that a large number of respondents be included in the sample. Conducting in-person interviews with such a large number of students would have been beyond the scope of this study. In addition, the mail-out survey strategy was not expected to be any more or less susceptible to sampling bias than the more common strategy used in this area of research to reach non- clinical samples -- that of surveying students in various introductory classes. Another reason for choosing this strategy was that previous studies of dating aggression which utilized this method have yielded valuable information. Investigators have reported that the resulting samples have not been found to differ from the population in ways that would threaten the validity of the results (Gwartney-Gibbs & Stockard, 1989; Gwartney-Gibbs, Stockard & Bohmer, 1987; Lane & Gwartney-Gibbs, 1985). Participation in the study was solicited by mail. Two sets of mailing labels were made, one of which was used to mail out the surveys. The other was used to send follow- up/reminder cards to the potential respondents. The initial mailing included (a) a letter inviting students to participate in the study, (b) the questionnaire, (c) a 152 consent form, (d) a stamped self-addressed return envelope for the questionnaire, and, (e) a separate one for the consent form. (Copies of the questionnaire, letter and consent form are included in Appendix A.) Since the consent forms were returned under separate cover, there was no way to determine the respondents' identity. Annonymity was thus ensured. The initial mailing occurred at the end of the semester. The respondents were asked to return the completed questionnaires by a date three weeks after the initial mailing. Five to seven days later, postcards thanking respondents who had returned questionnaires and reminding others to do so, were sent to the same 800 students. . In an attempt to increase the response rate, an incentive was offered to the students for filling out the questionnaires. They were given the opportunity to participate in two drawings, each for the award of a $500 gift-certificate from a local department store. Respondents were eligible for the drawings by returning their completed consent forms which stated that the student had completed and returned the questionnaire (although this was not discernible by the investigator, since the annonymity of the returned questionnaires would not permit verification). The consent form also included space for the respondents to provide information about how they could be contacted on the date of the drawings. The drawings were phased such that respondents who returned their completed participation agreements (and, 153 hopefully, questionnaires) in time for the earlier drawing were also included in the second one, thus providing an incentive for a quick response. A total of 294 questionnaires were returned, resulting in a return rate of 36.8%. Of those, 21 (7.1%) were eliminated because the respondent had had no dating experiences. An additional seven respondents (2.4%) were eliminated because they described either lesbian (N=3) or gay (N=4) relationships. Obviously, there were not enough of these individuals to conduct separate analyses. However, their responses indicated that rates of DA were unusually high among this small group of individuals. Among the seven, four indicated having experienced either SA (2 gay men) or NsPA (2 lesbian women) in their relationships. Although it is probable that this group is entirely unrepresentative, these rates suggests that the experiences of dating abuse in gay and lesbian relationships is an important area for study. An additional 14 questionnaires (4.8%) were eliminated because of concerns about validity: one because of excessive missing data, one because the instrument was clearly filled out facetiously, and, 12 because the respondents initially responded affirmatively to screening items which indicated clear experiences with either NsPA or SA (#74, 75, 80 or 81), and then changed those answers to indicate that they did not have experiences with these forms of DA. This last group was eliminated because of doubts about whether or not the dating relationships they described were really non—abusive. The 154 decision was made to drop them from the sample rather than to risk including them in the non-abusive comparison group erroneously. The final sample included 252 dating heterosexuals. Table 7 describes this sample in terms of the distribution of race, gender and class standing. The respondents' average age was 21.9. The majority of the sample reported that they lived off-campus in a privately owned apartment or house (68.7%). But a sizeable minority (21.4%) reported living in group living situations with other students, either in sororities, fraternities, co-ops or dorms. Finally, 8.7% reported that they lived at home with their parents and 1.2% indicated that they had some "other" living arrangement. The sample was significantly non-representative of the wider campus community in a number of ways. Even though over one-third (35.2%) of the respondents were non-white, as Table 7 indicates, whites were overrepresented in the sample when compared with the entire student body, as were women and seniors. But, in terms of age, living situation and parental income (7 = $54,000) they were representative of the population. Measures. The questionnaire used for this study is included in Appendix A. It contained the following measures. Table 7 155 Distribution of Sex, Class Standing and Race in the Final Sample. Asian/ Pac.Isl— anders Black Hispanic White 0_th_¢£ Totals Totals (Percentages in Parentheses. ) Jr.'s ll(4%) 2(1%) 2(1%) 30(12)- l(1%) 46(18) Sr.'s 30(12) 4(2%) 8(3%) 72(29) 3(1%) 117(46) J_r_._'§ ll(4%) l(1%) l(1%) 18(7%) _0___ 31(12) 163 (65%) Women 89 (35%) Men E N Mis gay;§, Lag Totals ll(4%) 63(25) l(1%) 8(3%) 3(1%) l4(6%) 42(17) 1 163(65) 0 4(2%) 57(23) 1 252 156 Demographic and personal information. The instrument included six items which assessed the participants' gender, age, racial/ethnic background, year in school, type of residence, and their estimates of their parents' annual income. This portion was based largely upon a questionnaire developed by Gwartney-Gibbs & Stockard (1989). Measure of non-sexual physical abuse (NsPA). The measure of non—sexual physical aggression, designed for this study, consisted of items #123 through 137 and 141 through 155 in the appended questionnaire. This measure was a substantially augmented version of the Conflict Tactics Scale (Straus, 1979). It assessed all of the information which is covered by The Violence Scale of The CTS, with four additional items concerning forms of violence overlooked by the CTS (#127 - 129, 135). In addition, the new measure assessed the perceived dangerousness of the most serious acts reported in each category and the occurrence of injuries resulting from each type of action. These additions were described and discussed in the previous chapter. Also included were questions about the severity of injuries, and the emotional and relationship consequences of experiences with NsPA, as well as a handful of open—ended questions which assessed the respondents' understanding of the context and causes of non- sexual physical aggression. Table 8 describes all of the scales and variables which were derived from items concerning the three forms of dating 1157 AotzcouV Apnu.uec5.< a.xuoncotuv sausasoo «econ. pec_auase ace beau..5c_ goon A.ueucoaec aeuce_ceaxo o>_usne >ssesxou 50 «05>» ucoceaw_u so cones: on» .e._v sentence «sou. 50 comes: as» we can on» .0 A an anecov .eaoc cu ocaeoeca sonco> we museums eacsoo -ceuc_ oc_>o: sac» uc_ocec upon .osxeu sseeucom >..ou_e>:5 ace e>_ucoou so once; a mousse:— .ouce.o_> so neoco_c05xe see oce>om no N ace F M mucous o>og a~ unsung psmx “cutaneous; as: mucoucoaooc on» *0 ~\— unaduc_ a» some maroon ucs>ax e>ec A— asocov aco.o_>-c02 "onsets caucuses ace a_uo_> goon 505 35.35 was 2.5 830.. 350355- ca 35:0 no 3383205 cocoon seem .auesuc_u an ue~>seca useu_ 9cm: .m o» p seem «cocoa .nn. ou n~P useu_ concedes euoa dds ou ce>wu ucpuac «moca30coocav «macaw: as» .ueusaeou cocoon necaeuasa ace noun..wc. goon .neucoaoc egos newcsqc. sea 5. ~ >9 pe..5_u.ss ace .ucpuec eeocasoconcev «someo- uaes ecu >5 no..5_u.:s .veuco_c05xe aucon_oc_ we cones: 0:» so can 0;» arms. u agas< a.xoenc05uv pausasoo cocoon nec_euesa ace venom.»c. sues .5_ceco.ue.ec ecu uc_csp 3:039: :35 :- 50 .395: 05 so .5. of mmmmmflummum .acowuousasou to; pocmwep menace oco>ea see p._: .osoum ocszoz co scans onsets oro>om ac. u._x .uco.o_>-coz .Aouv no: to .oue .u_w_uoau oc_;uesoa on o» o» ocmucseuua .omceaeo.*.oa .stoseom ocmussoc. sconce; 0.9.uaoa so once; a ocmco>oo scoop ec_z a.oc_ceuoecsu 05.. coon e>og p.300: 0» sacs; co c.a5 oc pounce e>eg p.300: eocw nemceu .cocucoa >9 acoucoaaec unc_eue neuuec_p ecom>ezon acesom> so aaocwsocoucon po>_eucoa so menace: .pecmeuusu «avenue. Lev ace anecasocoucou ue>_euco5 new ocmuzu_es mousse:- oe.< .ouo .ucez .55. co oc_ncen to oc_uo_su ace oc_xo;o uc_eg ac...:a uuc_c_ecuaec .>..ao_a>;5 «3.5 .osoum .o_uuoc uu_.ccou a. mac». ..< .oue can to oc_cx . sum: ac_cou.otsu .us_c . su_z oc_uu_; .uc_xu_¥ ..c_ssaro .uc_sa:5 .ouasuc_ commmwummmm x~mop .aoto a «mono “mums >o>tsm meuce_co5xm seaxom uman< .oaxwm.aa mm~-vv~ xeuu><5wzv o5aore s_uu_> ecu 5n—-m~P Acucwo<5mzv museum conceawo omsn< soo_a>z5 .osxom-c02 eo—-0mp eocodom> oc_uo_s~c_ Law sconce“ ump-n~p mmocm50comcoo snp-n~p nucs>szv e.oow eoce.o_> 30: sup .onp Aesop .esacumv .en.-on. swag. osaum .o_uoo» uu_.ecou .o~..n~p oaas< .oo.a>s5 .aaxo..coz ._ Ha among messe_ce> oasn< oc_uea we aco_u5_coeeo m o.neh 1fi58 .kooo._ . .n anotac sa_=uxkooo.. w ac. coo—.F . .~ 530.03 euecouos "boo—.p v eecooe .euOu <5. o>ec AP asocuv so; "unseen cepcewwo ace s_uo_> cues co5 .Au0u0_swc_ Lew m-. unoc_euese co» o~5.uec5.< «Loon—code 635980 3.63 vouu_:c. ace .8533» zuoo .eeeu. eeccu «segue ecooe sea. eoece>< .ueuo_.mc_ co» ems. “pec_euese Law n~o.uecad< u.soencocuv.ueuaceou eecooe neuo_.~c. ace voc_eua:a guom .eseu_ acu_e accrue ecooe seu_ eoece>< ..uooo_.cc_ to. mes. unoc_-u.:. to. ase.u.g5.< a.xoenc0cuv.ueusasoo eocooe secmeuese ace nouu_.~c_ coon .mEo». o>_e eaocue ocean seu_ ooeco>< .xuouu_.cc. rat has. unoc_.usa. toe so»...sa.< e.xoenc0cuv .ueusnsou eocooe nec_euasa ace neuu_.wc_ zuOO oufiflum 06m: ““0500 0500a Eflum 000L0>< .Aveuo_sec_ cow Ans. “successes cow oos.uec5.< e.xoencocuv.ueusasou necooe poc_euase ace so»o..»c_ cuom .eeeu_ e>_w neocoe etooa an». ooece>< .uoso.scc_ to. as». x noccoumaa to. .o—o.uec5.< e.xuenc0cuv.ueusnsoo cocoon sec_euese a pouu_swc_ cuon .ese». on «mecca ecooe so», uneco>< .eoc05 .eo_e>c5 uc_>so>c_ aucos_9c_ ecoaoc an nsocuv ece>ee "comuceou .eo_a>c5.coc we eucop_uc_ acoaoc aw 550cc» u._s "oueecooe e>ec aw asocov neu.apuo_>-coz ”unseen e_uu_> com .pueocoue mum e>ec aw 550cuv onsets e>_esn< “on eecoue mum e>es A" asocuv 550cc e>_esne-coa nuances conceaao co5 ammumflummuu .eco_ueusnsou cuoo neu_e.uos>.coc coosuen eue_uceceew_u eaaocu s_uo_> .QJOLu coco cm edges» *0 nxp «assoc. ou voc_meu .eocooa deuce <5. co semen unseen ece>om a ouecopox .304 .ucoEcwmcsa me co_uuowee co .e>oc55e oc_u.occu_: ace Located cu_s mac_cu cums mucus» at to eueu_cseeoo cu acmaawoc monssoc_ .uuo .mnce_cw a >._sem cu_: uoeucOu acmaecsoum_p ace ace uc_e_o_u_co “oc_uceam a es_uooc* acocucea newssocucOu u>u_.ep_wc_ vo ecomuemsooe «uneduc— .oue .aoc_._ec «cacao on» we egos. uc_xes co oc_cocu_ .auceeep .mpceseu >c0uumuecucoo uc_xes .uc_>. moussoc. .acom>ecob cmocu uc_~_omu_cu uce oc_.so_p_c so eucec e ace .Eecu eue_._ssc o» uc_uasouue .mesec cecucea acmsseu «uneduc— .eueeccu cacao ace eucoso.> we enacts» .cocucea scenes as» ue neuuoc_u uoc eucesow> deuwe>zn noussoc_ .uue .eco_>econ oc_..ocucoo .oc_s:umv.c a oc_e.omu_co .eueeccu e:o_ce> oc_xes eonsdocs .eecoue mum co semen eesnemce> .eeucewceaxe ece>ee ace u._e cu_s eeocu ace .eceacewwo ee em *0 eoce_ce5xe «socu_s ace no.3 eucoucoaeec ceesuob eue_ucecewm_p unseen concoeeo amwmmwcmwem -oo~ .55. .353..k. as. .3k. .35. oo~ .35, -~o. .kkp -mx. .-. an, .m». .oe. .ONF .nk. Fos-mop .~ep .Pa. ao~ -oo~ .55. .¢o~ -oo~ .50— .eo—.w~p mm~.¢e~ manna Acuu><>uas mayors c_»um> Arucco<>u5s manage roucoeco emsn< .eo_mo.oco>m5 oseom desecocu_3 -<5~ e.eum co_ue.om_ .208 a 38:8 -<5_ eseom mmecmsosseu secomuQEm -<5~ oseom 0.:o_u_x a Emmumumcu -<5_ edeum oucesom> a uneven» -<5_ OLouw d-uOh .<5sv oesn< . .eomoosocoxa5 mo sebace>c_ 0e3n< deowoodocu>e5._~_ Asuo> uce Aconceoemv onsets couceewo eesa< .esxom uqmmwumuummmmm x.e.ucooo a use.» 159 abuse. It displays the location of the items included in each scale (item numbers), a description of the content and meaning of each variable or scale and information about how each was created from the items. It also includes information about the internal consistence of each scale (Cronbach's Alpha). Six NsPA variables and scales, which were utilized to test the hypotheses for this study, are described in Table 8. They are: The Conflict Tactics Scale, The New Violence Scale, Dangerousness, Reasons for Inflicting Violence, and, two categorical variables distinguishing respondents who reported none, mild or severe experiences with NsPa (both as victims and offenders). Although it was hoped that the information from the open- ended questions would be useful in exploring qualitative differences in the NsPA experiences of men and women, the responses to those questions were judged to be too inconsistent to provide much valid information. In particular, a significant portion of respondents who had experiences with NsPA did not respond to the questions and there was tremendous variability in the amount of detail given by those who did - ranging from one word answers to several long paragraphs per question. Also, in general, women tended to respond more often and with more information than did men. Because of these problems with the methodology, the open-ended responses were neither coded nor analyzed. 160 Variables concerning injuries and the perceived emotional and relationship consequences of experiences-with NsPA were included in the analyses. The discussion of these is included below in the section concerning variables and scales related to the emotional and physical correlates of experiences with dating aggression. Measure of sexual abuse (SA). A modified version of The Sexual Experiences Survey (Koss & Oros, 1982) was used to measure respondents' experiences with sexual abuse (items 242 through 255 and 259 through 271). The SES, as discussed in the review of the literature, is a widely used measure of experiences with sexual aggression. In this study the 10- item form (Koss, Gidycz & Wisniewski, 1987) was adapted to include two items which were in the original version (Koss & Oros, 1982) - one of which was a positive sexual experience item; the other, an item which asked outright if the respondent had been raped. This latter item has been found to be useful in indicating the proportion of respondents who do not acknowledge their rape experiences as such. Three other items were also added. Two were included in the version of the measure adapted by Lane & Gwartney-Gibbs (1985): one concerned the use of drugs or alcohol to coerce sexual activity, the other was another positive sexual experience item. In addition, another item regarding the use of threats of force to coerce sexual contact other than intercourse was included because it was thought that this may 161 be a relatively prevalent occurrence which was not tapped by other items. In this study, the Sexual Experiences Survey (SES) was given in the same manner as the CTS. That is, both men and women answered all items and both genders responded as both potential victims and aggressors. Koss, Gidycz & Wisniewski (1987) reported that internal consistency coefficients for the SES were .74 for women and .89 for men. In the current study, Cronbach's Alpha for the modified version was .77 for the entire sample (men and women combined; based upon the data concerning experiences with victimization). Koss et al. reported that test-retest reliability at one-week retesting intervals was .93. In terms of validity, as described above in the literature review pertaining to SA, scores of the SES have been found to be associated with a range of attitudinal and behavioral measures in both laboratory and naturalistic studies (Gwartney-Gibbs, Stockard, & Bohner, 1987; Koss, 1985; Koss, Gidycz & Wisniewski, 1987; Koss & Leonard, 1984; Malamuth, 1986). Table 8 describes three scales or variables pertaining to sexual abuse that were used in the analyses: The SES scale itself, and two categorical variables distinguishing groups of respondents with different levels of experience with SA: one distinguishing between victims, and one between offenders. 162 Measure of psychological abuse (PsyA). A 34-item Inventory of Psychological Abuse (IPA), was developed for this study and a concurrent study of battered women in a post-shelter intervention program. This measure assessed respondents' experiences with this form of dating aggression (items 171 through 204 and 207 through 240). The items included in the IPA were based upon the earlier work of NiCarthy (1982), the results of Hoffman's (1984) study of psychological abuse, and the experiences of the measure's authors in working with battered women. The IPA asked respondents to indicate the frequency with which they and their partners engaged in any of the behaviors listed to annoy or hurt the other. The IPA data were examined to explore the possibility of creating meaningful sub-scales. This involved a number of steps which followed recommendations made by Jackson (1970) for rational-empirical scale construction. First of all, the distributions of responses to the IPA items were examined and four items (8195, 196, 198 and 199) were dropped because 95% or more of the responses fell into two adjacent response categories in both the ‘sustained' and ‘inflicted' series of questions. Next, the responses to the remaining items (for the data describing victimization experiences) were subjected to a series of principal components analyses, the results of which were used to guide the creation of rationally meaningful scales which would be most likely to perform well 163 empirically. This resulted in the creation of the five sub- scales which are described in Table 8: (a) Threats and Violence, (b) Criticism and Ridicule, (c) Emotional Callousness, (d) Control and Social Isolation, and, (e) Withdrawal. The internal consistency of each scale was examined and the resulting Cronbach's Alphas, for both the victim and the offender data, appear in Table 8 as well. The extent to which the scales defined empirically distinct phenomenon was assessed by comparing each item's correlation with the other items in its own scale to its correlation with all the other scales. The results of this analysis appear in Table 9. Corrected item-total correlations (i.e., correlations of items with the other items included in its own scale) are marked with asterisks and printed in boldface. A comparison of these correlations with the others in the same row (the same item's correlations with the other scales) reveals that in all cases except one, items correlate more highly with their own scales than they do with any other scale. The one exception is item #172 which correlates with its own scale (Control and Social Isolation) at .528 and with the Threats Scale at .530. Since it seems conceptually more closely related to processes which lead to social isolation (which are tapped by some of the items in the Control and Social Isolation Scale) than it is to items pertaining to threats or violence directed at inanimate objects, it was included in the former scale. 164 sauna an maouw wonuo 0:0 sums uses“ no coauewoomme madame ou poudmaoo .meaeom cso .mcowueaowwoo Heuou-aouq vouoewuoo a onn. pnm. orn. arse. oow. «asoco e ee rescue c30> 0.3ec_ c0 0.50.0.5 .oo— onn. ace. nun. anon. —0~. some.u c0 oaeu_cec .eoec emw_u>.ec cso> «.3ecm 50 e.su_u_¢ .>o- -e. non. nae. e—~>. «me. sane—sen vesse> ueoa cso> «sane. c0 assumvma .0»— wen. can. con. anon. own. 50_.nsn c. 30>.oe_o_u_cu c0 essu>u_¢ .nop >ee. 0~e. amho. mom. oqn. peoc_.eom cacao cso> >0 ecu_. exec c0 «cocoa .eo— own. «on. expo. ~om. mpn. sconce cso> >0 ecu_. axe: to «coco. .nmp nae. own. own. case. oem. «30> eue_._csc 00 >5» .~mp are. m>e. p—e. than. 0mm. neesec 30> ..eu .Pm— enn. one. e>~n. >oe. oo~. saueosooc c0 accuses >c0uumpecuc00 axe: .oop n~v. Pen. amen. 00¢. non. «50> consume >.ouecon_.op to 30> 0» e_4 .mnp .aom. can. ass. «~3. oa~. successaca5 .. co_uooet. to co_s.,uot55. ...>o.55. n.0s5013 .sk. >en. «one. p—n. >Pe. ooe. «ease 00c» c50> 0:05» 30> 30; socucou 0» >5» .>>— nm~. ease. n—~. eon. con. 5>ecos cso> ace5a 30> 30; socucou 00 >c» .o>— can. span. unn. n~n. pee. 5coc\s_c usocu_3 0:0 0m.0u 30> u_nc05 .mnp amen. Foe. >ne. n>e. m~n. «op 00 sauce: 50> uecu 30> cu_s auc_cu op 0» oeswoa .enp e¢~. m-. enhn. n~n. >-. peamcac0_uesoc season cacao coe\e_c usage 50> ..0» .n>p non. clan. enn. —~e. onn. «enmcec0_ue.oc .esxou c0300 uc_uces c0 uc_>eg *0 30> eesoo< .~>p than. eons nun. Fen. o>~. «30> 0» xdeu 0» cusses .—>p czecvcu_3 socucou «50.5eu sewo_u_cu sweets» ecoucou seu_ awn” mused Hmowmoaoconmm mo >wowco>cH on» now mcoawsamuwoo mamom\aoun m OHDMB 165 .mwnaom go .365» 5.” 253 uwsuo may sun: 80.: no coaumfiuommm mmwmmm ou 6959300 .mcowpmamuuoo Hmuoulfimufl umuoohuoo . 3.... an. .5. 3m. .20. :38. 2.8.. .5 F. Eufl Sc 88...: .230 3...: $8 2.... 3m. .3. 03. .E. t 5.3.3 853:, 5:. 2; c332: .BN New. ene. Jonln. 3e. ceno. 30 an 0:20.050» 30.50 ._0 :2 cu c3035 oem. ewe. a. new. .>me. 30 3 Lace 3 on} :0...) 03030 5.500 3 :33ch .SN nmucmz Sn. .3“. 03. :m. E. 2:... :5 on 0.80 no; t 028033... 9: 0c. 8 5:8,: .8~ . . «00:03.5 02. Se. ~e~. can. «nee. 50> r..— 030 9:59:00 0.0.3. ._0 5.2. .30....» £0.5wa £095 (.2 SN .30. 8m. alum. Se. WEE .3 3:5 .3... 8.383 .2: -~. .03. 0;. mm. 5. $2. 8.8 E .22: 8 3.2.: So» .322. .2; ~e~. .05. mlmn. 0:. Sn. 323.: .0 3:5: 50> 022:5 .8— EN. nen. on. «own. «305389.30 Saxon ._0\uco 093.5 .2: 03. me. one. .0? Wm. 38.2335 So 33.5.5 .02 to “50.5 9.0: mon. Fem. one. uni. ume. 30> :02: 300.53 *0 3.80 00.0...» .6 £59.33 .50> 0523.5 .0? Jam: 33:8 030:3 5303.75 309:: acoucou E3. '62. Eva— . 30.0008 0 033. 166 Thus, an examination of the patterns of item-to-scale correlations provides some support for the distinction between the five IPA scales. It should be noted, however, that many items are highly correlated with multiple scales, suggesting high interscale correlations. Table 10 displays the interscale correlations of the IPA subscales. And, the intercorrelations are, in fact, high -- ranging from .357 to .620, with an average of .515. In addition to the total scales and sub-scales of the IPA, Table 8 describes two other PsyA variables used in analyses. These are two categorical variables - similar to those described above related to NsPA and SA -- which distinguish between respondents with differing levels of experience with inflicting or sustaining psychological abuse. Psychological and physical status variables: The Brief Symptom Inventory. The 53-item Brief Symptom Inventory (Derogatis & Spencer, 1982) was included to assess psychological variables (items #7 through 59). The 831 was designed to measure low levels of psychological symptomatology among normal populations, as well as among psychiatric and medical patients. It consists of nine scales and a global index of psychological functioning, which are described in Table 11. The nine sub-scales are: Somatization, Obsessive-Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism. Internal consistency 167 q n H‘ co; save. team. :aNm. :hmn. 003303 8.. ..mav. twee. sen». 300m 0:0 Baeou 80:30:00 co. _ tn 3. an 3. 30039.5 0:883— oo; snow. can Emmoctu 0000—0; 8.. tea 882:. .3805; 3.3000 $0830.30 8&0:th 88.3; mcowumamuuoo manomuoHCH mmsnd Hmowmoaoco>mm mo >uoucm>cH on manna 1658 A.u.uCOuV .33. on ou vougoaog oca.< a.guancogu .Ameou_ mv ogouo so». ooaLo>< .ns. on on unugoaog asn.< u.;uoncogu .Amsoum my agouu no», anago>< .po. on o» tonnage; acad< n.5uancoLu .Ameou_ ov ogouu sou. ooogo>< .mo. on a» tenuous; aca.< u.:uoncoLu .Aueau. 0v .500. no». ooogo>< .es. on ou nougoaog acnd< ¢.;uoncoLu .Ameou_ e» 0500» so». ooago>< .no. on o» nougoaog ocnd< a.;uanco;u .nmeoum ow «goon no». ouugo>< .a~mop .Louconm a u_uouoLoov on. v9 cu nougoaog osa.< u.:uonco;u .Amsou_ NV «Lou» sou_ mongo>< .~s>.oeoguxo:v q o» Ag..o an woes» o 50;» coco“ .uonooog ago: «sou_ .ma ..< .~mn c_ neau_ mm ..o gov «Loon so». ouogo>< .ogo_>czon oucou_o>o ace owno;aagouo we meouasxa no zona usuaasxa o_noca o» non-dog auoguu_u «woodwoa .Looca saw: nouo_uoauo ago_>agon nee aucvdoow .uucaaozu o» tau-.05 uaoguu_v uuuo.*o¢ .uuocuaoduaog uca axoouuo o_coa .mwo:m:0>goc mo uncmdoow oc_u:.uc_ >u0_xcn mo meauasxa on vouo.og aaoguo_u auuo.+o¢ .augoaog» 3333 as 83:30.80... .a_..ofiu>u we auc_.oow no scan co_uaouauu we «sauaexo on wound»; nauguu.u auuo.*uu .agosuo :»_3 couwgoaaou coa_gaaeou c. >u_gomgo»c_ uco >uoauouoc. *o aoc_.oow o» vouo.og «moguu_u «woodwoa .gougoa_u o>_ud:nsou .o>_uao¢no gum: nou-_oommo >.do:aa aeouasxa so.» uc_u_ga unusuavu a»o0.»0¢ .co_uuc:.u>u >._uon .0 aco_uaougoa sou» uc_a_go anoguamu «woodwom .gucoe yawn oc_gan .o>o. «magnumu o.ucoucoamog mo co_uoo_uc_ ddogo>o mmflmmwumwum 3on 53 mm.o¢.~n.¢n.¢_ >»o_xc< u_no;¢ Ana: .mnv ~m.~¢.o¢.ov.~. >u.._uaoz mm axc< .mm. .pm.4¢.n~.np.~ >»o_xc< .aoo .muv on..¢.¢~--.m. co..oogaoo Auc— .mn. >uw>_u_ucom o¢.o~.-.o~ .ncougoagouc_ Au-o .mm. ~c.om .mn.~n.F~.F— o>_a.:asau.o>_aao¢no me.on nausea .mnv on.mm.o~.n..o co_»a~_uosom om-~ A_mov nouc_ >u_go>om .ogucoo A.~wow .goocoam a a_uaooLoov .n_mmv >Louco>c_ Ecuasxm *umgm mammq mammmummummmmm modnamgu> >L:_c_ new mauoam .ocomuoEw new dou_uo.05u>mm $0 aco_an_gomoo pp 0.30» .Aco>_u aco_u_c_wouv s>ga_:_ ogo>oas ou steam-umau .vougoaog swan go >ga_cm 0:: 30;» oc_uou_uc_ c ou F 555» uc_oco¢ >L:_c_ ungoz ozu yo >uvgu>ow 0:» yo uc_uo¢ .Aco>_u aco_u_c_*ouv s>g:_c_ ogo>oua ou :uoc.auuau .vougoaog c_oa Lo >g:_c_ 0:: 305* uc_uoo_uc_ q o» — eagw nausea >g:_c_ wage: on» $0 >u.go>ou u:» we uc_ua¢ .odouu uc.>:z c. odaou c_ vanadocm .goCuLoa sag» noc_ou¢au aa Lam .co_uaogooa *0 «550* “nos. n o.aoa < yo co_uogan ac. >u_go>on yo nogaaoox .nono.uasn.< o.=uonco;uv .co_uuogouc. doaxoa *o usgo* gosuo co nowon nogouo o>og ongaoogouc_ :. «aouco 0, ac: at on: ouuomnaa .aagh .aoado> uc_oams an voc_voo ago» ,0 yo noucoaaoa s.oc_>*a_uoo >go>s cu suc_>mu_uoa noes 505* .oag30ugoucp .oaxon cu uc_uoaz new 1; oc_ocag ugozuca ouoomuc_ ¢.— “nuance no: ouuo wag» nc_ua_x sag» ao_u_>_uuo doaxoa yo once; a on uouooog «souu .ansou. nv aagouu so». ooogo>< o uucwfi acoucoamog uc_>~o_uoa so; ogaaou: .noo.u .aamzmcovua.og c_oucmoe use nodo>vu ocn.< a.soancoLu .ameou_ «v ogoua aw». oocgo>< u:- agocuo uwaga ou xumdpno mogamoo: .uuo .ogoguo ou omodo acm_ou* Lo>oc .uo;._c:a on n.2ogu so» “any .augaaogu .F5. on on vougoaog gao> denucOu coo accuse» yoga acmxc_zu ma ozad< a.zuonc0gu .ameou_ my «Loon so». ouogo>< cusa «Ecuasxa on vouo.og amoguawu muuu.»o¢ .ouo .agonuo >9 uaono nexdau so vacuum: og- :o> was» .505» and 36> »_ 30> yo ooauco>uo ox.» .dmz >ocu was» .ns. on cu nougoaog .uouaagu on uoccou «gusuo uogu uc_.oov no 0m~ ”mp unp-n~F oo~-mo~ nm~.~m~ ocw-om— no-00 om.om.oq.o~.o ognd< u.:oonc05u .auEUu_ my choc. ecu. ouago>< gun» «ecuns>a cu vow-.05 nauguumu «woo.wo¢ sm.em.on.op.o— comumwuwmmm mamww u_go>om u_go>om no_L:_c_ <>au *0 yoga:— docopuOEm ma ~mmv Emwumuozu>ua Agog .mmv copycat. amocogoa «mummummummmmm ..u.»cou. .— o_nah 170 coefficients for the nine scales have been found to range from .71 to .85, with test-retest reliability coefficients ranging from .68 to .90. The 881 was chosen as a measure of psychological functioning for this study because the range of symptoms that it assesses are appropriate for exploring the psychological correlates of both the victimization experience and the phenomenon of perpetrating aggression against an intimate partner. Psychological and physical status variables: Other measures. Other measures of hypothesized psychological and physical correlates of dating aggression which are summarized in Table 11 include the Relationship Satisfaction and Sexual Satisfaction Scales, both of which were developed by Ross, Dinero, Seibel and Cox (1988) to measure social adjustment in rape victims. These scales and items were included because of the evidence, reviewed earlier, that victims of sexual assault experience prolonged difficulties with sexual and social adjustment. In addition to these measures, six other variables in this category were assessed: Three scales measuring the self- reported intensity and duration of the emotional impact of NsPa, SA and PsyA; the number of injuries experienced as a result of NsPa; the severity of those injuries; and, the severity of injuries sustained from experiences of SA. All of these variables are described in Table 11. 171 Resource variables. The questionnaire included several items that assessed respondents' and their dating partners' resources (items *6, 85 through 93 and 95 through 105). The items measured ten different economic, social and personal resources for both the respondents and their partners: GPA; age; educational level; parental income; personal income; amount of spending money; and ratings of overall levels of health, confidence, social activity and attractiveness. The intention was to create scale scores describing the respondents' level of resources, and also a discrepancy score indicating the difference between the respondent's and dating partner's resources. Since the resource items had different scales of measurement, the first step in scale construction was to create standardized scores for each item measuring a resource for the respondent and for each of the ten difference scores between the respondent's and partner's resources. Next, the inter-item correlations were computed within the two groups of variables (respondents' own resources and resource differences). In general, the items displayed fairly low inter-item correlations, suggesting that the combination of items into scales might not be warranted. For the ten items measuring the respondent's resources, the inter-item correlations ranged from -.158 to .384 with an average of .116 (average of the absolute values). The range of inter-item correlations for the difference scores were somewhat more promising, ranging from —.154 to .615, though the average of the absolute values was only .139. The 172 generally low correlations were not surprising. It would not be expected, for example, that age or educational level would be correlated with parental income or with ratings of health or attractiveness. In the interest of variable reduction (instead of having 20 distinct resource variables in the multivariate analyses), it was decided that scale scores should be constructed in spite of these low levels of association. Additionally, given the clear conceptual distinctions between clusters of resources (economic vs. personal and social for instance), it was felt that the possibility of meaningful sub-scales should also be explored. So, the respondents' ten resource scores and the ten scores measuring resource differences between the partners were each submitted to a series of factor analyses which were used to guide the creation of sub—scales. The final solution which performed most adequately, both conceptually and empirically, entailed three scales for the respondents: (a) Personal and Social Resources; (b) Economic Resources, and (c) Demographic Resources (which consisted of age and educational level); and two for the resource differences between the partners: (a) Differences in Personal and Social Resources; and, (b) Differences in Economic Resources. With the exception of the Demographic Resources scale, these are described in Table 12. The Demographic Resources scale was found to correlate spuriously with the abuse variables as an result of the way respondents were directed to choose relationships to describe 1373 av.ucouV .uoou coaueoa was» nocooa o>_uouoc “auocsoaoe “moo.uaga.< m.coonc0cuV .— »_o mg no >o3 «Sam oz» cm nousasou AFaa...;a.< ..gu.ncoruV .uocouu-~ onus» yo ounce). on» a. crane odoom .uoeooo oucocomv_u «easy com nousaeou noeoua.~ .uucocoasou o>_e ogu we soon com vousneoo nocouu oucocowmao Rama...ga.< ..;o.ncoru. .aucocoasoo oats» ozu yo gone co» nocooo.~ yo ouaeo>< .~nm.u.ga_< ..=oancoruv ..ucocoaeao o>mw on» we coco 50* nocouu.~ as» we oooeo>< acumo . u ago: no: «concoamoa uuaupuc_ mocoum o>pu_moa .do>o. doco_uooano new one .>ocoe oc_ucoaa .osooc_ w.coCueoa uco u.ucouconao¢ we «nevus; mousduc_ .aeocueoa ogu coozuob mouesoaoe 0.;rouosou\o_eocooo yo oucadon ozu we ocaaoox .odoum .acoa no aucocoasao «sum .0505 an; Lazaro; «as» .uoeooo o>_uoooc “Locate; cog» are: no; acouconuou ouoo_uc_ arenas: o>_»_ooa .ucoeueoa uc_uov as» as» cooauan «outsouec .n.uon a .ocoucoa we 00:0.99 ozu so ocnuoo: .xocos uc_ucoau u:- oeoufi .9823. .0285 3:28 .0 auc_uoc nona.oc_ .a_zuco_uudoc ecu oc.c:u noucsouoc o_eocooo u.ucouconuoe «o assuage ou.oonsou .<.a.o use unoco>_uoacuuo .>u_>_uoo .-_uoa .oucon.»cou .a... .g»..og ..u_.>;a .0 uuc_uoc «uneduc— .n_gnco_uadoe oc_uau as» uc_c:u «couscous .apuoo a decanted we .o>o. u.ucoucoaaoe we «caucus ou_aoa§ou amauaqummmu ow.ow mo..~o. .oo.no-ao.mo um.ow.o no-oo.mn manna a~ e_o nae moucsomoa 3.50m a .ocomcoa c_ moocoeo»»_o AP e_o ace noueaouou .o_uom a .acoocoa c. neocoeom»_o A.coouv noocsouom o.eocoom a.¢L0Av «outsouoa .a_uom a doconeoa mmmumMmmm .. o o e: no: uo.bo_cu> amsnco.uo.o¢ ten oucaoaox we aco.ua.cuuoo NF odnah 1I74 .An. 000E0>.0>0. .0 .0>0. 00.: 000 ..~. 00000>.0>0. .0 .0>0. 00000000 0... 00000>.0>0. .05000 000.00000 0.00.00> a00000>.0>0. .0 .0>0.: 0:0 00 .0 .0>0. 30. "00.0 .0300 .0 003000 .3050. 0;. .0 no.0:n_ru._u 000.00.: 000.0 ..c_oa ..o 030 «are. 00:..00 gu.g: 00.0..r.> .0:...00 00:0 :00: 00 gauco >.00 00:0 :00: 000. 0.000 0000 0.0 0. 0003 00000>0 00 0. 00000000 00000 0.000000 00.000 030 000 00.0 0030 300 .0 00.000 000 ..=000000 0.00 00 00.0000 00: 00 00 .: 00 30000 000000 0.00 00000 >.00 .Amom.u000.< .: 000. 00.000. 0.000 00.00 00>00 0.00000000. .000000.~ 030 00000 .0 00000>0 0. 00000 0.000 0. .00000.0000 0.0000.00.00 .0 0:. .0000. 0:0 0:0 .0 0000 .0. 00030000 0003 000000.N 00.000 00.03.00. 0.30000 00.000000 .0. 000. 000 >000 000000.00 000 03. 0030 :0: .0 000 0000 to. 0. 0.00:0.0..or 0:0 00.50 .000 003 .0 0000 300 .0 .00.000.0.000 ..000>0 .0 00.000 00 00003 .0.00 .Aomo.u000.< -00.00.00 00.000 000 00.: 00.000. 0.:0000000. .0000. 00.00 000 00 00000 00000 000. 00000>< -0.000 0.0000000000 .0 003000: .00000>.0000000 a >0.>.000 .0.000 .00000..000 .00.00: .>0000 00.00000 .00000. ..0>0. .000.000300 .<00 .000 .9800. .0000000 "00003000.. 0. .0 .0>0. 0.0000000 a 0.0000000000 .0 000.000 00 00000 ...+. 0000 000 0000000000 000 00.03 0. 000 000 .0. Am.m.u000.< 0.00000000. .00000 000000...0 .0000 0.00 000030000 00000.00 >.0>.00.00 0>00 00 00.030.000.0 000 0. 0.0000 000 .0 000.00 00000.: 000 0500 00000000 030 000 00.0: 0. 000 0.00.0 .00030. 0003000 00.00.300.00.0 >3 000..00 003000 .A.-. 0000000000. 0000 000030000 .000000.~ 000 00000 00.0000>0 >3 0000000 00000 0000.0...0 .0000 0000 0>00 0.000000 00.03 .000. .000000 000000...0 00000 00. 00030000 000000.~ 0. 000 .00.0 .0300 .0 .003000 .000030000 0. 000 .0 0000 00. 00030000 000000 000000...0 00000 000..00 00.03 0.00.00> co 0 .0000 . .uou 00.00000000 0°.-oo. 00—.03. m...—.. no—.ua .oo.mo.no.mo.o manna 0.00.00> 03000 00000>.0>0. .0 .0>0. 00000>.0>0. .0 .0>0. >0..030 0.0000.00.0¢ 00030.0..00000. 003000 0000.00 00030000 mqummuuummmum A.D.0couv Np 0.00» 175 on the questionnaires. Through a series of screening questions, they were directed to choose the most abusive relationship that they experienced during college. If they had no experiences with abusive relationships, they were to describe their current relationships. Thus, respondents describing abusive relationships were more often talking about relationships in the past, when they were younger and had attained lower educational levels. Because of this spurious relationship between this scale and the abuse variables, the scale was dropped from inclusion in any analyses. (Though measures of the differences between the partners' ages and educational levels were retained and are included in the Differences in Economic Resources Scale.) Table 12 includes information about the internal consistency of the remaining resource scales. It can be seen that the values of the Cronbach's Alphas are fairly low, ranging from .481 to .665, which reflects the low inter-item correlations just discussed. Tables 13 and 14 display the item/scale correlations between the resource items and the scale scores. Table 13 contains information about the variables assessing the respondents' resources, while Table 14 deals with those describing the resource differences between the dating partners. Both tables reveal that, with the exception of the respondents' GPA's, all items are more highly correlated with the combination of the other items in their own scale than with the other resource scale. The item assessing the respondents' GPA's was correlated similarly 176 Table 13 Item/Scale Correlations for the Resource Items (Standardized Scores) Ite- Contenm Parental Income Personal Income GPA Health Confidence Social Activi Attractiveness ' .247 n -.O31 .019 EdUcational Level“' -.030 0.052 These are corrected item-total correlations to assess the item's relationships to the other items on the same scale. These items were not included in either of these two scales. Table 14 Item/Scale Correlations for the Resource Difference Variables (Standardized Scores) m l Ite- Content Differences in Differences in Personal/Social Resources Economic Resources Differences In: Parental Income“r .081 .128 Personal Income .060 .643. Spending Money .085 .495' Age .007 .433' “ Educational Level .076 .002' I n on .193’ .013 II Health .268. .155 n confidence .405' .158 H Social Activity .237' c.059 H Attractiveness .185. -.046 “ . These are corrected itemrtotal correlations to assess the item's relationships to the other items ,. on the same scale. This item was not included in either of the resource difference scales. 177 (and at a low level) with both scales but was included in the Personal and Social Resources Scale because it was felt to be a better conceptual fit there. Regarding the two economic scales ~~ the one describing the respondent's economic resources and the other describing the economic differences between the dating partners -- the factor analyses indicated that these should consist of different combinations of variables. In particular, the differences between the partners' ages and educational levels were found to correlate fairly well with the differences in their economic resources as measured by the other, more obviously ‘economic', items. The same was not true in regard to the items describing the respondents' individual resources. This difference may reflect differences between the respondents and their dating partners, since the respondents were a more homogeneous group in terms of age and educational status than were the dating partners who they described. Whatever the reason, in this sample, being older and having attained a higher educational level was not associated with more economic resources for the respondents, but among dating partners greater discrepancies in age and educational level were associated with larger discrepancies in economic resources. Another difference between these two scales is that, among respondents, their parents' level of income was associated with the other economic items describing their resources (as might be expected for college students), but this 178 relationship was not strong for the dating partners, nor were the difference scores on this variable highly correlated with the economic differences between the partners. Thus, the differences in the contents of the two economic resource scales reflect these observed differences in the patterns of association between the two sets of resource variables. In addition to these resource scales, Table 12 includes the variable, Resource Balance Groups, which categorizes the sample into three groups based upon the balance of resources among the dating partners. Relationship variables. Table 12 also describes three relationship variables: Relationship Quality, Level of Involvement, and a categorical variable which differentiates between three levels of involvement. The specific item contents, and scale information are included in the table. Attitude variables. Table 15 describes six scales which assessed the respondents' or their dating partners' attitudes or the discrepancies between them. All of these scales were derived from the attitude questionnaire developed by Burt (1980) to assess cultural attitudes that support rape. The scales which were included in this study were, the Sex Role Stereotyping Scale (which was used to assess respondents' sex role attitudes, as well as their perceptions of their dating partners' attitudes, and two measures of discrepancy between the two), The Adversarial Sexual Beliefs Scale and The 1379 craggy ...o_.on 0.0.0 .0 0000000000 0000000 00000.00. 00000 0000.: .0..0000 000 .0.000 .00>00 00 00003 000 000 0003000 ...00.u000.< 0.00000000. .0000. 00.0 000000 00000 000. 0.0000.00.00 00>.00000 0000000000 .000. 0030. 00000>0 0. 00000 0.000 .00.0000 0000>00 00 0000000 0000. ..< 00.03 00 0000x0 000 .0 003000: Nnn-0~n 0.0..00 .0.00000>0< .0000.0.> .0 0000000000 0000000 00000.00. 00000 0000.: .00.0000.00.00 0000.00. 0.000. ..0mo.u000.< .0.00 .0 000000000 000.0000000 0.00000000. .0000. 0 000000 00000 000. 00000>0 00 0. 00 00 0000.0.) 0000000 0000000000 .000. .0030. 00000 0.000 .00.0000 0000>00 00 0000000 .nn 0 «mna 0000. 00.03 00 000000 000 .0 003000: omn-onn .wnm 0000.0.> .0 0000000000 .000000...0 000 .0 00.0000.0 000 .0 00.000...0000 03000.3 00030.000 0.00 .000 0.0000000 000 .00000000000 0. ..0 000000...0 0.00000 .0 03.0> 003.0000 00. >00000000.0 000 .0 0m.” .0 003000: .m.-¢.. .-n-m.m m0 000000...0 0.00x00 .0000000000 000 0000 .000.0.0000 000. 0. 0000000 0000 .0.000 0.00000 0.0000000 0000.00. 000000 0>.0000: .00030.000 0.000 0. .000.0.0000 000 000 0.000 0.00x00 000 00 «um& 0. 0000000 00000.00. 00000 0>.0.000 0 .0+ 00 0- 000. 00000 .0000000 0.0\000 .0 00000 000 00000 0.0.0000 000 00000 000000...0 00.0.3000 .00000 0.000 00030.000 0.00000 0.0000000000 000 0.00000 0.0000000 000. 0000000030 0. 00000 0.000 0.00000 0003000 >00000000.0 000 .0 003000: .~..0.. .-n-n.n .0 000000...0 0.00x00 .0.000 0.00000 0. 0000. 0000 00 0003000 0.0000000 .0 00000.000 ..00..0000.< 0.00000000. 0.0000000000 00 00000 00030.000 .0.000 0.00x00 000 00 00.00030000 000 00.0000 .00.0000. 0.00x00 0.0000000 .0 003000: .~..0.. 0.00x00 0.0000000 ..0~..0000.< 0.00000000. .00030.000 .000.0.0000 0000 .0000. 000.0 000 000000 00000 000. 00000>0 00 0. 00000 0.000 0000.00. 000000 0000.: ..000.0.0000 .00030.000 .000.0.0000 0000 0000.00. 000000 00.0 0000 00 000 00030.000 0.00000 0.0000000000 0.n0 00. 0000x0 .00.0000 0000>00 00 0000000 0003 0000. 00.03 00 0000x0 .0 003000: -n-m.n .000..0030. 0.00x00 "000000000 .. mm.0a.0000m .000.m0ummamu 00.0m.0000m 0000. «.00wV003000: 00.00.00> 00300.00.x0: 00.0.000 000 0030.00< .0 000.00.00000 m. 0.00. 1130 Anzac“: .a0000000000 000 000. 00000 00.0 000. 0.000 0. 00 3000 0003 .30000 000000 000x00 300 .0.0. 00.0.000 0000000000. .0000000 0.0000000 ..0.000 .000. 000 00 00 00000.0000 3..0303 00. ~00. .0.000 .0300< 00. 00..u000.< 0.00000000. .000.0.000 00300 0000. .03000 .03000 00.000 00.0000000000 00 0000 000 00030000 00000 00 00.0000 0000 0300 00.000 .000 000 0000 00300 0000 00. 00.000 00000.00. 00000 03.0.000 .3..000. .00.. 0.303 >000 000.00030 00003000 .0 0 0030 300 000 .000.0.000 0000000000. 0030 0.0000.00.00 00.000 000 "0.000 0000.0.000 0000000000.: 000 0. 00.00 000. .000. 0. 000 ..03000 0000000000 00300 0030 \.03000 00300 000 .0 0000 00. 0000000...0 000 .0 0000030 00. 300 0003000 >00000000.0 .0 003000: .000.0.000 0000 00000 0030 0.0000.00.00 .000. 00 0. 0300 00 00.. 0.303 0000000000 00003..0. 0030 300 00000.00. .00003..0. 03.00.00 0000 000 0000000000 0000.00. 000000 0000.: .000.0.000 0000000000. ..0000000 00000000000 000 000 0000 0000000 00 0.00000 00 00000.0000 >..0303 000.0.000 0030 000 ..0.000 .000. 0.0000000000 000 000 0000 0000000 00. 00.. ..0300< 00. 0.0.»000.< 0.00000000. .00.000 .000. 00 0.00000 00 00000.0000 >..0303 000 .0300< 0000 00. 0000. 00300 000000 00000 000. 000003< 00.0.000 0030 .0000000 00.000 .=.00.000 0003.0 0000.0 .0 an 000 2000.000 0003.0 0000.0 0.000 00 03.00.00 000 0000000000 0000000 3:: n. 0000 00 0000000 0.0.n.~ 0000.. 00003..0. .0 003000 00000.00. . amwmmwummumlmmamwmmmmmamm mmqwmwummwm 0300< .0 ..< 0.n.~.n.o.m .mon.0on.00~.so~ .00.... 3.0.0000 000.08.000.30... 03000 .0 ..< n.03000 0000. 00 00 0000000 0000. ~0~.00~.00~ .00~.~0~.0-.00~ ~0~.oo~.00~ .00~.~0~.00~.ms~ 00000 000000...0 .83 .0300< "000.0.000 0.0000000 0030 00300 000000 000000...0 .000. .0300< "000.0.000 0000000000. 0030 00300 0 . oomufluamwv! 0.0.008. m. 0.000 1E31 63.8.93 .35.. s.uco.a_oon go>o o>an udaoca «sexes suaozu .do oxafl u.:osu «sax»; acoao:ua an o» saco_a_uon .d. soaps-a 5.0:» oucoadwcp no 50309 oxoa u.:o;m _ gouachs .P can» unvocag «Loon so». odoc_m can: so; we uc_u-g a.ucuuconao¢ .goxon o>_u-.og vog_aou can: non-u_uc. ogoou 5050.: .oco_u.uon *o aon>u .da nacho. .a_guco_ua.og adaon_s :- c. Agocugon oc.uon - ow o>_uo.ogu ¢>ag on ox_. v.30: A.«nn.uagn.< a.goon:oguv uconcoaaog oz» ouco:.*c_ we .o>o. .msuu_ p~ ..n «wagon ogooa an». ouago>< uoL_aou ..ogo>o on» we ogauooe < .goaon o>_uo.og choc monou_ncm ogouu Loco_= .ogauous 0;» >3 nogo>ou aco_o_oon on» .d. Lo>o Locugoa oc_uou 5.0:“ o» o>_ua.og ..e¢m.uoza.< u.guoncoguv no: acoucoamog ocu cocoa.*c_ .mEou_ Fm ..a macho. ogouu so». ouogo>< *0 .o>o. ..ogo>o an» we chances < ~.co.u_uon doaxoa .LoguooOu oa_u ucoaa ou so: .goguo coco con cu cos: .u.ov ~.o.-uo anon_ aco.u_uou n_smco_un.og .doauas 50* o—o. no.-uu .aauo< no» ops.uazn.< a.gooncoguv on 0» voguu_¢cou >..oama aco_a_uon .o>oaa oo.ouu oc_ucoaaoggou on wean «cu nouaasou go>oo use». .o>ono no oc_coue «sum amauuwuumum.4uamwmuummamm ammmmwummmm Now c—n-eo~ no~-nh~ o>on< mo dd< can.mon .~on-oon.mo~.¢o~ mo~ .«o~ . S~é~£~£~ gua— o>_uox oucac.soo «.50cugoa oco_o_ooo dd< 505° gozou *0 acaos< .oot. Lo vog_aoo douoh A..uo<..ouo»v aco.a*uoa ..< gu>o some; .oauu< douo» ogoum oncogo»*_o .ouu_ .oauu< uuco_a_uoo nmzmco_uo.oa go>o Luzon o.owwuwg1mmmu au.ucouv mp cinch 182 Acceptance of Interpersonal Violence Scale. Specific information about the items included in each scale, as well as information concerning scale construction procedures and the internal consistence of the scales are presented in the table. These scales have been widely used in previous studies of dating aggression - especially studies of sexual aggression — and have been found to be related to a number of measures of involvement in SA (Check & Malamuth, 1983; Koss et al., 1987; Malamuth, 1984). Decision-makinq/Dower variables. Table 15 also describes twelve scales derived from a decision-making measure developed for this study. This measure contained 42 items designed to assess how decisions were made in the respondent's dating relationship, and how the same decisions would be made in an "ideal" relationship. Twenty-one different decisions that dating couples might face were presented and respondents were asked, (a) who actually influenced each of the decisions in their relationship, and, (b) who would influence that decision in an "ideal" dating relationship. Answers were given on a S-point scale ranging from, "I almost always decide", through "My partner and I usually have equal influence" to "My partner almost always decides". Seven items were designed to assess decisions which might commonly be considered to be the respondent's personal, independent decisions (item numbers are listed in Table 15). 183 These items included such things as "What clothing and hair styles are best for you, in general", "Who you spend your time with when you are not with your partner”, "How you spend your money", etc. Seven corresponding items assessed decisions which are the partner's personal, independent decisions. The remaining seven items were designed to tap decisions that are mutual, relationship decisions, such as, "When to see each other", "Whether or not to engage in sexual activities", and "How much time to spend together". The respondents ratings of these items confirmed exactly the classification of the 21 items into these three categories. For each of the seven items designed to assess the respondent's independent decisions, over 50% of the respondents said that they always made that decision. For those items designed to assess the dating partner's independent decisions, the majority said that their partner always made the decision; and for the "relationship" decisions, the majority reported that the two of them shared the influence equally. The responses to this measure were used to develop: (a) scales describing the actual dominance patterns that exist in relationships in each of the three categories (independent decisions, partner's decisions, and relationship decisions) and overall; (b) scales that describe the respondent's dominance ideals in each category and overall, and; (c) scales describing the discrepancy between the two in each of the three categories and overall. Internal consistency 184 information concerning each of these scales is included in Table 15. One additional variable assessing power and influence issues in the dating relationship, which appears in Table 15 is called, Partner's Dominance Motive. It is the respondent's rating of their partner's expectations of dominance on a single, five-point item. Additional variables. The last section of the questionnaire assessed respondents' perceptions, utilization and knowledge of the counseling services offered on UC Berkeley's campus and was of no relevance to the content of this study. The results of the data analyses will be presented in the next chapter. TOWARD AN UNDERSTANDING OF ABUSE IN THE DATING RELATIONSHIPS OF COLLEGE STUDENTS: THE RELEVANCE OF THE FENINIST/SOCIAL CONTROL MODEL Volume II BY Julia Anne Parisian A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1990 185 Results As discussed previously, the research questions addressed in this study pertain to the following three main areas: (a) achieving a comprehensive view of the phenomenon of dating aggression, (b) assessing the relevance of the feminist/social control model through exploring the impact of a variety of attitudes, resource distributions, and patterns of dominance upon the prevalence of dating aggression, and, (c) exploring gender differences in the experiences of dating aggression. The hypotheses and research questions relevant to each of these areas were discussed at length in an earlier chapter. In the discussion below, the results pertinent to each hypothesis or research question will be presented. As a starting point, to provide reference information for the results and discussion to follow, Tables 16 through 19 provide descriptive statistics for each variable used in the analyses, categorized as they were in the previous discussion of the measures. Table 16 presents information about all of the dating abuse variables; Table 17 pertains to the psychological status and injury variables; Table 18, the resource and relationship description variables; and, Table 19 refers to the attitude and decision-making variables. In each table, descriptive information is given for the entire combined sample. In addition, the women's and men's means 1536 Av. 080v ~Fo. Fm.~ u u o~. oo. o«. mF. e . 9 NF - o poqud+cF-mom o~o. F~.~ u u «w. mm. 0F.F me. o - o NF - o nocFoumsm-uom m... m«.F «s. m¢.F n . F n . F gate rucco (an: <¢.F om.F ms. o¢.F m . F n . F «are e.uo_> )z «o.wF F~.FF «n.0c mm.~F one . o . o nocFoumsm.uc.>:z om.F Fo.F F~.¢ wo.F Mm . o - o noquch_.mFu oF.n ~n.F s..o oo.~ mm - o p o noccau.:m-mcu o.mox o.cmeoa outta .qauo< o.aFa¢o¢ can: odnoFca> "moococoFFFo xom scoo_c_cacm c. moanmwuo> omsnd mcfiuoa you moaumflumum o>aumwuouoo ma wanna 1237 mo.F Fo.~ Fo. oo.F n . F n . F sonata coucoFFo <>ma co.F ~o.~ ~o. oo.F m - F n - F manage emuu_> <>ma oo.F oo.~ FF. Fo.~ oo.e . F q . F nouu_.Fc_ - .ozotocLF: oo.~ mo.F on. oo.F oo.o - F c . F pocFoumam - doxococuma ~n.F en.F Fe. nn.F um.m . F Q . F vauuF.FcF.Cu.omF a deucu q¢.F Fq.F mm. ~¢.F Fs.n - F o . F voCFQumsw-c».omF a .cucu mm.F oq.F Fm. Fm.F o¢.n - F q - F nouu_.Fc_.mmocm.oo ..oe~ ooo. ~o.n u u oo.F oo.F ~s. oo.F q . F nocFaumam-muoCu.oo ..osm Fn.F m~.F «n. F~.F n~.~ . F e . F nouuFchF.oou¢ a mucous» ~e.F m«.F mm. ¢¢.F -.n . F c - F uocFoumsm-suo¢ a muaotzF «F.F «F.F on. «F.F no.n . F e . F oequchF-oc.> a ouoocsF oF.F oF.F Fe. oF.F oo.¢ . F e . F nocFoumam-oc.> a nuance» QM.F nn.F ~n. Fn.F mm.~ . F c . F nauudecF doqu s¢.F Fm.F no. om.F om.n - F e . F nocFouasm .oqu Faa ooo. oo.F - .cm Fgu oF.F Fo.F Fm. FF.F N . F n . F «macro coucoFFo (u «8. 2.6 u .3 .5 m: on; 8. R; n - F n - F 2320 .38; «m oocouFFFchm «canon .39 to x x .n.- x omens cocoa «so: o.noFto> oan as” u Fe .3; Boos 3&8: 8.8 332 3888 "noococoFFFo xom «cooFchoFm FF F.u.uc00v 0F 0.00F 188 U. uCOU Foo. mn.m o«.~ Fo.~ oo.F Fn.m oo.m - F m - F Homoeo ..OEu <>ma Foo. ov.m mn.F Fo.m mF.F wu.~ om.o - F m - F Homes. o.05w ma . omm mmo. mum. on. NF. ow.m - o c - o emu - .mm mow. «mm. oq. om. om.~ . o o . o noca - _mm oon. Fun. No. mu. oo.m - o q - o mo: - ~mm ooo. mq.¢ one. co.F oo. ~o. mm.m - o e . o xc< . ”mm woo. Fwo. mu. Fo. om.m - o e - o Qoo . ~mm FNo. ~m.~ «on. oo.F Fm. No. oo.c - o c - o uc_ - me oo.F oF.F mu. «F.F uo.n - o o . o u . o . ~mm Foo. o~.m Fom. mmc. cc. Fm. 0N.n - o o - o umeom . me mmo. oo.~ nmo. cos. we. on. om.~ - o c - o Fmo - me >c0uco>c~ Ecuas>m Foch mucouFF_cmFm .Um m .m .o.» .m mmcom cocoa oEmz woancm> on to.uoum m.cox m.coeoz ncoco oaauu< monmmmoa u Fo o:.m> "moococoFFFo xom scou_c_cm_m c_ moono_cm> >L3_c_ can msumum .mco_u05m\.oo_mo.ozo>ma coF muFHmFHmum o>_uaFcomoo NF monoF 1239 no.P mF.F we. QM.F oo.¢ . F c - F mumcamcF *0 xumco>om 3. NS. 8.. fin. 2 - m - o 3:25 oocQUFFFcoFm «Um x x .o»w x omcoa oucam osmz odocho> _;o to.uaum m.mox m.moeoa ommco .mauu< monmmmoa u we mada> "moucocoFFFo xom ucmuFFchFm o. financFLcooo moonoFLo> >ca_c_ uco msuoum docoFLOmeoouFmo_ozo>ma cow moFumFLQLm o>Fqucomoo NF odomF 1190 Fo.~ ~o.~ ~o. ~o.~ n o“ F m o“ F oars-»coeo>.o>c_ oFo.. ~oo.. so. Fo.. so.F on oo.n. groan." ace>o>c~ so>os oo.n ~o.n so. so.n m a» F m on F ..-:a aFga..o¢ "aFgucoFuoooa coo. so.Fn awa.Fgu oFo. ”FF.. ~o. no. F+ o» F. F o» F- .mro.ouc.ao..o¢ coo. o~.o- .F Fan. oF~.- on. co. ~m.~ o» mq.n. arena-" .coum . FFFo ooo. Ano.. on. so. ~F.~ o» ~m.F. orouo-u a a a . FFFa one. oFo.. nu. co. ~e.~ o» o¢.F. mecca." oFuocoum oFo.. oFo. co. co. ~n.F cu «o.F- orooa-~ ..uom a ...roa "39:62:. .aFm .w» .;u x x u.« x one.“ use.“ «an: «snoFro> co .u-um.F gnaw: a.caaw: toque Fosuo< .uaoo Fa «a..> n 30:23 F F o xom .oFm F_ .mwanm«um> awnmcowumHom can moousomom you mofluuwumum o>wumanom0a ma OHQMB 1591 U. uCOU oo.~ mo.m om. oo.~ o~.m - mo.F m - F .mqu - omsuo< ono. 05F. om. «F. oo.F - oo.F- o+ . q- anmcoFumooa - .FoFo NF.n oo.m om. oo.m oo.m - mq.m m - F onmCOFHmoom - smooF ooo. vo.~- qo.m mm.~ «v. oo.~ o~.v - oF.F m - F oFngOFLmoom - .mouu< woo. on.~- oow. moo. em. NF. «F.m - oo.~- o+ - v- m.coc~cma - .FFFo Foo. om.m- on.F oo.F Fo. Fm.F oo.m - F m - F m.c0Cucma - .mmoF oo.F om.F om. om.F mc.v - oo.F m - F m.cocutma - omsuo< moo. ooo. mm. oo. oo.m - om.F. c+ . o. ”concoooocF ..F+Fo ~Fo. mm.~ Fm.w no.4 me. Fo.v oo.m- . oo.m m - F uconcoowocF - .mop. omo. no.F F¢.¢ mo.e mm. om.c oo.m - «F.~ m . F acoocoaoucF - Fmsuu< "c3a\mc_xmz commFuoo ooo. oF.e- . «w.~ om.~ ow. mm.~ oo.m - F n . F FoFCoo doFmeco>o< N8. 8.»: oo.F. E; %. oo.F 3A. - F F - F 05> to 83382 FF». mom. ~n. mo. ww.m - o o - o ma .FFFo ooocxom ooo. nh.m mmF.- cum. oF.F 0F. ww.n - mm.m- 0+ . 0. Fa .FFFo moccxom ow.~ Fo.~ FF.F Fo.~ m~.o - F n . F moccxom m.coCucma oFo. mn.~- oo.~ 0m.~ mo. o~.~ no.0 - F n - F ooocxom "mmoauFuu< oquuFFFCoFm Loam m. m .o.m m mocmm cocoa osmz moancm> u Fo oaom> m.coz m.coson cameo Fmauo< oEQmeoo "moococoFFFo xom ucooFFFCmFm FF moonaFco> cozooxmcmxmz coFmFooo new oosuFuu< LoF moFLmFuaum o>FunFcomoo 0F ooan 192 ooo. moé mud F~.m R. oo.F” m - F m - F 033: oocmcFEoo m.cocucma FF.F” mo.m -. Fo.m oo.m - Fw.~ m - F 3qu - .83 vocaUFFFcoFm Loam x x .o.m x omcmm cocoa osmz monoFLm> u Fo osom> m.cmz m.cmeoz newco omauu< monmeoa "moocmcoFFFo xom acouFFchFm FF financFHcouo mosnaFra> toxoa\mc_xm= cOFmFuao new unauFuu< rot muFumFumum u>FuaFrumoo oF moan 193 are given for each variable as well as information concerning the tests for significant sex differences. If no information appears in the last two columns of the table for a given variable, it means that no significant sex differences were found for that variable. The 252 respondents included in this sample described a wide range of dating relationships and experiences. The majority of them (61.1%) described relationships that they were currently involved in, while the remaining 38.9% focused on past dating relationships. Over two-thirds reported that they had dated these partners exclusively, while 26.1% described commitment levels that were more casual. On the other end of the spectrum, 6% reported that they were engaged to their dating partners. The average length of the dating relationships they described was 1.7 years, although this ranged from one date to nearly ten years in length. Almost half of the respondents (45.6%) reported that their partners were not students at the same university that they attended. In terms of these students' experiences with dating abuse, they reported high levels of involvement. Table 20 presents information concerning the proportion of students with experiences of sexual and non—sexual physical abuse in dating relationships during college. Fully 44.2% of the women reported experiences with either sexual or non-sexual physical abuse in the context of a dating relationship during college - either as a victim or as an offender. Of the men, 36% reported such experiences. The reported rates of 194 Table 20 Men's and Women's Rates of Inflicting and Sustaining Sexual and Non-sexual Physical Abuse. (Percentages in parentheses are based upon the number of valid cases: N=252 for SA variables; N=246 for NsPA and combined variables.) TYDg_of Datinq Abuse flgmgg gen Sexual Abuse - Inflicted 11 (6.7%) 17 (19.1%)** - Sustained 36 (22.1%) 11 (12.4%) Non-Sexual Abuse - Inflicted 49 (30.1%) 26 (29.2%) - Sustained 54 (33.1%) 26 (29.2%) Either SA or NsPA - Inflicted 52 (33.1%) 30 (33.7%) - Sustained 67 (42.7%) 29 (32.6%) Both SA and NsPA - Inflicted 8 (4.9%) 13 (14.6%)* - Sustained 23 (14.6%) 8 (9.0%) * Sex Difference significant at < .05 ** Sex Difference significant at < .01 195 involvement with both SA and NsPA are consistent with those found in previous studies (Lane & Gwartney-Gibbs, 1985; Sugarman & Hotaling, 1989). A Comprehensive View of the Phenomenon of Dating Aqqression For the purposes of this study dating aggression (DA) was defined as the combination of (a) sexual aggression (SA), (b) non-sexual physical aggression (NsPA), and (c) psychological abuse (PsyA). The following hypotheses and research questions were addressed to explore the relationships between SA, NsPA, and PsyA. Hypothesis 1: For, males and females, inflicting SA, inflicting NsPA and inflicting PsyA were expected to be significantly correlated with one another, as were sustaining SA, sustaining NsPA and sustaining PsyA. These relationships were explored by examining the bivariate correlations between the measures of each type of dating aggression for men and for women separately. Table 21 presents these correlations. For both men and women, all correlations between inflicting the three types of DA are significant. They range from .259 to .477 for men and from .305 to .385 for women. The same is true of the correlations between experiences of being victimized by all three types of DA - again, for both men and women. These correlations range from .275 to .534 for men and from .534 to .575 for women. These results support Hypothesis 1, and provide evidence that dating aggression may best be viewed as a combination of 196 Table 21 Correlations Between Non-Sexual Physical Abuse (NsPA), Sexual Abuse (SAL and Psychological Abuse (PsyA) Measures - Women's & Men's Data ‘WOMEN NsPa NsPa SA SA PsyA PsyA Victim Offender Victim Offender Victim Offender NsPA - Victim 1.00 Experiences NsPA - Offender ."286 1.00 Experiences SA - Victim .534" .212" 1.00 Experiences SA - Offender .275" .385" .286" 1.00 Experiences PsyA - Victim .575" .230" .555" .284" 1.00 Experiences PsyA - Offender .357" .380" .320" .305" .722” 1.00 Experiences IMEN NsPa NsPa SA SA PsyA PsyA Victim Offender Victim Offender Victim Offender NsPA - Victim 1.00 Experiences NsPA - Offender .926" 1.00 Experiences SA - Victim .367" .437" 1.00 Experiences SA - Offender .214' .259" .231' 1.00 Experiences PsyA - Victim .534” .574" .275" .284" 1.00 Experiences PsyA - Offender .452" .477" .307" .367" .883" 1.00 Experiences |/\ IA 2: H 'U'U E: U) 197 sexual abuse, non-sexual physical abuse and psychological abuse. The rates of inflicting and of sustaining these three types of abuse are clearly associated. Table 21 also provides support for the observation, made in numerous previous studies of dating aggression, that rates of inflicting and sustaining violence are associated. However, the patterns of correlations for men and for women are 13;! different in regard to their reported experiences with NsPA. Men's reports of victimization and offender experiences with NsPA are correlated .926, while for women, this value is only .286. To explore the possibility that this difference may be due to the inclusion of perceived dangerousness and injuries in the New Violence Scale, these same correlations were examined using The Conflict Tactic Scale items and scoring strategy. The differences, though less pronounced, remained. For men, victimization and offender experiences were correlated .868, while for women they were correlated .432. Thus, regardless of which measure of violence is used, it appears that men are mggh more likely to describe mutually violent relationships (in terms of NsPA experiences) than are women. For Psychological Abuse, victimization and offender experiences are very highly correlated among both men (.883) and women (.722). In contrast, the relationship between inflicting and sustaining sexual aggression is much weaker for both sexes (.231 for men and .286 for women). 198 While Table 21 clearly indicates that experiences with PsyA are associated with experiences of NsPA and SA, these relationships were further explored by examining the correlations of the IPA subscales to these other forms of DA. Table 22 presents the correlations between reports of partners' acts of PsyA and reports of victimization experiences with NsPA and SA. These analyses were conducted on men and women separately, but since there appeared to be few differences between the patterns of association, Table 22 includes data from both male and female respondents. Not surprisingly, the Threats and Violence Scale is very strongly associated with the NsPA variables, including the number of injuries sustained, the perceived dangerousness of the violent actions, and the severity of the injury reported. Also, the Criticism and Ridicule Scale generally has a stronger relationship to this set of NsPA variables than the remaining IPA subscales have. Thus, it appears that these two types of psychological abuse may be especially likely to occur in relationships where there are experiences of NsPA. In contrast, the experience of SA victimization seems to be most strongly associated with the Total Scale or the Emotional Callousness Scale. Additional research questions. In addition to this main hypothesis in this area the following analyses were conducted to further explore the relationship between the three forms of DA. 199 Table 22 Correlations Between Psychological Abuse Scales and Othe Dating Aggression Scales and Variables - Victimization Data Perceived Injury Sexual Injury NsPA No. of Danger— Severity Abuse Severity Scales Abuse Injuries ousness (NsPA) (SA) Threats of .654 .618 .570 .717 .454 .285 Violence Criticism & .513 .434 .516 .538 .385 .305 Ridicule Emotional .286 .212 .336 .330 .454 .322 Callousness Control & .460 .433 .494 .534 .340 .132 Isolation Withdrawal .326 .307 .315 .317 .326 .215 Total Scale .545 .482 .558 .595 .481 .312 * p for all correlations < .001 200 Comparing the feminist/social control model's relevance to each formiof DA. The patterns of relationship between inflicting each of these three types of dating aggression (dependent variables) and the independent variables predicted by the social control model to be related to the phenomenon of dating aggression were explored by means of a series of six discriminant function analyses - one each for the three forms of abuse, with separate analyses for women and men. Discriminant function analyses were chosen over regression analyses for two reasons. First of all, the distributions of the NsPA and SA variables were expected to be extremely skewed. Secondly, the variance of these dependent variables was not expected to be equal across levels of the predictor variables. (For example, since the variance on the NsPA variable was observed to be much larger among men than women, it would be expected that the variance on this measure would be unequal across levels of all the predictor variables which were found to be significantly correlated with sex.) Both of these conditions violate the assumptions necessary for multiple regression analysis (Norusis, 1988). Since the intercorrelations among the set of predictor variables has important implications for the interpretation of the results of a discriminant function and the variables which are identified as ‘discriminating', the first step was to examine these intercorrelations for the female and male respondent's separately. Tables 23 and 24 present these correlation matrices. A large number of the predictor 201 .U . UCOU Hoo. W Q I iii HFo. V. no mo . w o .. t. oo.F mF. FN. FF. so. oF.- oF.- oF.. oo.- Amouu< at a «it a: oo.F no. Fo. oo. so. quu Fo mucmuamuu< oo.F 06. Am. So. Acuuamc< cofluocsm ucmcflfifluomflo CH comb mmHQMHHm> youoflcmum Mo wcoflumHmuuoouwucH mm OHQMB 2()2 axono. mu no HAS. V. no mo. w o z .1 .oF. no.. mo. mo.. mo. ..F~.- .mF.. F~.- 80.. F~_omav UFEOCOUS . macarocc_o :33: 333 ..-. no. oF. oo.- so. oF.- oo. oo.- no.. a .aCOmroa . oucmrochO Fo.- so. Fo. oo. oo.- ko.- ~o.- mo.- ~o.. Acouwo u_eocouw oN. oF. o~. o~.. oo. -.- -.- F~.. oF.. Amroao saFuom a racomraa at. ta .5 area. at 1!: 89.5088 ~n.- Fo.. mF.. oF. oo. on. oN. on. oF. Arse. «snag roses ».rocuroa it I. I it... {ii iii t. ..so.- ~F.- .FF.. oo.- oo.- FF. FF. .oF. oF. FF.Foo mco_m_uoo.m_;mco_uo.u¢ oF.. Fo.. o~. FF. No. oF. FF. nF. mo. x~eFoo «coFm_uoo m.rocorma iii .1! .4. o~.- ~F. oo.. oo. oo.- o~. FF. oF. Fo.. FFSFQV acoFm_uao Fcoucuaonc_ It icC It! I 1 8.30m 09.20::— mo. Fo. oo. .FF. no. so. FF.- oo. NF. Fn.o_o aco_m_uomtm_;mcoFuasoa Fo. om. o~.. mF. oo. No. no. FN. AN. x~.oso «co_m_umo m.r~c3roa as; a: c a; :3. mo.- om.- Fo. oF.. FF.. oo. ~o.. -.- o~.- FF.o_o «coFmFUoo scoocmaooc_ tic it! i it. t! rose; .aooF mou< ~.o< :2 on: my: 32 nu: ~uu< Fuu< .F.o.ucooo mm anme 2()3 H00. Wm HO. IVIQ ill. mo. .v. o .. oo.F oo. oo. mo. -.- sF.- m~.- no. so. s~.- oF. FNFomos oFeocooo - oocoroooFo ii I. I! it. AFFomoo .o_oom oo.F oo. ms. FF.. F~.- mo.. oo.. No. No. so. s .ocosroa . oucoLsooFo .Ifii It oo.F .oF. so. ~o. Fo.- so.- No. ~o. ~o.- accuse o_socoom oo.F F~.- o~.- oo.- o~.- mF.. no. Ao.- Fmroos .oFoom o .mcosroo Ci Ii it C mflogaomou ...sm. oo. oF. FF. no. mo.. nexus moooo roses m.rocuroo oo.F oF. mN. on. Fo.- oo. FmsFos chFsFuoo mooscoFooooo the use oo.F ss.- oo. om. ~n.. F~oFos scoFm_uoo s.rocormo at use at oo.F oo. on.. on. FFoFoo mcoFscuoo scoocoooocF i: ate «0.30m coconut 3 oo.F Fo. NF. Anou_o chFmFuoo oogmcoFooooo oo.F tso~.- F~.o_o chFmFooo m.rscormo oo.F FF.o_s scOFsFooo scoocoomoco rosoo soon. ~Fomo FFomo couo mesa too moFo NsFo FoFo moo. ~.o_ F.o_ .A.o.ucooo mm magma 2(14 .muFHHFanU Foo. w o . I it. .FFV. ”v AF mo. w m : fi oo.F sF. oo. NF.- ~o.- Fo. so.- No. oo. Fmooo< C C! i?! oo.F oo.- mF.- ~F. «m~. FmFF Fo socooooou< oo.F FF.- oF. sm. Fsoo Houoflomum mo mcoflpmHmuuooumucH @N OHQMB 205 .muFAHFanu Foo . w a I flit. Ho . v m mo. w o : ' ~o.- so. Fo.- mo. oF.- Fo. mo.- mF. «F. FNFowao uFEocoow - oucocoFFFo FFFommv .mFoom mo.- no. am~.- oF.- mF. oo.- eom. oo.- mF.- a decomcoa - oucocoFFFo mo.- ~F.. mo. sF.- mo. so. No. 00.. oo. Fcoumv umeocoom oo.- mo.- oo.- sF.- FF. oF.. mF.. mF.. so. choav .mFuom a .ocomeoa moucaomoa m~.- -.- mo.- om.- nF. 0o.. mo. FF.- oF.- Ferny moeou carom m.cochoo «a a a; a 00.. oo.. Fo. om. om. mo. NF.. oF. ow. FMFFoo mcoFmFuoo chmcoFLoooa a: u so I «a oF.- o~.- -. mF. um. oo. oo.- kw. mm. FNFFoo wcoFmFuoo m.cocucma s cc c 5!. a; the mo. hm. Fo.- mo. no.- oF. no. «0.- oo.- FFFFoV mcoFmFUoo ucoocoooocF C?! .91. museum oucocowwmo mF. so. no. mF. mm. MF. oF.- «m. we. Anon—v mCOFmFuoo QFmeoFuooom it... I {it {58. oo.- mm. on.. NM. mF. ow. mo.- no. me. FwouFo mCOFmFooo m.eocucma {at a: «a a a :5 a: «F. ~¢.. om. mo.. mF. oo. mo.. m~.. oN.. FFEUFV mcoFmFooo ucoocomoDCF is as; at a. .638 .83 20¢ ~..o< Fdo< 03¢ muu< cuu< nuu< ~uu< Fuu< .A.o.ucooo so magma 2()6 Hoo. .v. a HO. WQ .3; mo. w o .. oo.F so.- .NN. mo. NF.- NF. no. so.. NF. no. FF.- FNFomoo o_eocooo - oucoroosFo FFFosoo .oFoom oo.F o~. No. F~. oo.- oo.. mN. so.. Fo.- no. o .mcosroa - ooooroooFo a. {it 8 ii oo.F ...~m. so.. oF. sF. so.. so. No. Fo. Accooo u_eocoom oo.F No. oo. oo. so. Fo.. so. ~o. Fmroao omFoom o oncomroo «oo.—Dome“ oo.F sF. ..om. oF.- oo.- oo. AF.- Frsoo moooo roaoo m.racurmo oo.F mm. no. so. os. ~o. “noFoo scoFm_ooo oFomcoFFoooo it! fit .Iii oo.F ~s.- om. oo. o~.. x~o_oo scoFmFuoo m.roc8r~o if: a: as; a. oo.F oo. oo.. s..om. FFoFoo scoFmFUoo Fcoocooooc_ mocoum 00:95::— oo.F ...os. oF. Fmoo_o scoFs_oao oFPmcoFomooo oo.F «ammf 3:63 mcoFmFuoo 935.com oo.F “Floss chFmFUoo Fcoocooooc_ carom Femo— ~_omo FFomo coom some Lao mFFo NoFo Fo_o m.o_ ~.o_ Foo. .A.o.ucooo so anoe 207 variables were found to be significantly correlated -- some very strongly so. This was particularly true of the relationships between the difference score variables and the variables from which they were derived (the two sex role difference scales, the decision-making power difference scales, and the resource difference scores). Also, the decision-making scales which describe the actual and ideal power over the respondent's independent decisions were strongly and negatively associated with the same scales pertaining to their partner's independent decisions. In other words, these two domains of influence tended to be experienced reciprocally, such that if a person reported having a lot of influence over their own decisions, they tended to report having little over their partner's decisions. The same was true of the ideal amount of influence they wanted to have over these types of decisions. Having such highly correlated predictor variables can be problematic to interpreting the results of a discriminant analysis. One solution to the problem, reccommended by Norusis (1988), is to utilize the correlations between the predictor variables and the resulting discriminant function to assess their relative importance in differentiating the groups of interest. This is preferred over examining the list of variables which are chosen to enter the equation or comparing the discriminant function coefficients for the predictor variables. Both of these latter methods of comparing the importance of the predictor variables are based 208 upon an assessment of each variable's unigue contribution to the function. Therefore, they will systematically underestimate the importance of variables which are strongly associated with group membership but also with other predictor variables which have already entered the equation. The criterion groups for the discriminant analyses were defined by the categorical variables discussed earlier and presented in Table 8. Three NsPa groups were defined on the basis of The New Violence Scale scores: Group 1 consisted of those individuals who did not report any experiences of inflicting NsPA upon their dating partners; groups 2 and 3 were of approximate equal size and contained individuals who reported mild and severe levels of inflicted NsPA respectively. The two SA groups were defined on the basis of their responses to the SES. Group 1 reported no incidents of inflicting SA against a partner, while Group 2 reported one or more. It was not possible to distinguish between mild and severe offenses, since so few individuals reported incidents of inflicting SA. Three PsyA groups were defined on the basis of the IPA Total Scale scores. The groups were created to be of approximately equal sizes, with Group 1 consisting of the lowest scoring one-third of the respondents, Group 2 consisting of the middle scoring one-third and Group 3, of the highest scoring one-third. Tables 25 and 26 display the means for each of the criterion groups on the predictor variables. Table 25 presents the data for the female respondents, and Table 26, 2()9 .o ucoo o0.~ Fo.~ mo.~ oo.m no.m oo.m no.m mo.m mcoFmFuoo aFmeoFumoom w0.F ~0.F mm.F oo.F ~0.F Fm.F om.F F0.F mCOFmFUoo m.cochma 00.0 00.0 Fn.0 00.0 00.0 nn.0 00.0 «0.0 mcoFmFooo acoucoaoocF "oozes omen— ow.~ Fo.~ mo.~ on.~ ww.~ ~m.~ «w.~ ow.~ mCOFmFuoo o_cmco_um.ou oo.F ms.F m~.F mm.F mm.F Nm.F mv.F sm.F mcoFmFooo m.cocLema ansz 00.0 00.0 mu.0 00.0 m0.s « om.s no.0 00.0 mcoFmFUoo ucoocoooocF "oozoo omsuu< os.~ ss.~ oF.~ sn.~ 0m.~ mm.~ ~0.~ 0~.~ mFoFme omFLomco>o< « 0~.F oo.F o0.F m~.F m~.F mo.F uw.F oF.F ouco.o_> +0 oucmuaooo< 00o.F wo0u. Funk. ammo.F sme. oo.F moon. oFow. Aom>.moom .umcuco.mFm ooo>om "masoco moan< .muFmoooco>ma .msxom-coz manage coocoFFo mso_cm> cF coeo: coo woonoFLm> cOLUFUoca co ocoom coo: mN moan 210 .Fo. w a an moucocoFFFo Queen coozuun ucaoFF_coFm mowu>or <>Oz< Fm: oco ¢. .3. w a on 3953:? anacm cmoxuon acmoCchFm 385.. <>Oz< >2. one « 0mF u .3qu mm" 2 mm" 2 we 2 0mF u .32 FF n 2 Q: n z omF n :32. FN u 2 mm n 2 N3 n z mw0m.- ow0F.. KFFI ummwf 35.. am}; Famof oooo. haw... oFEocoom wocooowto mmoof msuof 38.. 300. 33.. mFoof moo. 35.. m w a oococmwto NNNo. - ammo. no: . - «a NoFm. omoo. - omso. 0m? . .0.000. - o .2553 18sz 33.. «~09. FmFN. £3. 53.. 0an.. ooFF.- mmmo. 3.88 w oncomcwa "mQULJOwwm os.m mF.m oo.m « s0.m 0F.m om.m mF.m o~.m "momoo Lozoo m_omcooma mmom. ommF. FnoF. whom. NmoF. mw0N. ommm. ms0F. mcoFmFooo achcoFFaoom ano. soFo.. 0FoF. ooFF.- o00o. Ammo.. onmo.- mm0o. mcoFmFooo m.oococoa am\:~ 3mm. momof mmwof wowm. Nmso. arm}: 000m. mumm. smmof mcoFmFowo “Succumbs ”mmeoom mocoLmFFFo chucumFm ocm>om no: :3 3» oz . chucud Fm ooo>om B F: ocoz ”mango m N F N F m N F mm30< omoFmoFozoxma om30< owsxmm .mumecm oaoxom-coz omsn< A6. uaooo no. «32 211 6.28 NF.m oF.F” mo.m m~.m wo.m mo.m Fm.m oo.F“ wcoFmFuoo 3:93.533. 0&4 NwF 00F mod o0.F ¢m\:F FF.F Nm.~ wmé mCOFmFUoo mlocfimm F00 3.0 ~00 3.0 mmé 00.0 mi 0m..s mCOFmFUoo “concoooUcF "oozes .mwoF 0o.m wo.m 00d NF.m Fo.m oF.F” MF.M mo.~ mco_w_umo QFzmcoFumomm caM\:~..M\:F 05F mm.F F~.F a: mo.F Fm... :QJF owé n0.F -.F 23293 mloctmo ¢«m\:~..m\:F MFJ om...» FF..» .2. 09m F00 :M\3F..N\3F 3.0 NFJ N00 23225 ocwpcwaoch ”owzoa .msuo< F~.m 0n.~ F0.~ 00.~ mod oo.m No.~ SUN mwofiom omFmecm>U< 0o.~ 0m.~ 00F mF.~ nod om.~ him woé oocooo; *0 oucmuaouu< och. mam. me. 09F 00h. NMN. mo.F 005. Cm>.mnwm no: 303 mm» oz .umrucu.oFm oco>mm 03F: mcoz "menace m N F N F m N F eman< .moFmooozu>ma omsn< oaaxwm om30< .muFm>;o .asxmw-coz museum consowwo mooFLu> CF co: Loo moFQaFLa> coquuoru co «Loom coo: 0m 030... 212 a on 383320 aaoco c0053 ucmuFtcmFm Emmi: <>Oz< .6: mco f. .FD. W .3. w a on moocmcowto 9.6.5 5253 ucmuFtcmFm 283.. <>o=< >2. 0:0 I. Nonomqu 0Nuz mNuz oan mwnEFoF NF": 00": monomSF sFuz NFuz mmnz msm. 0mm. mom. m0.N KN N00. 3m. mNm. oFFoocoow oucooowto 3M\:F..N\:F FNN. FON. mNN.. ; mFm. NNo. c NNN. 0Fm. mFo.- w a a 358:3 oFN. 3N. moFf RF. omo. Roi mos. 08. 0:353 Fmo. me. «oF.- 3o. mmof Foo. who. moo! 388 a 3:033 "moucaomom me moN $.N m0.N EN 2N mwN m0.N "momoo Luzon Preston .53556 303m 00: :3 mo» oz FmoucoéFm 836m 3:. 0:02 "mango m N F N F m N F 33.2 333w ow3< 3292.... omaxoméoz om3< .muFmooocoxmo To. £58 3 3...: 213 the male's. The tables are included to aid in the interpretation of the results of the discriminant function analyses. One-way ANOVA's were conducted on each of the predictor variables, comparing the means of the groups defined by the three different DA variables. In the case of the NsPA and PsyA groups, if there were significant group differences, post-hoc comparisons were computed to determine which group means were significantly different from one another. The results of these significance tests and post- hoc comparisons appear in both tables. Tables 27 through 32 present the results of the six discriminant function analyses: two for each of the three types of dating aggression, one for women and one for men. Tables 27 and 28 display the results pertaining to the NsPA groups; Tables 29 and 30, those pertaining to the SA groups; and Tables 31 and 32, to the PsyA groups. These analyses are relevant to all of the three main research questions addressed by this study: the similarities and differences between the three types of DA; the role of the Feminist/Social Control Model to understanding dating violence; and, gender differences in the experiences of DA. For the moment, the focus will be on the results that are relevant to the first question. The pertinence of these analyses to the other two issues will be discussed later. There are a number of ways to assess the effectiveness of a discriminant function. The first question is whether or not the function achieves significance. If it does, one can 214 .oocmco >0 ooFFmem.u >ouoocooo on ooze: xw.w0 .mmmmo Fo Rm.FF >Fmemoo >ouooocoo mcoFuocao xo.oF No.m NoMN. moF. Fo.mF msomo. N xN.mF RN.mF oFoo. oom. on.Fo FmoFF. F can poucaouo< coo poucaooo< oucmuFF_coFm coFNoFooeou mucoFoo> Fo a usoo> cooFm coFuocso x m>Fuaosesau oucmFoa> oo>comoo ooooF x .ao_cocoo wcoFmFooo ucoocoaoocF «NNs. ooF. Foo. 0F0. eu>o euros *0 .ue< Favo— mcoFmFooo acoucommocF 00F. enos.. Foo. on». co>o carom Fasuu< ooo.- amns. ooo. New. oFoFFom FoFLamco>o< moocaoooa uFeocoum 00N.- «NNm. Foo. sow. m.ooa30u cF mmococoFFFo mcoFmFuoo acoucoaaUF 0mm. asF0. 0oo. 0mo. oooum .FFFo .mouu<-.moo_ N coFNcho F coFuocau oucmuFFchFm moose; m.xo_3 ooooucw moonoFom> mcoFFocsm a modnoFom> comzuom mcoFuoooccou .5603 .8395 30530 - "omen? 339:5 338.0752 - ESE-5 $932.2 acocFEFeomFo NN ooooF 215 60:20 .3 Eta-.0 30000.80 3 3:0: woos. .0030 F0 amfik 3.5030 3000.200 2.0.505“. Rofim No.2 Leno. an. x5.m~ NBS... ~ __ NF .on RF .3” oooo . m~o. Ram . .K whoqo. F .8; “0005003 .3; .0353: 00:00:75; cote—00.80 00:02; 00 a 33> :33 c3005: u 03033550 005...; 002030 .30... x 30.580 9.0.3.53 3:95:33. moo: co-. ooo. com. 00>0 .530; duauu< $30.83 0:00:80“!— Nom. Lomf ooo. F~m. ..0>o 0010.. .333 .35.- Nmo. ooo. Fem. 3003003. 0.5.953 000.530“ 8.358 F2. LE. 8o. 03. «.398 E 382038 woof .FFF. ooo. ~No. 00>Fuox 0030a 9.55.50 ‘- 9.3303 0105.5... .80. o-. 08. 30. .53 .532 00 0522 :33 .03.- 8F. ooo. mp... 080.0; we 00500003 23303 0105.50 so.. INS. ooo. ooo. .63 .530.. .033 ~ 8.80:3 F 339'qu 00:00:.E3m 03:3 0.0:; .0335 0032...; 05205; a 333.5 c0033 202300.80 .5: .2595 Lou—00:0 . "03 .0032... 33.3%.: . tug-5m £0392 05:25.30; ”N Snub 216 .00cmcu >0 DQFFmemou >.0000000 on 0‘30: xm~.mo .00000 Fo xm~.mo nonmemou >.u000000 coF00c3g RF.FF wooo. nmm. XOQF oommF. F 00¢ noucaouu< 00:0F00> 00:00FFFCmFm 09530 .33 a coFuadmggou .00_cocoo 00:0F00> F0 a oaom> c0m_m co_00c3u -~. Foo. ooo. moogsomoa .mFuom a .mcomgoa mmuaumuu< ooogxom Fan. moo. ooo. gocugaa cF 000:000FFFo co_m_00o 0.00:0000 o-.- «oo. moo. 0000m.FFo .0300<-.mmn_ mco_m_00o QFmeoFuaoom Fo~.- coo. hFo. 00>o 00300 *0 0c:OE< .000” hmq. moo. mmo. 0>Fuoz 00300 0.00:0000 oFo. ooo. omo. mougaomma 0Fsoco0m F 00000; 00:00FFchFm 009504 0.x._3 nogmucm moonmmgm> wcomuucau UCQ madnawLfl> cmmzumm mcomumdwggou .c0eoz .ma300o goncoFFo . "0m:n< .maxom - >0meeam mmeomc< FCoCFEFgumFo ow 0.00F 217 60:20 .5 00.0—000.0 30000000 00 0.30: 30.8 .00000 00 «00.00 00.0.0030 30000000 0030030 _ no...” 88. 0%. .82 32.... . __ 000 03.50% 0000.0; 030008 .30» 00 a 00:00.25; 00.00.0008 .00.:800 0000.0; 00 a 0:.0> 0080 00.0050 03. ooo. .000. 00003000.. 0.50008 00003000.. .0.000 a -m. ooo. 30. .0000000 0.0.0300 0. 0000000.... 0020.000 0.0000000 000. ooo. 00.0. 00>0 00:00 00 0030.02 .030< 000.0..000 0:000 00:. Fmof ooo. 000. 0006 00:00 00 00:82 .000. 0:03.000 02.000.00.00 mmw. ooo. SK. 00>0 00:00 00 0039.2 .000. ”0.30.000 «F0... ooo. 000.. 0000000000. $8 00300 .033 F Ema. 0003.00.00; 0005.. 01:: 03300 $30.03, 00.00030 000 00.00.02. 5000500 gmuldofau :0: .0905 000:0va .0039 .0360 - 30.0.50 0.0305 000550003 on 0.00... .000000 3 00.0.0030 30000000 on 0.30: uni» .00000 .0 3...}. 3.0030 30000000 000.0050 uném Rank .3... kN. uo.o~ 938. ~ «0.0. no.0. ooo. ~n0. x0.nh mom-. . 000 0005003 000 0005002 000003.00.» 00.00.0008 .00.:003 0000...; .0 a 0:.0> 000.0 00.0050 a 3.00.2050 00:0...0> 02.0008 .000. a 2113 .5... 3m. 08. .00.. 83:0: 32. .30 £398 E 89.20:... .52. 00.... 08. ~00. 82:83 0.8800 0.3030 c. 30500.03 ~00. .03. 08. E. 30:00 .2822: «00.5081 .00... :0. 08. 02. .380. 0 .082: 0.0.98 c. 8823.3 .3». a. 08. 08. 8288.. 0.3800 .30. .- 08. 90. 08.0.80 080000000. 026 00.60 .032 05. .2»: .8. 80. 20.280 3.55.: .53 023.. .302 to. 50.. 08. 000. 80.58: .380 0 .208: ~ 8:050 . 8:050 308.....5 .38.. 5:: 0235 830...; 000.0050 0 0030...; 000300 000.00.00.80 .0060: .0995 00000.3 30:2 30.00.2030 . Elia 0.0305 0000.07.00... .n 0.00. 69.2.0 .5 3.3330 339:3 on 3:03 nmfin .326 “.0 5.3 33.330 332.3“. 2.335. 219 x3... x5: «8. 3.... .86... :3... ~ 5.3 5.3 8o. .3». 5.2 3R0. . .3. v9.53: .3. “.3533 088.25.... 5.3.3.60 .3233 35...; we a 33> 53m 5.35“. a 2,333.50 one-t; not-Bo .30» x .3. NS. 8... NS. 88:3... 32.3... 3.8:: «5.203 8... 8.. 8o. 5. 3.5333: .33 .28.. ,6 “595 .33 30.393 n3. $3.- 8o. 3. “5.8%. .56 .58.. B “522 .33 cum. moo. ooo. 03. 30.533. 3.53m «.393 c. 3895:... 2..- .3. 8o. 5. 22389 9:23:32 .26 .83.. :32 .3... to. So. o3. 9:2. 3:2. 9.8:: .63.- “8.. 8o. 3“. 8228a 22:8-28... .ts .3838! 6%.. SN. ooo. new. 33.09 .3330): 3...- .03. 08. ooo. 39530.. 35:03 «3.530.. .208 S»..- .03. 8o. .33. B. .8039. 2398 5 8889:; .am. fig. 8... 3». 2338.. «7.8:... :3 .28.. :33 ~ 8:82 v 8:82 858:8; .35. 3:5 2:35 83.2.5 "‘ H h 23.35“. amniotg c0233 93339.58 :0: 690.5 .5233 33$ 30332.33 $.35!» 3939.! 22.57.32. ~n 0...... 220 reject the null hypothesis that the population means for the criterion groups (on the discriminant function) are all equal. This significance test is based upon the statistic Wilks Lambda, which is smaller for cases where the between group differences are much larger than the within group differences. The second way to assess the resulting function is by examining the percentage of the total observed variance that is accounted for by group differences. Finally, it is important to consider the function's ability to classify cases into criterion groups correctly on the basis of the observed values of the predictor variables. In evaluating the effectiveness of the function in this last regard, it is important to compare its performance to what would be achieved by chance. In the case of groups of equal size, this is a straight forward matter: if there are two groups of equal size, you could achieve a correct classification of 50% of the cases by chance; with three groups of equal size (as with the PsyA groups), 33% could be correctly classified, etc. In the case of the NsPA and SA variables, though, the situation is a little different. For both of these variables, group 1 (the non-abusive group) is much larger than the others. In these cases, the percent which would be correctly classified by chance is equal to the "hit rate" that would be achieved if all cases were arbitrarily classified in Group 1. 221 The information necessary to evaluate the effectiveness of the functions according to all three criteria are included in the Tables 27 through 32. In comparing the effectiveness of the discriminant functions which resulted for each of the three types of DA, it appears that this set of predictor variables do the best job of discriminating between the PsyA groups. As compared to the functions discriminating between groups defined on the basis of the other two forms of DA, the resulting PsyA functions, for both women and men, achieve a smaller value of Wilks Lambda, account for more of the observed variance and do a better job of classifying the cases correctly when compared to chance. Using these same criteria, the functions discriminating between the NsPA groups appear to perform next best, and those for the SA groups do least well, which is not surprising, given the small number of individuals of either sex who report experiences with inflicting SA against a dating partner. In comparing the variables which enter the functions, it is helpful to examine Table 33. It displays the correlations between each of the predictor variables and the first discriminant function for each of the three types of DA. As discussed above, this is the preferred way to identify the relative importance of each of the predictor variables to the function when the predictor variables are intercorrelated. The table includes only the correlations of variables with the first function derived for two reasons. The first 222 Table 33 Predictor Variables' Correlation with the First Discriminant Functions for Each Type of DA. (Rankings, from high to low, appear in parentheses.) U 0 H E L M E Ll Variables NsPA SA PsyA NsPA SA PsyA Sex Role .106 -.075 .235 .061 .163 (9) .088 Partner's Sle .087 .283 (5) .502 (2) .050 .200 (7) .097 i Sex Role Dif 1 .007 .391 (3) .366 -.017 .016 -.006 a Sex Role Dif 2 .088 .221 .310 -.012 .072 (17) .063 Acpt. Violence .018 -.060 .075 .188 (9) -.062 (18) -.008 Advers. Beliefs .675 (6) -.052 .667 (3) .006 .136 (12) .261 (5) a ‘3 * Actual Power: lndep. Decisions -.693 (3) -.159 -.386 (6) -.507 (2) -.716 (1) -.336 (3) *a t *a ta * Partner's Dec'ns .160 .095 .379 (5) .612 (1) .676 (3) .632 (2) a *3 *8 * Rel'shp Dec'ns .015 -.106 -.026 .220 (7)8 .118 (13) .163 (9) 3 Ideal Power: lndep. Decisions .108 .020 -.225 -.295 (3) -.051 (19) -.268 (7) a a Partner's Dec'ns -.165 -.183 .302 .220 (7)* .316 (5) .250 (6) Rel'shp Dec'ns .076 -.291 (6) .130 .021 .255 (6)3 .100 a Actl/Idl Diffs.: lndep. Decisions .616 (1) .163 .196 .239 (5)* .695 (2)* .116 *3 Partner's Dec'ns -.319 -.278 (6) -.033 -.252 (6)* -.177 (8)* -.065 a Rel'shp Dec'ns .011 -.000 .070 -.151 (10) .101 (15) -.028 Partnr's Dominance .060 .657 (2) .262 .111 (12) -.161 (11) .017 a 0 (cont'a.) * The criterion groups had significantly different scores on these variables when compared with One-way ANOVAs. 9 These variables entered the first discriminant function. Table 33 (cont'd.) 223 H 0 H E N H E L Variables NsPA SA PsyA NsPA SA PsyA Resources: Pers'. 8. Social -.256 .221 (7) -.566 (1) -.021 .168 (10) .158 (8) a *3 Economic -.060 .610 (1) .213 .032 .106 (16) .266 (6) *a a a Diffs - P 8 S -.080 .103 .011 .126 (11) .322 (6) .659 (1) fl .3 is Diffs - Econ. .522 (2) -.031 -.229 .212 (8)3 .085 (16) .065 224 function is always the one which accounts for the greatest proportion of the variance. When there are three criterion groups (as there were with NsPA and PsyA) two functions will always be derived. (In analyses with two groups, though, there is only one resulting discriminant function.) In the three-group case, the first function attempts to identify those variables which discriminate between group 1 (the non- abusive individuals, here) and the other groups (those indicating experiences with inflicting abuse). The second function attempts to identify variables that discriminate between group 2 and the other groups (in this case, between individuals with moderate levels of experience with abusive behaviors and the others). In this study the variables' relationship to the first function -- which distinguished between abusive and non-abusive individuals -- have more theoretical importance. The correlations for each discriminant function are ranked from highest to lowest in Table 33 according to the size of their absolute values. The number of predictor variables which were ranked varies from function to function (from four to 19). Variables were ranked until all of those which were entered into the function as significantly distinguishing between the groups were included. The rankings are presented to facilitate making comparisons across the functions. The table also indicates, for each predictor variable, whether the criterion groups were found to differ significantly in 225 the One-way ANOVA's and whether or not the variable was entered into the first discriminant function. The table indicates some clear and important sex differences, which will be discussed more later. For now, suffice it say that this set of predictor variables does a better job of distinguishing between abusive and non-abusive men than women. And, further, fewer patterns -- either similarities or dissimilarities -- emerge in comparing, across the three types of abuse, the functions discriminating between the women's criterion groups. But, even when considering the men's data, few consistent distinctions between the three types of DA are apparent. Some variables which seem to differentiate between the three types of abuse in the discriminant functions, did not have significantly different group means when compared via the One-way ANOVA's. For example, the Ideal Power Scale for Relationship Decisions was a ranked variable for both men and women only in distinguishing the SA groups. Yet, while men who inflicted SA did report having ideals of higher amounts of power over relationship decisions than did non-abusive men, the differences were not significant. And, for women, the difference between the two groups on this variable was miniscule. It appears that there are more similarities between the three types of DA than differences, especially in regard to the decison—making power variables. Among both men and women, increasing levels of abuse are associated with 226 decreasing amounts of actual influence over one's own independent decisions. This is true for all three types of abuse among the men and for NsPA and PsyA among the women. In addition -- for men, but not women -- increasing amounts of actual influence over the dating partner's independent decisions was associated with increasing rates of inflicted abuse of all three types. The other similarity across all three types of abuse, apparent from examining the men's data, is that increases in abusiveness are associated with significant increases in the reported difference between the two dating partners' Personal and Social Resources. As abusiveness increases, men report having more such resources than they describe their partners as having. What can be concluded from the results of these discriminant analyses, as they pertain to the issue of the relationship between the three forms of DA, is that, this set of predictor variables accounts for a significant portion of the variance between both male and female criterion groups defined on the basis of each of the three forms of dating abuse. Furthermore, there appear to be important similarities between NsPA, SA and PsyA —— especially in men's experiences as offenders. Emotional consequences of experiences with SA. NsPA. and PsyA. Respondents who reported victimization experiences with each of the three forms of DA were asked to rate the seriousness of the emotional effects of their experiences with each type. A Repeated Measures ANOVA (Gender X Type of 227 DA) was utilized to compare the differences between the reported emotional consequences. The results of this analysis are presented in Figure 3 and Table 34. Respondents who reported at least one experience of victimization with both NsPA and SA and who were also in either the moderate or severe PsyA group were included in the analyses. There were few respondents who met these criteria: 26 in all; 18 women and 8 men. The dependent variables were the three Emotional Impact Scales described in Table 11. The analyses indicated a significant main effect for sex, a main effect for type of DA and a significant interaction effect. Women reported severe emotional reactions to all three types of DA (PsyA i = 3.36; NsPA Y = 3.31; SA Y = 3.28). In contrast, while men reported having similar severe emotional reactions to experiences of PsyA (Y = 3.19), they report being successively less emotionally affected by incidents of NsPA victimization (Y = 2.63) and SA victimization (Y = 1.63). Since the number of respondents eligible for inclusion in the Repeated Measure's ANOVA was so small, these results were checked by examining the mean scores for the whole sample on the three emotional impact variables. T-tests were conducted to compare the significance of the observed differences between the female and male respondents. The results were similar. Women reported similarly severe emotfonal reactions to all three types of DA victimization (PsyA'Y = 2.91; NsPA Y = 3.09; SA'Y = 3.07). However, these analyses resulted in a slightly different pattern of reactions to the three types of 228 5“ u u 00 3' g 3.31 3;; ° Woman (N218) £10704“. WAC? N N u \. u... (m) ‘ in 9:911 st’A SA rm: or name Anus: £1!me Figure 3 Comparison of the Emotional Reactions to Victimization Experiences Across the Three Types of Dating Abuse (Repeated Measures ANOVA) Table 34 Results of Repeated Measures Analysis of Variance Comparing Men's and Women's Emotional Reactions to Three Forms of Dating Abuse ANALYSIS OF VARIANCE Source of Sums of D.F. MS. F Sig. Variation Squares Between-Subjects Within Cells 57.22 24 2.38 -- -- Effects ' Constant 557.78 1 557.78 233.95 .000 Sex 11.60 1 11.60 4.87 .037 Within-Subject Within Cells 23.58 48 .49 -- -- Emu“ DA Type 7.66 2 3.83 7.79 .001 Sex X 6.26 2 3.13 6.37 .004 DA Type 229 DA for men. They reported significantly less emotional upset than women to both PsyA and SA (PsyA Y = 2.49; SA Y = 1.75). They reported less emotional upset in response to NsPA victimization than did women as well (7 = 2.71), but the difference was not significant. In combination, these results indicate that women reported similarly severe emotional reactions to all three types of DA victimization. In contrast, men found SA victimization to be only mildly upsetting, but reported stronger emotional reactions to the other two types of DA, though they reported generally less severe emotional reactions than did the women. Physical consequences of sustaining SA and NsPA. Injuries sustained by respondents as a result of their victimization experiences with SA and NsPA were also compared. Fifty-four women (33.1%) reported NsPA victimization experiences. Of them, 32.7% reported suffering some injury as a result. For the 26 (29.2%) men who reported being the victim of NsPA, injuries were somewhat less common: 21.4% of them reported suffering an injury. None of the men reported experiencing anything but mild injuries, while 9.6% of the women (n=5) reported moderate or severe injuries. Moderate injuries were defined as, "including serious cuts, black eyes, extensive bruising or swelling, a broken nose or finger, strained muscles or joints". None of these differences were statistically significant, however, due to the small sample sizes. 230 The situation was somewhat different for experiences of SA victimization -- where the sample sizes were even smaller. A total of 36 women (22.1%) and 11 men (12.4%) reported experiences of SA victimization. Of the women, 22.9% reported suffering mild or moderate injuries. Moderate injuries are defined as above; while mild injuries, "include redness, lingering pain or soreness, small cuts, lumps or bruises, etc.". None of the men reported any injuries. Again, because of the small sample size, the differences were not significant. To summarize these results concerning the comparison of the physical consequences of NsPA and SA victimization experiences, for women both types of experiences result in injuries fairly often, though NsPA seems to be slightly more likely to result in injuries and is responsible for all of the severe injuries, and most of the moderately severe injuries reported by this sample. For men, SA victimization did not result in any injuries among these subjects. NsPA victimization experiences hold a much greater risk of injury for men than SA experiences, but no men in this sample experienced any moderate or severe injuries. Thus, in general, NsPA appears to have the greater risk of physical injury among a college dating population and women are at greater risk of serious injury. Theirelationship between DA experiences and measures of psychological symptoms and social adjustment. It was predicted that victims of all three types of DA would report 231 significantly more problems with depression, anxiety, interpersonal sensitivity, somatization, hostility, and a global index of psychological symptomatology than would individuals who did not report such victimization experiences. Tables 35 through 37 display the results of a series of ANOVAs which were conducted to explore the relationships between victimization experiences with all three types of DA and psychological symptomatology, as measured by the BSI scales. Separate analyses were conducted for men and women. Table 35 displays the data exploring these relationships with NsPA experiences; Table 36 with SA experiences; and Table 37, with PsyA. Increased levels of victimization for all three types of DA were found to be associated with increased levels of symptomatolgy on the Somatization, Anxiety, Hostility, and Global Severity Index Scales, as predicted. This was true for both men and women. In addition, all three types of DA victimization were found to be associated with increased scores on the Paranoid Ideation Scale. Interpersonal Sensitivity Scale scores were found to be associated with NsPA and PsyA victimization experiences, as predicted, but not with SA experiences. Increases in scores on the Depression Scale, contrary to predictions, were not associated with higher levels of either SA or NsPA victimization, but they were associated with experiences of PsyA victimization. In comparing the patterns of association between psychological symptomotology and experiences of victimization 232 me bro. mc w~.m no.m «v.3 cm.m Pp.q 0P.q >umdmso a_cmcomumsom m: we we o—.m F—.m cm.m mq.m om.m em.m comuumwm_umm .msxom mc woo. ~mo. Fm.m wm.m on.m em.m wo.c cw.m co_uumcm_umm a_zmco_umdom mc ooo. ovo. mu. on. F0. no." Po. we. _mo-_mm we m: m: um. Pk. oo. kw. FR. mm. >ma._mm mc Poo. we on. us. mm. mF.F ow. no. cma-_mm m: m: m: NN. on. FN. 03. Fe. om. noca-~mm mc ooo. we w~.— Na. um. No.P oo.F mo. mo:._mm mc ooo. ooo. ch. as. oo. mm.F r~.P mo. xc<-~mm me we we ks. mm. oo. -.F vo.w ww. amo-~mm mc nmo. bro. wk. co. m». me.w no. Po. uc~-_mm m: —mo. mc NP.F om.F mo.w om.F em.— mo.F ulo-~mm mc ooo. coo. cm. on. «N. mm. mm. ~m. umeom-_mm mcomuom-couc_ sacco-uoocmm c_m: xow-uoocwm c_mz oco>mm od_x ocoz oco>om odwz ocoz modnm_cm> “concoamo mpuwmmm mo muz om3n< .mu_m>ca .myxom-coz A.<>oz< mo muadmou can mcmoz manouov .Coaoa one so: you moaanum> Damaumsnp< Hmwoom paw Hmowwoaogohmm ou moocowuomxm cowumnwawuow> HmonKSm Hudxom-coz mo awnmcowumHom mm manna 233 Table 36 Relationship of Sexual Victimization Experiences to Psychological and Social Adjustment Variables for Men and Women. (Group means and results of ANOVA.) Dependent Main Main Variables Effect- Effect- Inter- Relationship Sexual Relationship Zihh an use an ~..n 6A.. no.6 an.n no.. on.¢ massage nanaou ueaem QG Noo. a: DO.N nc.n oc.n on.n an.n oo.n caducewnauam Hesse» as one. oco. nn.n Ho.» ss.n no.n ns.n oo.n noduoemnuuam canneoauoaom ac ooo. M“ Na. No. en. no. as. no. Hmoiumn .c “as. am. no. on. on. an. an. as. A.A-Hmn n6 ooo. ac HO.H 0c. No. oo.H 00. um. HemiHmn «a has. .a on. an. as. 6.. SN. RN. nona-Hmm._ ad 000. .6 DO.H 50. 0c. :9. on. an. nomiHmn n6 ooo. 000. on. as. we. on.d no. on. Xd<|Hmn we ooo. mm so.a an. Hm. an.“ no. «A. moo-Hmn an woo. 0H0. so. as. oo. nfi.H On.u sh. UGHIHmn 6: n36. mm ow.” No.“ so. an.“ oo.” no.“ u o-Hmn n6 nHO. noo. 0N. nN. «N. am. 0n. Nn. udfiémlen , 3.3 uue mmouo blew eu e>em eueuevqx v.32 eu e>em eueuevo: v.3: e931. ue> iueucH nuoemum nuoewmu unevneaea can! can: mkouchm ho mozoz< mo madame“ one name: mmsouov .coaoz can so: you moanmaum> ucmaunancd Hmfioom can Hmowmoaono>nm ou neocowuomxm coaumNHawuow> omsnc Hmowmoaonommm Ho QfinchHumaom hm GHQMB 235 across the three types of DA, it seems clear that all three types of DA victimization experiences are associated with psychological symptoms across a wide range of areas covered by the BSI scales. In the case of PsyA victimization experiences, the ANOVAs conducted on the three groups defined according to this variable indicated that these experiences were associated with increased scores on eyegy BSI scale. Victimization experiences with NsPA were also shown to be associated with a large number of the BSI scales (seven in all -- those mentioned above, plus the Obsessive Compulsive Scale). SA victimization experiences were associated with elevated scores on five of the BSI scales. However, given the smaller number of respondents who reported SA victimization, the lower number of significant main effects found for victimization-group membership here could be due to the smaller number of subjects. What is clear from these results is that college students who are experiencing dating abuse in their intimate relationships also report experiencing a wide range of symptoms of psychological distress. Of course, a correlational study such as this cannot unravel the direction of causality in these relationships. Also included in the questionnaire were three scales that measured different aspects of social adjustment: The Relationship Satisfaction Scale (which measured the respondents inclination and ability to develop close relationships with others) and Sexual Satisfaction Scale 236 (Ross, Dinero, Seibel & Cox; 1988), and the Relationship Quality Scale (which measures the quality of the current dating relationship). Tables 35 through 37 include results pertaining to these scales also. Again, increased levels of PsyA victimization were found to be associated with decreased scores on all three scales. With NsPA experiences, respondents who reported increased levels of victimization also reported decreased Relationship Satisfaction and decreased Relationship Quality. Not surprisingly, SA victimization experiences were found to be associated with decreased Sexual Satisfaction, although they were not associated with the other two scales. The psychological symptoms reported by those who inflicted DA were also examined. A series of ANOVAs corresponding to those described above was conducted on groups defined on the basis of their reported experiences with inflicting the three types of DA. Tables 38 through 40 display the results of these analyses. The results were similar to those pertaining to the victimization experiences, but with some notable exceptions. As with the PsyA victim groups, there were main effects found for PsyA offender-group membership on every BSI and social adjustment scale. But, for the NsPA groups, inflicting abuse against a dating partner was found to be significantly related to more of the BSI scales than were victimization experiences. Offender-group membership was associated with increased psychological symptoms on all the BSI scales, as in the case of PsyA. However, reporting 237 Table 38 Relationship of Non-sexual Physical Abuse Offender Experiences to Psychological and Social Adjustment Variables for Men and Women. (Groups Means and results of ANOVA.) M - exua Ph ice A ff Gr - Mean 9 0 M E M M E M SIGNIFICANCE OF EFFECTS Depen- Main Main In- dent Effect- Effect- terse- Var's. None Mild Severe None Mild Severe Sex Group tions I 851 Samar .36 .76 .55 .26 .31 .33 .006 .000 ns BSI ' o_c 1.03 1.50 1.36 1.02 1.28 1.18 ns .003 ns BSI Int .88 1.27 1.25 .72 1.06 .75 .016 .005 ns F. Dep .80 1.25 1.16 .87 1.07 .75 ns .013 ns Ins: ‘ Anx .93 1.28 1.31 .63 .96 .70 .000 .002 ns BSI Nos .66 .85 1.25 .56 .95 1.17 ns .000 ns BSI Phob .29 .69 .62 .20 .63 .26 ns .007 ns BSI Par .65 1.10 1.05 .50 .85 .93 ns .000 ns BSI Psy .56 .86 .72 .58 .85 .56 ns .027 ns 8“ I GSI .68 1.06 1.00 .59 .86 .76 .037 .000 ns rt Rel'shp Satis. 3.79 3.66 3.67 3.68 3.56 3.61 .006 ns ns flmml Satis. 3.57 3.61 3.26 3.36 3.09 3.21 . 068 ns ns Rel'shp Dual. 3.87 3.59 3.86 6.16 3.67 3.78 ns ns ns 238 Table 39 Relationship of Sexual Abuse Offender Experiences to Psychological and Social Adjustment Variables for Men and Women. (Group means and results of ANOVA.) Dependent Main Main Variables Effect- Effect- Inter- Relationship Sexual Relationship 2239 mo «oo. o: so.m oo.m no.3 om.m mo.m mo.o >o_omoo a_;mco_om.oo mo omo. mo oo.m No.m oo.m om.m Ps.m Fo.m co_oomow_omm Emsxom mo MNo. moo. Pm.m o6.m oo.» om.m m~.m mo.m co_ooocm_omm a_cwco_umomm mo ooo. omo. Po. mm. mm. oo.F ma. om. _oo-1mo mo ooo. no on. om. mm. to. oo. no. >mo-omo mo ooo. o: oo.. Am. oo. oo.F Fo. no. roo-_mo mo ooo. m: an. ~N. me. Pm. mm. mm. oooo-_mo mo ooo. mo oo.F Am. no. oF.F on. me. mo:-_mo mo ooo. ooo. oo. No. 66. mo.. Fo. .o. xc<-_oo mo Noo. no oo.. so. oo. oo.? oo. oo. ooo-_mo mo ooo. oFo. oo.. so. so. o~._ oo. oh. oc_-_mo mo ooo. no mo.. Po. ~o. "6.3 No.3 no. ouo-_mo mo moo. ooo. on. NN. oo. oo. oo. ~m. omsoo._mo mco_uumcouc_ cacao xom oeo>om ouncepoz p._: oce>om ouocoooz od_z modnm_cm> -ouotoo -oomcow ocoocoooo c_az c_mz mooollm no oozma A.<>oz< mo muadwou one name: msouov .co803 use 5oz mom moanmwum> uCoEumSmo¢ Hmwoom can Hmowmoaocoxmm op moocowuomxm “obsommo omsn< HmowwOHOSommm mo awnmcofiumaom 0: dance 240 offender experiences was not associated with any of the social adjustment scales. This is again, unlike the results with the victimization experiences, which were found to be related to both of the Relationship scales. As in the case of the SA victim groups, SA offender-group membership was found to be significantly related to fewer of the BSI scales. Those who reported inflicting SA against a dating partner were found to report more symptoms of hostility and paranoid ideation, and to have higher Global Symptom Inventory scores. Inflicting SA, as with NsPA, was not found to be significantly related to any of the measures of social adjustment; and most notably, not with the SexualSatisfaction Scale scores which were found to be associated with SA victimization experiences. One last observation concerning the results of the analyses presented in Tables 35 through 40, is that women and men were found to have nearly identical patterns of relationship between DA experiences and measures of psychological and social adjustment. While there were a number of main effects for sex found, which indicated that, on the whole women tended to report more symptoms on several of the BSI scales, and to score higher on the Relationship Satisfaction Scale, there was only one significant interaction effect indicating different patterns of response to DA experiences: Men who reported SA victimization experiences had high scores on the Hostility Scale, which was not true for women who reported these experiences. 241 Assessing the Impact of Attitudes and Patterns of Dominance. In exploring the relevance of the feminist/social control model to the understanding of the experience of dating aggression in a college student population, this study examined the relationships of several attitude, resource, and power variables to the DA variables. This section will present the results pertinent to the hypotheses and research questions in this area. Results pertaining to the following variables will be presented: sex-role stereotypes, attitudes towards violence and aggression in relationships, resource distributions, and patterns of power distribution and dominance in decision-making. Sex-role stereotyping and dating aggression. The following hypotheses, relevant to the feminst/social control model's predictions about dating aggression, were tested: Hypothesis 2: For males, it was expected there would be a small but significant association between sex-role stereotypes and inflicting all forms of DA, such that the more traditional were his stereotypes, the more experiences with inflicting DA he would report. For females, it was not expected that these would be associated. Hypothesis 2 was tested by computing the correlations between the stereotype scores and the three different measures of DA for men and women separately. The results of these analyses indicated that respondents' sex-role stereotyping attitudes were pg; related to their rates of inflicting either NsPA nor SA. This was true for both men and women. Neither were they related to the Total Scale Score on the Inventory of Psychological Abuse (IPA). However for both men and women, 242 rates of inflicting PsyA on one of the subscales of the IPA were significantly correlated with sex-role attitudes, although the correlations were very small. For men, Sex-role attitudes were correlated .198 (p = .034) with scores on the Criticism and Ridicule Subscale. For women, there were small correlations between sex-role attitudes and the Withdrawal scale (r = .135; p = .043). The correlations are so small as to be fairly meaningless, however. The results for women are consistent with the expectations put forth in this hypothesis. However, the predicted relationship between men's sex-role atttitudes and their rates of inflicting the three forms of dating aggression, was not substantiated. Hypothesis 2a: For males and females, the magnitude of the sex-role stereotypes discrepancy score (regardless of the direction of the difference) was expected to be associated with higher rates of inflicting all three forms of DA. Rates were expected to be highest, though, for the discrepancy scores which reflected that the male‘s stereotypes were more traditional than the female's. To test this hypothesis, two different categorical variables were created: one on the basis of the Sex Role Difference Score (er1 Difl) and one on the basis of the absolute values of these scores (er1 Dif2). In both cases, the sample was divided into three groups. For the erl Difl Groups, one group included the one-third of the respondents whose Sex Role Difference Scores were closest to zero (which would indicate exact agreement between the partners, from the respondent's perspective). Those individuals who fell outside of this area on either side were included in the other two groups. The group of respondents with more extreme 243 negative scores described relationships in which the respondent maintained more traditional sex role attitudes than did their partner. Those with positive score described relationships in which their partner maintained more traditional sex role attitudes. For the erl Dif2 groups, the three groups were of approximately equal size and consisted of respondents with the lowest, moderate and highest absolute value discrepancy scores. A series of ANOVAs were conducted to compare the groups, controlling for sex as well, on their rates of inflicting each of the three types of DA. The results of these analyses appear in Tables 41 and 42. The results of the analyses examining the rates of inflicting NsPA and SA provide some support for the first prediction of Hypothesis 2a. Table 42 indicates that, in both cases, increasing discrepancies between the two dating partners' sex role attitudes — regardless of the direction of the difference - were associated with increased levels of inflicted abuse, although the findings relevant to SA were only marginally significant. The rates of inflicting PsyA also increased with increased discrepancies, but these differences were not significant. Hypothesis 23 also predicted, however, that the groups which described relationships wherein the male is more traditional than the female would report the highest levels of violence. The group means on the NsPA scale in Table 41 are consistent with this prediction: The highest rates of Table 41 244 A Comparison of the Mean Rates of Inflicted Dating Aggression Across Groups Defined on the Basis of Sex Role Difference #1. W O M E N H E N Significance Resp’nt Partner Resp'nt Partner Main Main Inter More Prtnrs More More Prtnrs More Efct Efct ac DA Type Trad'l Eggal Trad’l Trad'l Equal Trad'l Sex Grp tion NsPA* 6.68 2.76 10.68 23.06 8.15 13.95 .086 .320 .665 SA .03 .06 .15 .22 .26 .50 .001 .066 .555 PsyA 1.29 1.32 1.65 1.36 1.62 1.38 .613 .032 .266 N - 37 N - 53 N - 65 N - 32 N - 36 N - 20 * NsPA scores are based upon the scores on The New Violence Scale. Table 42 A Comparison of the Mean Rates of Inflicted Dating Aggression Across Groups Defined on the Basis of Sex Role Difference #2 (Absolute Values). W O M B N H B N 81 ificance ‘ Hoder- Moder- Main Main Inter Prtnrs ate Extreme Prtnrs ate Extreme Efct Efct ac DA Type Eggal Diffs. Diffs. Eggal Diffs. Diffs. Se! Grp tion NaPA* 2.76 6.82 12.87 8.15 2,61 33.00 .123 .012 .210 SA .06 .06 .20 .26 .22 .61 .006 .055 .972 Pa A 1.32 1.36 1.66 1.62 1.31 1.60 .880 .092 .235 g N - 53 N - 57 N - 65 N - 36 N - 23 N - 29 * NsPA scores are based upon the scores on The New Violence Scale. 245 inflicting NsPA were reported by the women who described their partners as more traditional, and by the men who described themselves as more traditional. In both cases, the reported abuse scores were close to twice that of the other two groups. However, these results were not statistically significant. The results displayed in Table 41 pertinent to SA and PsyA were not consistent with the hypothesis, though. For rates of inflicted SA, in all groups, the men's scores were higher than the women's. But, contrary to predictions, for papa sexes, the rates of inflicted SA were highest among those respondents who described their partners as more traditional in their sex role orientation than they were themselves. This may be due to related differences between the dating partners in their attitudes toward and willingness to engage in sexual activity. In other words, it would not be surprising if partners who were described as more discrepant in their sex role attitudes were also more discrepant in their attitudes toward sexuality, which may be one factor which increases the likelihood of coercive sexual interactions. However, the data from this study did not allow for an examination of that possibility. In regard to the results in Table 41 pertinent to PsyA, the statistically significant effect for the group membership based upon er1 Diffl was also inconsistent with the predictions. Here again, independent of sex, the rate of inflicting PsyA was highest for the group who described their 246 partner as more traditional than themselves (Y = 1.44). The other two groups - those described as equal and those where the respondent described themselves as more traditional — had means of 1.36 and 1.31 respectively. Thus, these analyses did indicate that increasing sex role discrepancies were associated with higher rates of inflicting NsPA, and, to a lesser extent, SA. However, there was little evidence that those dating relationships in which the males had more traditional sex role attitudes than their partners were more abusive than when the discrepancies were in the opposite direction. Acceptance of aqggaaaion and adversarial beliefs. The following hypotheses were put forth concerning the relationship between experiences of DA and the acceptance of aggression and adversarial beliefs about the intimate relationships between men and women Hypothesis 3: Men's level of acceptance of violence in interpersonal relationships and of an adversarial/aggressive perspective about male-female relationships in general, were expected to be be correlated with higher rates of inflicting all three forms of DA. Hypothesis 3a: Women's scores on these attitude measures were expected to be unrelated to their rates of inflicting violence at low levels of involvement in their relationships, but positively associated at high levels of relationship involvement. Hypotheses 3 and 3a were first tested by means of bivariate correlations between the attitude measures and measures of inflicting the three types of DA. These were, computed for men and women separately. The Acceptance of 247 Violence Scores were not found to be significantly correlated with any of the measures of inflicting DA among either the men or women respondents. For men, The Adversarial Beliefs Scale was found to be significantly correlated with the measures of inflicting PsyA; both with the total scale score (r = .314, p = .002), and with all of the sub-scales except Withdrawal. These correlations ranged from .238 for Threats and Violence to .309 for Control and Social Isolation (p < .01). This attitude scale was not found to be related to measures of inflicting NsPA or SA among men. Among the women, a very similar pattern of correlations between this attitude scale and measures of DA emerged. It was found to correlate with The Total Score of the Inventory of Psychological Abuse (r = .230, p = .002); and also with all of the subscales except one. For women, only the Threats and Violence Scale was ppt_related to Adversarial Beliefs. The correlations between this scale and the other IPA sub- scales ranged from .160 (Control and Social Isolation; p = .022) to .254 for The Withdrawal Scale (p = .001). In all cases (except the Withdrawal Scale) the observed relationships were less strong for the women's data than for the men's. To further explore the differential prediction for men and women put forth in Hypothesis 3a, a series of ANOVAs were used to explore the impact of the level of involvement on the relationship between attitudes and inflicting dating abuse. 248 Level of involvement, gender, and level of attitude acceptance (low, medium, high) served as the independent variables (factors) in these analyses, with the rates of the three types of inflicted DA serving as the dependent variables. In regard to all three types of DA, a significant main effect for Level of Involvement was observed, indicating that the rates of inflicting all of these increased as respondents reported deeper levels of involvement. For all three types of DA, these effects seemed to result from the fact that the group indicating the highest level of involvement reported much higher levels of inflicting DA than either of the other two groups. No main effects or interaction effects involving The Acceptance of Violence Scale resulted from these analyses for inflicting either SA or PsyA. However, a significant interaction effect for Level of Involvement X Acceptance of Violence (F = 2.59; p = .038) did result in the analysis focused on experiences with inflicting NsPA (as measured by the New Violence Scale). Figure 4 depicts these results. For men and women combined, the group that described themselves as most deeply involved and who most strongly endorsed attitudes of acceptance toward violence (N=13) reported extremely high rates of inflicted violence in comparison with all other groups. In regard to the results concerning the relationship between the Adversarial Beliefs Scale and inflicting DA, no 249 50 (N=13) High Involvement 40% 30‘ < a in z 20‘ (N233) N227 10.(N=‘7)L\ ( ) N331 N=18 N=27 ( ) \) ( ) 1 Moderate Involvement N:E§) (N=17)aa Low MVdvmneni O x v i 1 LOW MED|UM HIGH ACCEPTANCE OF VIOLENCE Figure 4 Rates of Inflicting Non-Sexual Physical Abuse Among Students With Differing Levels of Relationship Involvement and Differing Levels of Acceptance of Violence in Intimate Relationships. (NsPA measured by New Violence Scale). 250 main effects or interaction effects involving this scale resulted in the analyses focused on NsPa or SA experiences. However, consistent with the correlations computed, a main effect for adversarial beliefs did occur as a result of the ANOVA pertaining to the PsyA Total Scale (F = 4.50; p = .012). Similar main effects occurred in the analyses conducted on the IPA subscales, Criticism and Ridicule (F = 5.13; p = .007), and Withdrawal (F = 4.61; p = .011). A similar, marginally significant effect occurred in the analysis focused on the subscale Control and Social Isolation (F = 2.99; p = .053). In all cases, rates of inflicting PsyA increased with increasing endorsement of Adversarial Beliefs about relationships between men and women. These results provide minimal support for the predictions put forth in Hypotheses 3 and 3a. Acceptance of Violence was not found to be related to any measures of inflicting DA for men or women. The proposed interaction between these attitude scales and level of relationship involvement was only found in the case of NsPA and the Acceptance of Violence Scale, and no significant sex differences were observed. The Adversarial Beliefs Scale was found to be related to inflicting PsyA as predicted, although not to NsPA or PsyA. And, in the case of PsyA, it was significantly associated with both men's and women's reports of inflicting that type of DA, contrary to expectations. Resources and dominance in decision-making. As discussed in previous chapters, the social control model predicts that 251 a dating couple's distribution of resources and the extent to which the male endorses the goal of male dominance are expected to be important determinants of DA. These predictions are expressed in the following hypotheses: Hypothesis 4a: Males who expressed dominance goals for male-female relationships and who reported having fewer resources than their partners were expected to be more likely to report inflicting all three forms of DA. This hypothesis was tested with a series of ANOVAs with Resource Balance Groups (see Table 12 for description) and a categorical variable (Rel'shp Pwr Goals) which defined three groups based upon The Total Ideal Power Scale score from the decision—making questionnaire as the independent variables, or factors, and the measures of inflicting the three forms of DA as the dependent variables. The three Rel'shp Pwr Goals groups were defined as follows. A group describing a goal of relative submission (Submission Goal Group) consists of those individuals whose total Ideal Power Scale scores were less than 3 (the score indicating equality). This was the smallest group, with only 16 members. The Equality Goal Group includes 158 respondents who described goals of exact equality. The remaining 87 respondents, whose scores were greater than 3, indicating they desired more influence than their partners in at least one area of decision-making in their relationship, comprised The Dominance Group. The analyses pertaining to NsPA did not result in any significant main or interaction effects. With rates of inflicting SA against a dating partner as the dependent 252 variable, the ANOVA resulted in significant main effects for sex as well as a number of significant interaction effects. Table 43 and Figure 5 present the results of this analysis. Figure 5 clearly indicates men's higher rates of inflicting SA. But women's rates of inflicting are highest for those who have goals of equality, regardless of the resource balance between the partners. Women who report wanting dominance and who are in relationships with partners who have more resources than they do are also more likely to report inflicting SA. For men, the more resources they have relative to their partner, the more likely they are to report inflicting SA, regardless of what their dominance goals are. The highest rates are reported by those men who report either dominance or submission goals (although the means of the group indicating goals of submission should be considered cautiously, given that there are so few subjects in this group). The men in this sample who reported having fewer resources than their partners (n = 14) did not inflict apy SA against dating partners. The ANOVA performed with rates of inflicting PsyA as the dependent variable resulted in a significant interaction effect for Resource Balance Group X Sex (F = 5.87; p = .003). As with the SA results, for men, rates of inflicting PsyA increased as the amount resources, relative to their partners, increased. In contrast, for women, rates of inflicting PsyA were highest when they reported having fewer Table 43 Results of Analysis of Variance - Inflicted Sexual Abuse BY 253 Resource Balance, Decision Making Power Ideals, and, Sex. Source of Variation Main Effects Resource Dal. Power Ideals Sex Z-way Interactions Res.8al X Pwr Idls Res Bal x Sex Pwr Idls X Sex 3-way Interaction ResBlXPerleSex Explained Residual Total Sun of Sean! 3.656 .569 .605 1.765 6.976 1.829 1.326 3.039 2.531 2.531 10.960 69.669 60.609 DF NNUI 58‘NNfiO-P 220 237 .226 .256 F 3.063 1.261 .898 7.820 2.755 2.026 2.938 6.736 2.803 2.803 2.857 256 2.5 2.0 ~ % 3 1.5 - m < .J < D g; 1.0 ~ (I) —- MEN-SUBMISSION IDEALS MEN-DOMINANCE IDEALS (15* /// MEN-EQUALHY DEALS / WOMEN-EQUALITY IDEALS __ WOMEN-DOMINANCE IDEALS a . PARTNER EQUAL RESPONDENT MORE MORE RESOURCE BALANCE Figure 5 Mean Rate of Inflicting Sexual Abuse for Groups Defined by Resource Balance, Power Ideals and Sex 255 resources than their partners and decreased when they had equal or more resources. These results were not consistent with the hypothesis. Men who reported having fewer resources than their partners were rare, but those who did, reported very low levels of inflicting SA and PsyA against their partners. Contrary to the predictions, rates of inflicting both of these types of DA increased as men reported having equal or more resources than their partners. Dominance goals were related to inflicting SA (in an interaction effect), but were not related to inflicting either PsyA or NsPA. Hypothesis 4b: Women who reported having equal or more resources than their dating partners and who described them as having expectations of dominance were expected to report sustaining more of all three forms of DA. This was also tested with a series of ANOVAs, conducted on the women respondents' responses. The victimization rates of the three different types of DA served as dependent variables, and Resource Balance Groups and Partner's Dominance Goals (coded as submissive, equal, or dominant) were factors. Across all three types of DA, the mean scores for sustaining abuse increased as women rated their partners as desiring more dominance, but this effect reached significance with only the SA (F = 5.977; p = .003) and PsyA (F = 11.549; p = .000) dependent variables. The resource balance between the partners was not significantly related to experiences of any type of DA victimization - either in a main effect or in interaction effects. Although not 256 significant, it is interesting to note that, across all three forms of DA, the rates of victimization were highest when women reported having fewer resources than their partners, not equal or more resources, as predicted by this hypothesis. Hypothesis 4c: For both women and men who reported relationships at moderate and high levels of commitment, their experiences with all three forms of DA were expected to be associated with more male dominant actual power distributions in their relationships. To test this hypothesis 3 series of ANOVAs were conducted with the actual relationship power distribution variable (The Total Actual Power Scale from the decision-making questionnaire) as the dependent variable and the level of DA experienced, sex, and level of relationship commitment as the factors. For this analysis, a variable defining overall involvement in dating violence was created, combining information about victimization from NsPA, SA and PsyA experiences. Low, moderate and severe groups were defined. The Low Victimization Group included those who reported having no experiences with either NsPA or SA victimization and also reported low levels of PsyA victimization. The Moderate Victimization Group included those in the lowest groups on all but one or two of the three areas with a maximum of moderate experiences on those. Also included are those reporting high levels of PsyA victimization with no NsPA or SA experiences. The Severe Victimization Group included respondents who were in the severe NsPA or SA groups regardless of experiences with other forms of abuse. Also included are those with moderate levels of all three forms of 257 abuse and those with high PsyA and moderate experiences with NsPA or SA victimization. The results of this analysis are presented in Table 44 and Figures 6 and 7. There were two significant interaction effects: Victimization X Sex (Figure 6), and Level of Involvement X Sex (Figure 7). Figure 6 indicates that, at low levels of involvement, women reported having less influence over all the decisions made and men more. This reversed at moderate levels of involvement. At high levels of involvement men reported having slightly more influence over decisions than did women, but the mean scores were very similar. Figure 7 displays the interaction effect which bears directly upon this hypothesis and the results are as predicted. Women reported continually decreasing amounts of influence over decisions as the level of victimization increased. In contrast, while men also reported decreased influence at moderate levels of victimization, they described increased influence over decisions at the most severe level of victimization. Contrary to the prediction, this relationship exists independent of level of relationship involvement. In all dating relationships, moderate levels of DA victimization are associated with decreased influence over decisions for both sexes. However, severe levels of involvement with DA are associated with men having more power and women less -- in other words, with male-dominant power structures. This was true, even though the measure of DA 258 Table 44 Results of Analysis of Variance - Total Actual Amount of Power Over Relationship Decisions BY Combined DA Victimization Group, Level of Relationship Involvement and Sex. Sum of Mean §pugcg 9; Variation Sgpagea 22 §gpa£g F Sippif. Main Effects .379 5 .076 1.365 .266 Victim Grps .285 2 .163 2.529 .082 Invlvunt .087 2 .063 .771 .666 Sex .026 1 .026 .656 .500 2-way Interactions .882 8 .110 1.956 .053 Vctim X Invlvmnt .216 6 .056 .957 .632 Vctm X Sex .516 2 .258 6.579 .011 Invlvmnt x Sex .366 2 .173 3.073 .068 3-way Interaction .523 6 .131 2.321 .058 VcthInvleSex .523 6 .131 2.321 .058 Explained 1.786 17 .105 1.862 .022 Residual 12.627 226 .056 Total 16.612 261 .050 259 3.2 3.1 - 3.0 ., MEN WOMEN ACTUAL POWER 2.9 ‘ 2 .8 I I LOW MEDIUM HIGH RELATIONSHIP INVOLVEMENT Figure 6 Mean Actual Relationship Power Scores for Groups Defined by Level of Relationship Involvement and Sex 3.2 3.1 ~ 0: ‘i O MEN 0. _, 3.0 - < D p— 0 < 2.9 - WOMEN 2.8 I I 1 LOW MODERATE SEVERE VICTIMIZATION GROUP Figure 7 Mean Actual Relationship Power Scores for Groups Defined by Level of DA Victimization and Sex 260 involvement was based upon victimization rates for both sexes. Hypothesis 4d: Among both women and men, reports of experiences with DA were expected to be associated with greater discrepancies between the amount of influence they actually experienced and their reported ideals (i.e. they would report having less power than they wanted). This hypothesis was tested by means of a series of ANOVAs with the Total Difference Score from the Decision Making Questionnaire as the dependent variable and gender and level of DA victimization as factors. Table 45 and Figure 8 present the results of the analysis which utilized the victimization groups defined on the basis of the three types of DA experiences combined. Figure 8 reveals that, the discrepancy between the amount of influence which the respondents experienced and the amount they rated as ideal increased as the severity of victimization increased. At the highest level of victimization, women's discrepancy scores are higher than mens, though Table 45 indicates that this interaction effect did not achieve significance. The main effect for victimization level, however, was significant. To further explore the impact of victimization experiences on the power discrepancy scores, analyses were conducted utilizing victimization groups based upon reported experiences with each of the three types of DA separately. In all three cases, significant main effects for level of victimization were achieved. For both NsPA and PsyA experiences, women reported increasingly larger discrepancy scores as victimization experiences increased in severity. 261 Table 45 Results of Analysis of Variance - Ideal/Actual Difference Score (Total Scale) BY Victimization Group and Sex. Sum of Mean Source 0; Variation Sgpareg DE §gpaga E §$5p;£. Main Effects .731 3 .266 2.630 .051 Victim Groups .663 2 .321 3.667 .033 Sex .113 l .113 1.222 .270 2-wey Interactions .319 2 .159 1.720 .181 Victim Grp X Sex .319 2 .159 1.720 .181 Explained 1.050 5 .210 2.266 .069 Residual 22.059 238 .093 Total 23.109 263 .095 262 0.25 WOMEN 0.20 - / 0.15“ MEN 0.10“ TOTAL DIFFERENCE SCORE 0.05“ 0.00 v . I MILD MODERATE SEVERE VICTIMIZATION GROUP Figure 8 Mean Total Difference Scores for Male & Female Victim Groups 263 In contrast, men’s scores did not increase between moderate and severe levels of victimization, and, in fact, in both cases, they decreased slightly. The overall result was that, at the most severe levels of victimization, women reported greater dissatisfaction with the power they have in their dating relationships than do men, although the interaction effect did not reach significance in either case (PsyA, p = .339; NsPA, p = .052). This trend, though not significant, is consistent with the results presented above indicating that, at severe levels of abuse, women experienced decreased influence over decisions in a relationship, while men experienced increased influence. Hypothesis 4e: Among those who reported experiences with inflicting DA, the males were expected to report male dominance ideals, while the women were expected to report ideals of equality in decision making. To test this hypothesis respondents were designated as aggressive or non-aggressive on each of the three DA measures (i.e. dichotomous variables were constructed). Then a series of ANOVAs were conducted with the Total Ideal Power Scale from the Decision Making Questionnaire as the dependent variable, and sex and involvement in inflicting the three types of DA as factors. The group means resulting from these analyses are depicted in Figures 9, 10 and 11. On the whole, both men and women overwhelmingly espoused equality ideals for dating relationships (an ideal of perfect equality in dating decision—making would be reflected in a score of 3). However, significant main effects for sex resulted from the 264 33 m 3.2 - (N226) LAJ ; M94 0 CL 0 LiJ 0_C If) LIJ O 3.1“ (N261) (N=49) (Nzl9EL_.——,———~———r~*““"’"’+ WOMEN 10 . , NO YES Figure 9 Mean Scores for Desired Power in Decision-Making (Ideal Power - Total Power Score) for Students Who Do and Do Not Inflict Abuse: Non-Sexual Physical Abuse - Inflicted 265 33 (N217) MEN 32“ cr LLJ 3 O (L C) LLJ g; If) 8 3.1 “ (N270) (N=152) (N211) WOMEN 3.0 . . NO YES Figure 10 Mean Scores for Desired Power in Decision-Making (Ideal Power - Total Scale Power) for Students Who Do and Do Not Inflict Abuse: Sexual Abuse - Inflicted 266 3.3 3.7 “ (N256) mm 3.1 “ N229 ( ) (N=109) WOMEN (N252)+#fiflflyfl,,lfl—wfl*’”””'flflflfifl+ 3.0 I T LOW MODERATE & SEVERE Figure 11 Mean Scores for Desired Power in Decision-Making (Ideal Power - Total Scale Score) for Students Who Do and Do Not Inflict Abuse: Psychological Abusa - Inflicted 267 analyses of both NsPA and PsyA groups (for NsPa, F = 5.77, p = .017; for PsyA, F = 5.326, p = .022). In the case of SA, the main effect for sex did not achieve significance (p = .058), but there was a significant main effect for being an offender (F = 5.406, p = .021), and a significant interaction effect for offender group X sex. Figure 10 indicates that sexually abusive men have ideals of male dominance, while all women (and non-abusive men) are more likely to espouse ideals of equality. This same pattern - abusive men expressing ideals of male dominance, while all other respondents expressed ideals of equality - was apparent in the case of NsPA and PsyA also, but the interaction effect did not achieve significance (for NsPA, F = 2.28, p = .132; for PsyA, F = .862, p = .354). However, given the fact that a significant main effect for sex occurred in both of those analyses and that the sex differences were most extreme between the men and women who reported inflicting DA against their partners, it can be assumed that the abusive groups' scores are significantly different. These results are consistent with this hypothesis and the null hypothesis can be rejected. Assessing the Over All Impact of Attitudes. Resources and Dominance. The results of the discriminant function analyses presented earlier provide a test of the following research questions concerning the relationship between DA experiences 268 and the attitude, resource and dominance variables central to the feminist/social control model of dating aggression. Relavance of thayfaminist/social control madel to understanding men's abusive bahaviors. Respondents' scores on the variables (specified below) which are predicted by the feminist/social control model to foster male violence and abuse against women were expected to discriminate well between those males who inflicted SA, NsPA, and PsyA and those who did not. The variables which the social control model predicts will be related to higher levels of all three forms of male aggression toward dating partners are as follows’: (a) lower absolute levels of resources (Personal and Social Resources; Economic Resources), (b) fewer resources in comparison with partner's (Diffs. in Personal & Social Resources, Diffs in Economic Resources), (c) endorsement of the ideal of male dominance in relationship decision making (Ideal Power - Independent Decisions, Ideal Power - Partner's Decisions, Ideal Power - Relationship Decisions), (d) higher levels of reported actual decision-making power in one's relationship (Actual Power - Independent Decisions, Actual Power - Partner's Decisions, Actual Power - Relationship Decisions), (e) reportedly having less decision-making power in one's relationship than one's ideal (Difference Score - Independent Decisions, Difference Score - Partner's 7 The names of the specific measures used in this study to assess each variable are enclosed in parentheses. 269 Decisions, Difference Score - Relationship Decisions), (f) maintaining traditional sex-role stereotypes (Sex Role Attitude Scale), (9) reported discrepancies between one's own and one's partner's sex-role stereotypes (Sex Role Differences #1, Sex Role Differences #2), (h) endorsement of attitudes characterizing male/female relationships as manipulative and antagonistic (Adversarial Beliefs), and (i) acceptance of violence toward women (Acceptance of Violence). Overall, the results of the discriminant function analyses which were computed on the men's data, presented previously in Tables 28, 30 and 32, indicated that these variables, in combination g;g_discriminate significantly between abusive and non-abusive men in the case of all three different types of DA. They accounted for from 35.6% to 67.8% of the observed variance in the discriminant function scores between the offender groups and, in all cases, they permitted a better classification of cases into the correct offender groups than would have been possible by chance (although the performance was not a lot better than chance in the case of both SA and NsPA). Another way of evaluating the overall accuracy of the feminist/social control model's specific predictions as they related to the DA experiences of the men in this study is to compare the number of specific predictions made to the number of variables which did in fact statistically discriminate between the different offender groups (either in the One-way ANOVA results presented in Table 26, or by entering the first 270 discriminant function for one of the types of DA). Specific predictions concerning 18 different variables were put forth above. Of these 18 variables, 11 were found to significantly discriminate (or be significantly different) between offender groups in at least one of the three types of DA. Of these 11, eight were in the predicted directions, while three were in the opposite direction of that predicted. These three were: Differences in Personal & Social Resources, Differences in Economic Resources, and the Ideal/Actual Power Difference Score pertaining to the dating partner's decisions. In the case of Personal and Social Resources, for all three types of DA, increased levels of abusiveness were associated with reports that the men had mppa resources than their partners while the non-abusive men were more likely to report equal or less resources than their partners. In the case with Differences in Economic Resources, in general, across all groups, men reported having more economic resources than their partners. Furthermore, in the case of both NsPA and PsyA groups, contrary to what was predicted, the most abusive men reported a greater differential between themselves and their partners than did the less abusive or non-abusive men. Although these differences were not significant in the One-way ANOVAs, this variable aaa discriminate between men in the different NsPA offender groups. 271 While the Ideal/Actual Power Difference Score pertaining to the dating partner's decisions turned out to discriminate between abusive and non-abusive men in the opposite direction of that predicted, the results are actually consistent with the feminist/social control model's perspective anyway. The prediction was that, with increased levels of reported abusiveness, men would express increasing dissatisfaction (reflected in greater difference scores) with the amount of influence and power they experienced in all domains in their relationship. In fact, the difference scores pertaining to influence over the decisions normally considered to be the partner's personal domain were lgya; for the men in the most severe NsPA and PsyA offender groups and among those who reported inflicting SA against their partners. An examination of the group mean scores on the actual power variable pertaining to influence over the partner's decisions in Table 26 reveals that this is due to the fact that, with increasing abusiveness, men experience increasing levels of actual influence over this domain of decisions and so their dissatisfaction decreases. In concluding, the results regarding the relevance of the feminist/social control model to explaining men's experiences with dating aggression in this sample of college students were mixed. A number of the model's predictions were not borne out, and yet the combined group of variables predicted by the model to discriminate between aggressive and non- aggressive men did, in fact, significantly discriminate 272 between the abusive and non-abusive groups in this sample and number of the model's specific predictions were supported by the data as well. Relayance of feminist/social control_anel to understanding women's abusive bahaviora. Women respondents' scores on the following variables (which are predicted by the feminist/social control model to account for reported rates of female violence and abuse inflicted against their dating partners) were expected to discriminate well between those who reported inflicting SA, NsPA, and PsyA and those who did not”: (a) reporting that partners have fewer resources than themselves (Differences in Personal and Social Resources, Differences in Economic Resources), (b) rating partners as subscribing to more traditional sex-role stereotypes (Partner's Sex Role Attitudes), (c) larger discrepancies between their own and their partner's sex—role stereotypes (Sex Role Differences 1, Sex Role Differences 2), (d) rating partners as subscribing to more male-dominant decision making ideals (Parner's Dominance Motive), and, (e) reports of lower levels of actual decision-making power in the relationship (Actual Power - Independent Decisions, Actual Power - Partner's Decisions, Actual Power - Relationship Decisions). The results of the discriminant function analyses which were computed on the women's data were presented previously 9 The specific measures which assessed each variable in this study are listed in parentheses. 273 in Tables 27, 29 and 31. As with the analyses focusing on men's abusive behaviors, the discriminant functions indicated that these variables, in combination g;g_discriminate significantly between those women who report inflicting abuse against a dating partner in the case of all three different types of DA. However, the resulting functions did not perform as well as those that pertain to the men's data. For the women, they only accounted for from 11.1% to 19.0% of the observed variance in the discriminant function scores between the offender groups. In the case of SA the resulting discriminant function does no better than chance in correctly classifying cases to offender groups, and with NsPA offender groups, the functions perform poorly as well. As with the men's data, it is possible to evaluate the accuracy of the feminist/social control model's specific predictions about the relationships between dating abuse and particular variables by looking at the number of predicted variables which did in fact statistically discriminate between the different offender groups. For the women, the model specifies predictions concerning nine different variables. Of these 9 variables, 6 were found to significantly discriminate (or be significantly different) between offender groups in at least one of the three types of DA. Of these, two were in the opposite direction of what was predicted. The discrepancies were similar to those found between the predictions and actual results regarding the 274 men's DA experiences. Differences in Economic Resources, and the Actual Power Over the Dating Partner's Decisions. The same finding regarding the balance of resources between the dating partners emerged from the women's data: increased dating abuse (in this case NsPA) was found to be related to reports that their male dating partners had relatively gppa economic resources than the respondents, not less, as the model predicts. The feminist/social control model predicts that women who report inflicting DA against a dating partner will report lower levels of actual decision-making power in their relationships than will women without such experiences. However, in regard the discriminant function pertaining to the criterion groups based upon PsyA score, the Actual Power Scale for Influence over Partner's Decisions entered as a discriminating variable, indicating that increasing levels of abuse were associated with increased influence over the dating partner's independent decisions, rather than withdecreases in power. Increasing levels of abuse were also associated with increased power over the partner's decisions for men, as discussed above. While that finding pertaining to men is consistent with the feminist/social control model, it is not so easy to understand women's increased power over their partners' decisions from the perspective of this model. However, the comparison of the three PsyA groups' mean scores on this variable, presented in Table 25, do not reveal any significant differences as assessed by the One—way ANOVA. 275 In summary, the variables predicted by the feminist/social control model to explain women's experiences with DA do a poorer job of differentiating between women who inflict aggression against a dating partner and those who do not than they do with men. In addition, again, as with the men's data, the results pertaining to the model's specific predictions are mixed. Some of the model's predictions are supported by this data, while others are not. Gender Differences The results of some of the analyses described above demonstrated many gender differences in these college students' experiences with the three different forms of dating aggression. In addition to the findings relevant to gender differences which have already been addressed, the following hypotheses were tested: Hypothesis 5: More women were expected to sustain more serious forms of sexual aggression and physical battery than men, to experience more psychological abuse, and to experience more injuries and more negative psychological effects associated with these DA experiences than did men. In addition, women were expected to rate the incidents of NsPA victimization that they experienced as more dangerous and to indicate more frequently than men that their acts of inflicted NsPA abuse were self-defensive. This set of predictions was tested by comparing, by means of a series of T-tests, women's and men's levels of sustained DA (all three forms), sustained injuries, reported psychological effects, rated dangerousness of NsPA incidents and reasons given for inflicting NsPA. Table 16 presented information about these tests which pertained to all of the 276 dating abuse variables. There were no statistically significant differences between the experiences of men and women with regard to NsPA victimization - whether measured by the CTS or the New Violence Measure. In addition to the Total Scale scores, women did not report experiencing more serious types of NsPA. While they did rate the incidents of victimization they reported to be slightly more dangerous than did the men, this difference was not statistically significant. There also were no statistically significant differences between men and women in the reasons they gave for inflicting NsPA. Women were no more likely than men to indicate that their actions were in self-defense. There also were no sex differences on the PsyA victimization total scale, although women did report higher rates of victimization on the Emotional Callousness Scale than did men. There were significant differences, though, in the rates of SA victimization reported by men and women. Women sustained more SA as measured by the total scale, as has been consistently found in previous studies. The sex differences in regard to the emotional and physical consequences of DA victimization experiences were presented earlier. It will be recalled that the reported immediate emotional reactions to DA experiences were found to be more severe among women than men. In addition, the women reported sustaining more and more severe injuries as a result of both NsPA and SA, but the differences were not statistically significant. 277 Comparing the CTS and The New Violence Measure. As described previously, a major goal of the current study was to develop an alternative measure of NsPA which would be more sensitive than The Conflict Tactics Scale to the qualitative differences in the experience of this form of DA. The New Violence Measure included assessments of perceived dangerousness and sustained injuries in the overall violence score. It was predicted that this measure would better reflect the gender differences in the experience of NsPA in intimate relationships. Hypothesis 6: The NsPA scores which resulted from the qualitative scoring strategy, as compared with the CTS-based scoring strategy, were expected to be more highly correlated with the measures of reported emotional distress, and with scores on the depression, anxiety, interpersonal sensitivity, somatization, hostility, and global severity scales of the Brief Symptom Inventory, as well as with the measures of decision-making power distributions in relationships. The responses to the NsPA questions were scored in two ways: one which included only the count of aggressive acts that would be assessed with the Violence Scale of the CTS; and another which incorporated all of the violence items (including those which were added) and used the dangerousness and injury information as weighting variables. This hypothesis was tested by computing a series of bivariate correlations between (a) the CTS and New Violence Scale scores and, (b) the BSI Scales, Emotional Impact Scale, and Actual Decision-Making Power Scales. The results of these analyses are presented in Table 46. The table also includes information about the relationship between these 278 Table 46 Correlations of The CTS-based measure’, The New Violence Measure, Dangerousness and Injury (dichotomous) with Brief Symptom Inventory, Emotional Impact, and Actual Decision- Making Power Scales. Scales CTS-based New Violence Dangerousness Injury BSI - Somatization .205** .283** .279** .227** 0bsess.- Coupulsive .087 .097 .153" .090 Interpersonal Sensit. .038 .068 .129' .106 Depression .096 .119‘ .162* .096 Anxiety .057 .106‘ . 206“ .163* Hostility .281** .331** .196** .155* Phobic Anxiety .066 .118* .131* .113* Paranoid Ideation .211** .191** .201** .161' Psychoticism .167" .198“ .138“ .158* Gen. Severity Index .175* .197** .231** .173. Emotional lupact - NsPA .676“ .505" .761“ .570“ Decision-Making Power Total Scale -.070 -.238** -.268** -.332** Independent Dec'ns -.387** *.619** -.321** -.357** Partner's Dec'ns .321" .186"' .109* .100 Relationship Dec'ns .006 -.081 -.166* -.212* * [)‘< .05 ** p < .001 9 For both the CTS and The New Violence Measure, the scale scores indicating NsPA victimization experiences were used in computing these correlations. 279 same scales and the two weighting variables used in the New Violence Measure: the highest perceived dangerousness ratings and a dichotomous variable indicating whether or not the respondent reported any injuries as a result of NsPA victimization. As predicted, the Depression, Anxiety, Interpersonal Sensitivity, Somatization, Hostility, and Global Severity BSI scale scores were more highly correlated with the New Violence Scale than with the the CTS Scale scores. In fact, all of the BSI scale scores, with the exception of the Paranoid Ideation Scale, are more highly correlated with the New Violence Scale, although a number of these differences are very small. In all, the CTS scores were significantly correlated with five of the BSI scales, while eight of the correlations between these scales and the New Violence scores were significant. The measure of emotional distress was also more strongly associated with the New Violence Scale than with the CTS scores, as were two of the scales measuring actual decision- making power in the relationship: the Total Scale and the scale measuring influence over Independent Decisions. The scale measuring power over the Partner's Decisions was, however, more strongly associated with the CTS scores. In general, the results of these analyses were consistent with the predictions that the New Violence Measure, with its inclusion of assessments of perceived dangerousness and injury, would perform more effectively in accounting for some of the expected consequences of experiences with NsPA. 280 In addition to testing this hypothesis, the gender differences in the reports of NsPA experiences which resulted from the new measure were explored, as presented above. Contrary to expectations, the New Violence Measure did not reveal any sex differences in the experiences of inflicting or sustaining NsPA, nor did the measures of perceived dangerousness or injuries. This was true even though the responses to all of the items that were added to the CTS in creating the new measure indicated that they tapped behaviors that were more frequently sustained by women than by men. However, in only one case - physical restraint - was the difference significant (t = 4.01, p = .000). Whatever differences exist between men's and women's experiences of NsPA in dating relationships, they were not tapped any more effectively by The New Violence Measure than they were by The Conflict Tactics-Based Scale. 281 VI. Discussion The discussion below will address the results as they pertain to the following areas: (a) the development of a comprehensive view of dating abuse as the combination of sexual, non-sexual physical, and psychological abuse; (b) a comparison of the performance of the Conflict Tactic Scale to that of The New Violence Measure developed for this study; and (c) an assessment of the relevance of the feminist/social control model's predictions concerning attitudes, resource distribution and patterns of dominance to understanding dating abuse. Achieving a Comprehensive Viaw of Dating Aggression The results of this study clearly indicate that non~ sexual physical, sexual, and psychological abuse are related phenomena. Relationships in which there are experiences of one type of abuse are likely to be the arena for abusive interactions of the other two types as well. Not only were the three types of DA found to be significantly correlated with one another, but the results of the discriminant function analyses indicated that they also were similarly related to some important relationship patterns. Specifically, men's reports of inflicting these three types of abuse were similarly related to some relationship power and dominance variables, as well as to the balance of some kinds of resources between the dating partners. 282 But there were also important differences observed between the three types of DA. For instance, SA was found to be much more frequently inflicted by men against women, and when it was inflicted against men, it had much less severe emotional and physical consequences. These gender differences in rates of inflicting were not found, or were not as pronounced, in regard to NsPA or PsyA. Other differences between the three types of DA were also found. For instance, the relationship between inflicting abuse and sex role attitude discrepancy scores were found to be different for the three types of abuse. Rates of inflicting NsPA were higher in both directions of discrepancy, while rates of inflicted PsyA were highest only for the group who described their partners as more traditional than themselves, but not when they described the difference in attitudes to be in the other direction. The rates of inflicted SA also followed this pattern (i.e., were highest where the partner was described as more traditional), but the effect was not significant. These facts concerning the similarities and differences between the three types of DA suggest that it would be helpful to view them as interrelated, yet distinct, components of the overall phenomenon of abuse in dating relationships. Assuming this perspective would require the inclusion of measures of all the different forms of abuse in future investigations. In the absence of this broader perspective, our understanding of the phenomena of violence 283 and aggression in dating relationships is bound to be limited by the artificial limitations of our constructs. As Gwartney-Gibbs and her colleagues have suggested previously (Gwartney-Gibbs & Stockard, 1989, Gwartney-Gibbs, Stockard & Bohmer, 1987; Lane & Gwartney-Gibbs, 1985), it seems especially ill advised to study violence in intimate relationships without including measures of sexual abuse. The results of this and previous studies have found SA to be a widespread and common component of the abuse that occurs in intimate relationships. An examination of the percentages of men and women who reported some experience with inflicting and/or sustaining SA and NsPA in Table 20 (page 194) demonstrates how common experiences with SA in dating relationships are among this sample of college students. The percentages in this table also show how the exclusion of measures of SA experiences in studies of "dating violence" may lead to misguided conclusions concerning men's and women's relative involvement in DA experiences. When just NsPA incidents are considered, nearly equal proportions of men and women report inflicting and sustaining DA in a dating relationship. However, when the proportions of men and women who report experiences with either SA or NsPA are compared, while again nearly equal proportions report inflicting one or the other type of DA, more women report victimization experiences (42.7% of the women as compared to 32.6% of the men), though the difference is not statistically significant. Comparing the percentage of women and men who 284 report experiences with both SA and NsPA reveals sex differences, also. More men than women report inflicting both types of DA against their dating partner (14.6% as compared to 4.9%), while more women than men report sustaining both types of violence from their dating partner (14.6% of the women, and 9% of the men). However, the latter difference, again, was not statistically significant. Still, these differences and trends make it clear that, including consideration of SA experiences is extremely important to achieving a more complete and accurate contrast of men's and women's experiences with DA in their relationships. Perhaps the original conceptualization of NsPA as a "conflict tactic" has led to the unfortunate historical separation of investigations into these two aspects of intimate violence. Whatever the reason, it is important to move beyond a limited perception of the scope of aggressive, coercive behaviors which occur in intimate relationships between men and women. We can only achieve a comprehensive understanding of the common issues, patterns, and underlying dynamics that lead to all_forms of aggression in intimate relationships by looking at the "big picture". This study also defined psychological abuse as an important aspect of dating aggression. In fact, one of the study's major contributions was to begin to explore the role of this type of abuse in the overall pattern of dating abuse. As a first step, it was extremely fruitful. Just as predicted by the clinical literature on battered women, 285 psychological abuse, as measured by the Inventory of Psychological Abuse, was found to be an extremely important and powerful aspect of abusive interactions in intimate relationships. In a correlational study such as this one, no conclusions can be drawn concerning the direction of causality, of course, but the strong relationships between reported rates of inflicting and sustaining PsyA and ayapy measure of psychological and social adjustment were remarkable. The IPA performed very well also. The fact that respondents' scores on this measure were so strongly associated with measures of sexual and non-sexual physical abuse and with measures of relationship adjustment (sexual satisfaction and relationship quality) provides good evidence of the measure's construct validity. In addition, the factor structure which was derived on the basis of this sample's responses was interesting and may be theoretically important. Specifically, the subscales' different patterns of correlation with the measures of sexual and non-sexual physical abuse were promising. The differences suggest the possibility that some particular types of psychologically abusive interactions (which were reported with relatively high frequency among this non-clinical sample) may provide risk markers for other, less common, and more physically dangerous forms of abuse. Thus, the IPA may also hold promise, not only as a research instrument, but as an assessment and/or educational 286 device in working with individuals or couples who have concerns about their relationships. Additional research is necessary in order to develop norms and confirm the factor structure of the subscales with other populations, as well as to assess the measure's performance with clinical populations of interest. The Conflict Tactic Scale and The Naw Violence Measapa The New Violence Measure was, as predicted, found to be more highly correlated with measures of reported emotional distress, as well as with the BSI scales measuring symptoms of Somatization, Depression, Anxiety, Hostility, Phobic Anxiety, Psychoticism and The General Severity Index. The correlations with the measures of the actual power distributions in the relationship gave mixed results. While the New Violence Measure was more strongly associated with the measure of the respondents' total amount of decision- making power in their relationships, and, to a lesser extent with the measure of influence over independent decisions, it was laaa strongly associated with the measure of influence over the partner's decisions. It has been argued that it is essential to include, in a measure of NsPA, an assessment of physical consequences, as well as other contextual factors (such as the perceived dangerousness rating used in The New Violence Measure) in order to achieve a more accurate picture of NsPA experiences and outcomes. And, the injury and dangerousness variables, 287 which the New Violence Measure used as weighting factors, yapa, in fact, found to be strongly related to measures of psychological symptoms, as well as to respondents' ratings of their emotiOnal reactions to NsPA incidents and to the relationship power variables. This provides a strong argument for their inclusion in an instrument that assesses the important dimensions of NsPA experiences. The CTS' lack of attention to such issues has been cited as one of it's major weaknesses and these shortcomings have been seen as responsible for its consistent failure to find sex differences in the descriptions of NsPA incidents (Walker, 1989). The revisions which were made to the CTS, that resulted in the creation of The New Violence Measure, were designed to address these weaknesses and to include an assessment of important contextual features in the measurement of NsPA. Contrary to expectations, though, the inclusion of these additional dimensions of NsPA did not reveal any significant sex differences in experiences of either inflicting or sustaining NsPA. However, even though the differences were not significant, it is interesting to compare the picture of men's and women's experiences with NsPA as reflected by the two different measurement strategies in Table 16 (p. 186). The changes made to the New Violence Measure (taking injuries and perceived dangerousness into account, and adding measures of several kinds of physical abuse thought to be common in incidents of sexual abuse) do result in a shift in men's and 288 women's mean scores in the directions predicted by the critics of the CTS. Men's NsPA scores on the CTS, indicate that, on the average, they reported sustaining over twice as much NsPA as did the women respondents (i = 3.10 as compared to 1.32). However, when assessed by The New Violence Measure, while men still reported sustaining more NsPA, the means were, proportionally, not as discrepant (for men, Y'= 14.04; for women, 11.21). The impact of the different measurement strategies on the mean scores for inflicted violence is similarly interesting. With the CTS scoring strategy, women reported inflicting slightly more violence than did men, but the means were very close (for women, Y = 1.67; for men, 1.56). With The New Violence Measure, however, the picture was different, with men reporting inflicting over twice as much NsPA (Y = 14.54) as women (NI: 6.09). Thus, the New Violence Measure, as compared with the CTS, shifts the overall picture of men's and women's NsPA experiences, in the directions predicted by the critics of the CTS. But, nonetheless, the resulting gender differences were not significant in any case. There are a number of possibilities to explain this failure of The New Violence Measure to reveal significant gender differences. First of all, it may be that idiosyncracies about this particular sample were responsible for the failure of the new measure to find sex differences to which it may, in fact, be sensitive. A number of the findings suggest this possibility. For instance, it is 289 inconsistent with previous studies that there were no statistically significant sex differences found in this study in either the rates of reported injuries or in the ratings of the severity of injuries sustained. Also, Sugarman and Hotaling (1989) found, in their review of gender differences in the dating violence literature, that previous studies utilizing the CTS have generally found women to be slightly more likely to report victimization experiences than men. But, among this sample (as mentioned above), though the difference was not significant, the men reported an average CTS victimization score that was twice that of the women's average. These discrepancies between these respondents' reported experiences and those reported by subjects in previous studies call into question whether this sample is a representative one. It will be recalled that the sample was shown to be non-representative of the larger university community with respect to a number of demographic variables. Thus, it is possible that, because they are not representative, this study was not an adequate test of the new measure's ability to assess gender differences. Another possibility is that the specific revisions which were made to the CTS failed to measure adequately the important contextual dimensions of NsPA incidents which the critics of the CTS maintain must be assessed to gain an accurate picture of the differences in men's and women's experiences. However, The New Violence measure did attempt to assess those contextual features which have been mentioned 290 most frequently -- injuries, and variations in intensity or dangerousness. Another item in the questionnaire attempted to assess whether or not the reported NsPA was engaged in for self-defensive purposes, another contextual feature which has been explicitly mentioned by the critics of the CTS. But, here too, the data failed to reveal any gender differences. So, if it is true that men's and women's experiences with NsPA behaviors in dating relationships are substantially different in regard to the context of the experiences, those differences do not appear to be in these specific contextual features which have been mentioned most frequently in the criticisms of the CTS (at least not in this sample, as these features were measured here). Another possible reason for the absence of hypothesized sex differences is that The New Violence Measure failed to correct more fundamental weaknesses of the CTS. For instance, The New Violence Measure, like the CTS, does not include direct measures of SA experiences, the omission of which, as discussed above, creates an inaccurate picture of DA experiences, perhaps especially in regard to sex differences. Another fundamental weakness of the CTS was discussed by walker (1989), who argued that apy tally of violent behaviors is bound to be biased by differential reporting effects that have been noted in the work with clinical samples of couples with experiences of domestic violence. As discussed in the literature review, both empirical studies and the clinical 291 literature on domestic violence have suggested that, among clinically identified samples, males who inflict violence tend to consistently underreport their own violence (Browning & Dutton, 1986; Ganley, 1981; Jouriles & O'Leary, 1985; Ptacek, 1988b; Sonkin, Martin & Walker, 1985). Walker notes that battered women tend to underreport their experiences of victimization as well. It is possible that these same biases affect the reports of NsPA experiences in dating relationships among a non-clinical sample. Certainly, if these consistent biasing effects fig occur in reporting NsPA experiences, then the New Violence Measure would be no less susceptible to them than would the CTS. There is no way to tell whether or not the current data is affected by such a differential reporting bias. But, among the current sample, one very dramatic difference between men's and women's reports of their NsPA experiences was found: Men's reports of inflicting and sustaining NsPA were correlated .93, while for women, this correlation was .29. Of course, since the men and women in this sample were not describing the same relationships (i.e. were not couples), this may reflect true differences in their) experiences. However, the result is consistent with those of Browning and Dutton's (1986) study with 30 clinically identified assaultive couples. They found that, even with couples, while men tended to describe mutually violent relationships, women did not. Thus, it seems probable that the difference found here reflects a perceptual or reporting 292 difference between men and women. Of course, the data from the current study does not permit a judgement about whether men's or women's reports (or neither) more accurately reflect the true experiences with NsPA in dating relationships. But, the finding is suggestive of gender reporting differences like those discussed by Walker (1989) and others. The final possibility concerning the absence of statistically significant sex differences in rates of reported NsPA experiences is, of course, that the data accurately reflect men's and women's experiences in dating relationships. It may be that the finding of no significant sex differences in rates of NsPA, even with the New Violence Measure, is not due to the unrepresentative nature nor small size of the present sample, nor to the inadequacy of the measurement techniques employed, but instead, is a valid depiction of the NsPA interactions between dating partners. Perhaps there are no significant sex differences in the actual incidence of specific NsPA behaviors between dating partners in a non-clinical population. Some support for this conclusion comes from a review of the literature in the field. As discussed in the previous literature review, similar findings have been reported by other investigators and have led some to conclude that women and men are comparably violent, or even that women are slightly more violent than men, in intimate relationships. This conclusion has been made both about non-clinical dating populations (relevant 293 studies are summarized in Sugarman & Hotaling, 1989) and about community samples of married and co-habitating adults (Straus, Gelles & Steinmetz, 1980). So previous empirical findings which are in accord with those of the current study regarding a lack of gender differences in the incidence of NsPA provide some evidence that the findings from the present study may be valid. Additional arguments for this conclusion come from the discussions in the literature about the probable causes of violence in intimate relationships. These discussions propose a number of social and psychological factors which could be expected to operate equally upon both males and females in fostering violent behaviors against intimate partners. The models of dating violence proposed by Riggs and O'Leary (1989) and presented earlier incorporate or allude to a large number of such factors. They include, for example: exposure to models of aggression in intimate relationships in the past, not only in the family of origin but also via the media and elsewhere; childhood experiences as the victim of physical punishment; the prevalent acceptance of violence as an appropriate means of expression in intimate relationships -- especially in the family (Gelles & Straus, 1988; Riggs and O'Leary, 1989; Straus, 1980a); and, high levels of stress and conflict associated with intimacy (Farrington, 1980; Foss, 1980; Gelles & Straus, 1988; Riggs and O'Leary, 1989; Straus, 1980a). All of these influences 294 are clearly present throughout the lives of REED men and women. So, while it may be expected that men would be more prone to acting aggressively, for the reasons explicated by the Feminist/Social Control theorists, still, there are powerful social forces operating in this culture which would be expected to foster violence in intimate relationships on the part of pagg men and women. Thus, it seems reasonable to expect that women, who grow up in a culture which is so supportive of violence in intimate and family relationships, would not be immune from those socialization forces. It seems reasonable, further, to assume that they, as well as men, would be likely to engage in violence against their partners in intimate relationships with some frequency and that that violence would not be exclusively, or perhaps even primarily, self-defensive, contrary to what the Feminist/Social Control Model predicts. And, in fact, the results of this study regarding self- defensive violence indicated that women reported much more non—self-defensive NsPA directed against dating partners than the model would lead one to expect. Only 23% of the women who indicated that they had engaged in NsPA in a dating relationship indicated that they did so, even in part, "to defend themselves". A similar proportion of men (21%) indicated this as one of the reasons for their NsPA actions against their partners. Makepeace (1986) found a much higher proportion of women reporting self-defense as a motive for 295 their NsPA actions. Here again, the differences between the present and previous results may be due to idiosyncracies of the present sample, so the current findings should be regarded cautiously. Still, 37 women in this sample (77% of those who reported inflicting NsPA) reported that the NsPA that they inflicted against their dating partners was not motivated by self-defense. Even if the proportions are inaccurate due to an unrepresentative sample, these women's reports of their experiences still cannot be totally disregarded, and they are not consistent with the Feminist/Social Control model of intimate violence. So, if valid, the prevalence of non-self-defensive NsPA inflicted by women against their partners, in combination with the non-significant differences in men's and women's rates of inflicting and sustaining NsPA, present a major challenge to the position of the Feminist/Social Control model as it is currently articulated, which -- it will be remembered -- posits that women's violence in intimate relationships is overwhelmingly self-defensive and that women are the victims of violence, while men are the perpetrators. This position, though, seems overly restrictive and limiting, and (given the consistency with which non-self- defensive NsPA is found among women in non-clinical samples) it, unfortunately, leaves the Feminist/Social Control model extremely vulnerable to offhanded rejection -- "unfortunate", because the model has other significant and important contributions to make to the understanding of intimate 296 violence. It is argued here that this position of denying the frequency with which women engage in non-self-defensive acts of violence against intimate partners is app necessary to the Feminist/Social Control perspective on domestic violence. It is possible to acknowledge that women engage in violent behaviors against intimate partners, for reasons other than self-defense, without equating men's and women's experiences with intimate violence. Even if women do inflict a substantial number of non-self-defensive violent behaviors against their partners, it is clear that violence and aggression in intimate relationships is not an "equal problem" for men and women. The experience of intimate violence is very different for men and women. In the first place, as others have argued (Saunders, 1988; Straus, 1980), even though women are not the sole victims of intimate violence, they are, as the Feminist/Social Control Model predicts, more severely victimized by intimate violence than are men. The evidence concerning the physical consequences of NsPA in intimate relationships, both in the domestic violence and in the dating violence fields, is scarce, but consistent. Berk, Berk, Loseke and Rauma (1983) studied police reports from 262 domestic violence incidents and found that women were much more likely to be injured in such incidents than were their partners -- both in terms of the number and the severity of injuries sustained. Sugarman and Hotaling (1989) draw the 297 same conclusion from their review of the few studies of DA which have focused on the physical consequences. The results of the present study provide some support for this conclusion. Consistent with the earlier studies, more women than men reported injuries (32.7% of the women experiencing NsPA as compared with 21.4% of the men), but this difference was not statistically significant. In addition, as in earlier studies (Makepeace, 1986), even though most reported injuries from NsPA experiences were mild, all reported moderate and severe injuries were sustained by women. Aside from physical injuries, Makepeace (1986) also reported that women experienced more emotional distress as a result of their DA victimization experiences. The results of the present study provide some evidence for this as well. Among those who reported being victimized by all three types of DA, women reported experiencing more severe emotional trauma than men as a result of both NsPA and SA. The same difference existed between men and women in their reported immediate emotional reactions to SA experiences among the larger group of men and women who had SA victimization experiences. However, among all respondents who reported NsPA experiences, men and women reported similarly high levels of emotional trauma. But, considered as a whole, the evidence from this study and from previous ones does suggest that, even though both men and women suffer negative 298 consequences of DA victimization, women experience more serious consequences -- both physically and emotionally. In addition, as will be discussed at greater length in the following section, the results of the current study further suggest that the relationship consequences of violence between intimate partners may be more severe for women than for men. No matter whether men or women reported inflicting the violence against their intimate partner, it was associated with patterns of male dominance in decision making. It is in making sense of these findings, that the Feminist/Social Control Model is particularly relevant. In conclusion, the New Violence Measure showed some promise in that it provided a better predictor of psychological and adjustment outcomes for victims of NsPA than did the CTS-based scoring strategy. It also affected the average scores of men's and women's rates of inflicting and sustaining NsPA in the predicted directions, although the differences were not significant. However, it is questionable whether or not the present sample is an adequately representative one and the measure merits further testing. The question of whether or not the absence of gender differences in the rates of inflicted and sustained violence in dating relationships is a valid and reliable finding will have to await further research with a more representative sample. Research with dating couples would be particularly 299 helpful in this regard. It seems clear at this juncture, though, that accurate gender comparisons would be more likely with a measure, like The New Violence Measure, that takes into account injuries and that assesses the forms of physical violence such as physical restraint, which are known to occur frequently in the context of sexual assaults. And, as Sugarman and Hotaling (1989) have noted, it is time to move beyond the superficial, quantitative approach to dating aggression and assess more contextual and consequential factors. In addition, in order to make valid gender comparisons in the experiences of DA, it is essential, as discussed above, to expand the perspective on intimate violence so that sexual abuse is always considered and assessed in studies of "dating violence". The Relevance of the Feminist-Social Control Model In regard to some overall comparisons of men's and women's responses, the findings of this study supported some of the Feminist/Social Control theorists' general assertions concerning the relationships between men and women. For instance, as predicted by the model, men, as compared to women, tended to see violent behavior as more appropriate in intimate male/female relationships. This finding is consistent with those from a number of previous studies as well (Emery, 1987; Finn, 1986; Fischer, 1986). Also, men tended to subscribe to more traditional positions vis a vis sex-role attitudes, which would be expected by the 300 Feminist/Social Control theorists, since these attitudes would be more supportive of male positions of dominance in intimate relationships. As would also be expected by these theorists, men reported ideals of having more power over their partners' independent decisions than did the women, and they expressed more dissatisfaction with the amount of power that they had in this arena, indicating, on the average, that they wanted more than they had. This was not true for the women respondents. In addition, men reported having more decision-making power over relationship decisions than did the women. Also, women, on the whole, rated their partners as being more dominant than men rated their partners: Women rated their partners as wanting to have slightly mgpa control over decisions than they, themselves had, while men rated their partners as being slightly on the submissive end of the scale (i.e., as wanting laaa control over decisions than their male partners). These results are all consistent with the Feminist/Social Control theorists' position that male- dominance is normative in intimate relationships in this culture. The results also provided some support for the Feminist/Social Control Models' predictions concerning the role of the attitude, resource and power distribution variables in violence and abuse in intimate relationships. The complete set of these variables, as a whole, was found to discriminate significantly between all criterion groups in the six discriminant function analyses. In other words, 301 these predictors accounted for significant amounts of variation between abusive and non-abusive individuals across all three types of inflicted dating abuse. This was true for both men and women. This fact provides some general support for the feminist/social control model's conception of dating abuse and suggests that the model is, in fact, relevant to the study of dating abuse, contrary to the recent assertions by several investigators in the field (Makepeace, 1989; Riggs & O'Leary, 1989). In addition to these findings, the results that pertained to the decision-making power variables were very much in line with the feminist/social control model, but with some interesting surprises. The analyses pertaining to the effects of different levels of DA victimization upon overall levels of decision-making power in the relationship revealed clear sex differences. For women, increasing levels of victimization were associated with decreasing levels of overall power in the relationship, as might be expected. However, for men the results were very different. While moderate levels of victimization were associated with decreased reports of decision-making power, the men who reported the highest levels of victimization reported having the highest levels of overall relationship power. This is a particularly dramatic effect in that it is based, not upon men's reports of the abuse they inflict upon their partners, but on their reported experiences of victimization. This finding is consistent with the feminist/social control model, 302 since it views all reports of violence primarily as indicators of male aggression (even male victimization reports, since female violence is seen as primarily self- defensive or reactive to male initiated violence), and increased male aggression would be expected to be associated with increased male power. However, outside of the Feminist/Social Control model, this result is not a prediction one would make about the consequences of victimization. Also consistent with the predictions of the feminist/social control model were the findings that, across all three types of DA, abusive males were more likely to espouse dominance ideals, but women who reported inflicting DA, in contrast, endorsed goals of equality. Other interesting results were observed in the relationships between power and dominance issues and inflicted abuse.‘° Both men and women reported having significantly decreasing levels of power over their own independent decisions as they reported increasing levels of inflicting DA. This finding may be seen as being consistent with that of Arias and Beach (1987) in their study of married couples. They found that reports of violence, for both men and women, were negatively correlated with scores on the Autonomy Scale of The Personality Research Form. In the current study, however, important gender differences were 1° This discussion draws upon the data that is presented in Table 33, on page 222. 303 also found. Abusive men simultaneously reported this loss of autonomy (decreased power over independent decisions) and corresponding increases in the amount of power over their partners' decisions and over relationship decisions, across all three types of DA. In contrast, women reported increases in power over their partner's decisions only in relationship to reported levels of inflicted PsyA, but not of inflicted SA or NsPA. And, even in the case of PsyA, the One-way ANOVAs revealed that the differences between the offender groups were not significant. Similarly, men also reported that they had higher goals for dominance over their partners' decisions as abusiveness increased and lpya; goals for retaining power over their own independent decisions, whereas for the women, these trends were neither significant nor did they discriminate significantly between the offender groups. Also, it was predicted that both men and women who are involved in abusive relationships would report having less power than they wanted. However, in fact, as men reported increasing levels of dominance over their partners' decisions as abusiveness increased, they also reported increasing satisfaction with their power over this domain of decision- making. Consistent with men's reports of having more influence over their partner's decisions, were women's reports of decreasing power over their own independent decisions as abusiveness increased, and also the fact that, in general, women tended to rate their partners as having dominance goals 304 in relationships. However, men also reported decreasing levels of influence over their own decisions with increased abusiveness, and in fact, they reported even lower levels of autonomy over these decisions than did women (in general, across all groups). And yet, these reports were not similarly corroborated by women's reporting increased influence over their male partners' independent decisions. And, in fact, the men themselves reported that their partners tended to have 'submission' goals in relationships (i.e. men in general said their partners believed that the men should have more control over decisions). And this was true for all of the male offender groups - in other words, no group rated their partners as endorsing dominance goals. 80, men's reported experience of the loss of autonomy over their own decisions, is confusing. It may be that these results reflect the observed tendency for abusive men to be unusually dependent upon their partners, as has been noted in the clinical literature pertaining to batterers in the domestic violence field (Ganley, 1981; Sonkin, Martin & walker, 1985; Walker, 1979). In fact, these authors propose that the abusive man's dependence upon his partner is threatening to his sense of self and that his violence is, in part, motivated by a need to defend against his feelings of dependency. The possibility that this pattern may also exist between dating partners who experience abuse is interesting and warrants further research. 305 In combination, these results pertinent to dominance and control issues indicate that experiences with dating abuse agaJ in fact, clearly related to patterns of male dominance and to men's goals of dominance in their dating relationships, as the feminist/social control model predicts. Furthermore, these effects appear to be in the decision- making domain that is related specifically to the woman's independent decisions. Abusive men - as compared to non- abusive men and to women - reported striving for and achieving more control over their partners' personal, independent decisions. They wanted and had more power over their dating partners' personal decisions: how she spends her money and wears her hair, what she does when she's not with him, and what classes and career plans she makes for herself. These results appear to be consistent with those of Dutton and Browning (1988), who found that violent husbands, in comparison with non-violent, reported feeling more anger in response to a vignette depicting a marital argument about the wife's desire for more independent activity. The violent husbands also rated this vignette as being more relevant to their own marital disputes than did non-violent men. Dutton and Browning interpreted these results as being consistent with their view that male violence in intimate relationships results from intimacy anxieties, which include the fear of abandonment. They propose that a woman's autonomous activity triggers this fear. If so, then struggles about the woman's autonomous decisions would be expected to be especially 306 relevant to male violence, as the results of the current study suggest. Even though these results were not predicted, they are consistent with the Feminist/Social Control Model's view of dating aggression. It has been suggested before, both in the domestic violence literature as well as by Stets and Straus (1989) in their recent article about dating violence, that conflicts over control trigger incidents of NsPA. However, the results of this study, and of Dutton and Browning's (1988), suggest that the content of these conflicts may be highly specific. The results of the current study additionally indicate that they are probably not gender neutral. Instead, they suggest that the conflicts that precede violence are more likely to involve disagreements over the man's attempts to control his partner's independent choices and behaviors, but not vice versa. In other words, NsPA appears to arise out of conflicts which surround the attempts of the male partner to dominate the woman, just as the Feminist/Social Control Model predicts. If that is true, then, no matter who does what in the physical fight, NsPA is a very different experience for the man and the woman involved, with different meanings and implications, as well as different consequences. Gender-neutral theories, like social learning theory, or the models based upon that theory and put forth by Riggs and O'Leary (1989) -- while they are consistent with a number of the findings from the current study -- are not sufficient to explain these important gender 307 differences. Of the existing theories of intimate violence, only The Feminist/Social Control Model is consistent with these findings. However, the results of this study did pp§_conform with the predictions of the feminist/social control model in several important instances. For example, the feminist/social control model predicts that men who have goals of male dominance in their relationships but who do not have legitimate resources to achieve it will use violence as a "resource of last resort". However, at least among this group of relatively advantaged, upper-middle and middle-class college students, the results suggested instead that, for men, having resources that provide a legitimate base of dominance over one's partner is associated with higher rates of sexual and psychological abuse. And men who reported having fewer resources than their partners, although in the minority, reported very low rates of abuse, independent of their dominance goals. The results from the analyses pertaining to the women's reports of rates of victimization were also disconfirming of the predictions regarding the role of resource balance in the etiology of DA. While women's rates of SA and PsyA victimization yaga associated with their ratings of their partner's dominance goals, their ratings of the balance of resources was not significantly related to their experiences of NsPA victimization (though, here again, contrary to expectations, the mean reported rates were actually higher 308 for those women who reported having fewer resources than their partners). It is difficult to make sense of these results pertaining to the resource balance between dating partners. It is possible that resource imbalances that tend to support or create conditions of domination by the male actually foster sexual and psychological abuse, but that is inconsistent with the results of previous studies of NsPA in both the dating abuse field (DeMaris, 1987; Sigelman, Berry & Wiles, 1984), and in the domestic violence field (Allen & Straus, 1980). Given that the present sample was shown to be significantly unrepresentative of the larger university community in regard to a number of demographic variables, it is possible that these results are due to idiosyncracies of the sample. Further research is required in order to sort out these inconsistencies in the findings from this and previous studies concerning the relationship between resource balance and intimate violence. 80, overall, in some ways the Feminist/Social Control Model was supported by the results of the current study, and in some ways it was not. The major challenges to the model stem from these findings regarding resources, and from the failure to find predicted gender differences in the rates of involvement in NsPA, which was discussed in the previous section. One additional finding from the discriminant function analyses is interesting to note in connection with 309 this lack of observed gender differences in the rates of NsPA: The variables which were predicted by this model to account for abuse in dating relationships did a much better job, across all three types of DA, in accounting for the observed variance in men's experiences with inflicting abuse than with women's. Since the feminist/social control model views women as victims and ‘reactors' and men as instigators in the abuse in intimate relationships, the theory was developed to account for violence instigated by males against women in intimate relationships. This difference, then, is, perhaps, not surprising. Still, in combination with the significant proportion of women who report inflicting non- self-defensive violence toward their dating partners, this difference reinforces the position, taken above, that the Feminist/Social Control Model underestimates and does not adequately account for the incidence of woman-inflicted violence in intimate relationships among a non-clinical sample. As discussed in the previous section, it is probable: (a) that social and psychological forces in addition to those pinpointed by the Feminist/Social Control theorists (which are: widely accepted norms of male dominance in intimate relationships and attitudes accepting and encouraging of male violence) foster NsPA in intimate relationships, and; (b) that these forces encourage violence among women as well as men. Some of these social and psychological influences mentioned in the earlier discussion (and discussed by 310 previous theorists in the field) include; exposure to models of aggression in intimate relationships (both in vivo and via the media), childhood experiences with physical punishment and/or violence in family relationships, the prevalent acceptance of violence as an appropriate means of dealing with stress and conflict in intimate relationships (especially in families) -- by both men and women, and high levels of stress and conflict associated with intimacy (Farrington, 1980; Foss, 1980; Gelles & Straus, 1988; Riggs & O'Leary, 1989; Straus, 1980a). The Feminist/Social Control theorists focus upon the acceptance of violence exclusively as a component of the traditional masculine gender-role. In contrast, the position taken here is that, while traditional gender roles undoubtedly socialize men to be more accepting of violence and to express themselves more easily with physical aggression in a variety of social roles, the norm of violence as an acceptable means of expression in interpersonal relationships is a widespread, pervasive influence which affects not only men, but women as well -- especially as an appropriate way to manage conflict and stress within family and other intimate relationships (Gelles & Straus, 1988; Straus, 1980a). So, while men might be expected to be more prone to acting aggressively —- especially outside of family and intimate relationships -— there are powerful social influences which are expected to encourage both men and women to inflict violence against their intimate partners. 311 On the face of it, it seems unfeasible that the widespread acceptance of violence as an appropriate means of expression in intimate and family relationships among QQEE men and women could be related to traditional gender roles. Yet a number of feminist historians and sociologists have suggested that it is (Kennedy, 1970; Polk, 1974; Starhawk, 1987). Polk, for instance, asserts that, because men have more economic and social power, the values which are imbedded in the masculine gender role (including aggressiveness and the acceptance of violence) ”constitute the dominant and visible culture of society" (Polk, 1974, p. 417). She further suggests that, while men are socialized almost exclusively to the masculine value system, women receive dual socialization because of the dominance of the masculine values stemming from the pervasive influence of male- controlled institutions. Thus, from this perspective, the widespread acceptance of violence and aggression, even as it is expressed among women, is interpreted by some feminist theorists as a reflection of the male gender role. This becomes a dominant influence in shaping cultural norms because of the economic and social inequalities between men and women, which results in the traditionally "masculine” values being more highly valued by and more strongly promoted throughout the culture. This position seems to be largely compatible with the Feminist/Social Control Model of intimate violence, except that, as the model is currently articulated it proposes that 312 women are only -- or certainly primarily -- the targets of NsPA behaviors and men the perpetrators. But, as stated earlier, that position is not well supported by the results of this or other previous studies, and seems not only unwarranted, but unnecessary. It seems nearly undeniable that women express themselves via physical aggression in non- self-defensive ways with some frequency -- especially in intimate and family relationships. And to acknowledge that fact does not affect the main positions of the Feminist/Social Control Model: (a) that violence is intricately associated with patterns of male-dominance in intimate relationships; and (b) that women are the primary victims of this violence, in that they suffer more severely from its consequences physically, emotionally, and in terms of relationship consequences. But the current formulation of the Feminist/Social Control model maintains that the association between NsPA and patterns of male dominance stem from the fact that men who have dominance goals which are unobtainable due to a lack of legitimate resources, resort to using violence as a means to those ends. If BREE men and women inflict violence against intimate partners, how is the association between this violence and male-dominance to be explained? The results of this study suggest that violence may occur in an intimate relationship in the context of arguments stemming from the man's attempt to control or dominate the woman's independent actions. Then, once the violence occurs (regardless of who 313 initiates it), even though both partners may suffer negative consequences, it is likely that the balance of power between the partners begins to shift more or less gradually over time resulting in the woman losing more and more influence in the relationship while the man gains power accordingly. Thus, intimate violence would be expected to result in the establishment of patterns of male-dominance, no matter who inflicts what specific acts of non-sexual physical aggression against whom. This would be expected to occur because of several factors: (a) women experience more severe consequences from the violence both physically and emotionally; (b) women and men undergo different processes of psychological development and gender-role socialization experiences which are expected to result in their having different reactions to violent conflicts in intimate relationships; and, (c) different economic and social conditions confront men and women in a sexist society and these result in their having different "stakes" in establishing and maintaining an intimate relationship which, in turn, differentially affects their reactions to relationship conflict. These factors will be discussed below. The fact that women experience more severe physical and emotional consequences from the victimization experiences of intimate violence has been established in both the domestic violence and the dating abuse literature and was also found to be true (although not so unequivocally so) among the 314 current sample of college students, as discussed previously. This fact undoubtedly stems from men's greater physical stature and strength. In addition, it is likely that, by the time men and women become involved in intimate relationships, most men have had more direct experiences with physical combat. Although the strength and importance of the association has been recently questioned (Deaux, 1984, 1985), it has been a consistent finding that it is more common for boys and young men to express themselves via physical aggression than for females (Block, 1983; Deaux, 1984, 1985; Donelson, 1977; Frieze, Parsons, Johnson, Ruble & Zellman, 1978). In addition, it seems probable that males are more likely to be exposed to physically aggressive activities in a variety of arenas, including, play and other informal interactions with peers, organized sports, and military training and service. So, men's increased exposure to physical combat in combination with their larger physical stature are likely to be important factors in leading to the observed situation where men are more likely to inflict harm in physical confrontations than are women. It is also likely that they would tend to be less intimidated and frightened by these types of interactions, although not necessarily otherwise less "upset" by the content of the argument. The differences in these emotional reactions to incidents of physical violence may be exacerbated by the different gender—role socialization experiences that boys and girls experience. While aggressive behavior tends to be 315 discouraged among little girls, it is much more likely to be encouraged or tolerated by parents when exhibited by boys (Ferguson, 1977). Since physical aggression is considered to be a more acceptable, gender-appropriate behavior for males than females, it is also likely that it is more ”ego- syntonic" for males. The fact than men, in general, tend to view interpersonal violence as somewhat more acceptable in intimate relationships is consistent with this position and has been established in this study as well as in previous investigations, as discussed earlier. It is possible that this greater acceptance of violent behavior results in men being less remorseful or self-condemning than are women, as well as less judgmental of their partner's aggressive behaviors, after incidents of intimate violence have occurred. If so, this may be an additional factor in men's being generally less upset by incidents of physical violence in their intimate relationships than are women. If the man experiences less severe physical and emotional consequences than his partner does from incidents of physical violence in intimate contexts, as a result of the above factors, than it is likely that he will not be as motivated as she is to avoid similar physical confrontations in the future. If so, the woman is more likely to "give in" and subordinate her own needs and desires to his in subsequent conflicts in order to keep the peace and avoid escalation to physical violence. This would result in a shift toward a situation wherein she is more submissive and 316 he more dominant in subsequent decision—making. If, as the results of this study suggest, the conflicts tend to revolve around issues pertaining to her independent decisions, she may give up more and more of her own autonomy in order to avoid conflicts that might lead to physical violence. Women's gender-role socialization experiences may result in an even greater likelihood that, over time, in conflictful intimate relationships such as those in which incidents of violence occur, women will generally tend to lose power and control while men become more powerful and dominant. Developments in feminist psychological theory (Chodorow, 1978; Gilligan, 1982; Miller, 1976) and the feminist popular and clinical literature (Becker, 1987; Friedan, 1963; Russianoff, 1981; Sanford & Donovan, 1984) converge on the assertion that women are, in general, more strongly oriented 'toward the establishment and maintenance of intimate relationships than are men. Though these authors differ in their emphases, they (andothers) agree that women are systematically taught to be more strongly oriented toward relationships by both direct and indirect cultural influences in child—rearing and educational practices, by the traditional strictly gender-differentiated structure of family life and by other social and economic inequities in a sexist society. Miller (1976) proposes that, since the intimate relationships between women and men occur in a sexist and male-dominated context, they incorporate essential 317 characteristics of any extremely power-differentiated relationship between a dominant and a subordinate individual. Since girls are raised within this context, Miller suggests that they develop more relationship-oriented skills as a means of survival in a culture which restricts their independent options for economic self-sufficiency and/or self-expression. Thus, a woman's social status and economic survival is much more dependent upon her being in a relationship than is a man's. This results in women and men's psychological development proceeding along very different courses. She describes some of the consequences of these conditions on women's development: One central feature is that women stay with, build on, and develop in a context of attachment and affiliation with others. Indeed, women's sense of self becomes very much organized around being able to make and then to maintain affiliations and relationships. Eventually, for many women the threat of disruption of an affiliation is perceived not as just a loss of a relationship but as something closer to a total loss of self. (p. 83) This, she and others maintain, results in a situation wherein women, more than men, are likely to assume the primary responsibility for smoothing out and resolving whatever problems arise in their intimate relationships. Women become the "peace-keepers". They are more likely to perceive the signs of interpersonal conflict when they occur, to be concerned about the meaning of those signs, and to initiate attempts (either overt or covert) to resolve the conflicts -- often by giving up their own desires or needs in pursuit of maintaining or regaining interpersonal harmony. This would 318 be especially true in regard to giving up power over autonomous, independent decisions, if (as this study suggests may be common) their male intimate partners had goals of having more influence over them in these areas and this was an area of frequent disagreement and conflict in the relationship. A substantial amount of empirical evidence from recent research on gender differences is consistent with these theoretical assertions. Deaux, in her recent reviews of this research (1984, 1985), concludes that studies on gender differences in conformity and social influence have consistently found women to be "more easily influenced than men and...more likely to conform in group pressure situations..." (1985, p.61). These conclusions are based upon research conducted in social situations with other research subjects with whom the women do not have intimate or significant relationships. Although it is not certain that these same gender differences would emerge when the interpersonal relationships of interest are more intimate, it is probable that women's tendencies toward conforming to the pressures of others would be even more pronounced under those circumstances. Additional research has indicated that females have also been found to both encode and decode non-verbal interpersonal cues more accurately (Deaux, 1985); to exhibit more accurate empathic skills; to choose more affiliative, cooperative, and nurturant activities; and to develop more intense and 319 intimate friendship patterns than do males (Block, 1983). All of these findings are consistent with these feminist theories of gender differences in psychological development. One implication of these theories and empirical findings is that women would tend to lose power and influence, and men gain it as a direct corollary, in apy highly conflictful intimate relationship between a man and woman. This would be especially likely if the man wanted to have more control over the woman's independent decisions and activities and if he initiated attempts to exercise that control. If the woman resists these attempts, and a conflict ensues, it is likely that, over time, the woman will give up her power in this area in order to avoid future conflicts, to the extent that she wishes to maintain the relationship. This would happen because of the woman's tendency to be more sensitive and responsive to relationship problems and to assume primary responsibility for fixing or avoiding them in the future. This would be expected to be an especially strong effect in relationships where conflicts escalated to the point of physical violence, given that women would be particularly intimidated by these incidents, due to the more severe consequences of their victimization experiences, as discussed above. While there may appear to be some similarities between this position and that taken in the recent explosion of literature on "co-dependency" (Cermak, 1989; Subby, 1987) and "women who love too much" (Norwood, 1985), there is one 320 essential difference. The feminist theorists maintain that women's tendency to be more strongly oriented toward their connections with others than are men and to assume greater responsibility for maintaining these relationships, often by means of self-sacrifice, is normative. They see it as stemming from strict gender-differentiated family roles, strong social pressures, and economic inequities which, in combination, make it virtually imperative for women's survival and sense of well-being that they be thus oriented toward relationships. These authors also maintain that these relationship-building, cooperative traits and skills are extremely important and valuable ones for society as a whole; and, further, that they tend to be undervalued precisely because they are associated more strongly with women. The Co-Dependency literature, in contrast, suggests that a strong orientation toward maintaining relationships at the cost of sacrificing one's own interests is a mental illness which stems from growing up in "dysfunctional" families. From this perspective this is a disease from which one must "recover" via intense psychological treatment (albeit, often in peer guided, 12-step programs). This group of authors virtually ignores the social and economic realities of a sexist society which necessitate that women adopt a stronger orientation toward establishing and maintaining intimate relationships with men. What is proposed here are several major refinements of the Feminist/Social Control Model of intimate violence. 321 Because of the pervasive climate of the acceptance of violence —- especially as an appropriate means of dealing with conflict and stress in family and other intimate relationships —- non-sexual physical violence is wide-spread in intimate relationships and it is perpetrated by women as well as men. It is possible that this climate of acceptance of aggression and violence is due to the fact that the prevailing cultural climate reflects more strongly the values of the traditional masculine gender role. Even though both men and women inflict NsPA in relationships, women are more severely victimized, intimidated and dominated by violence in intimate relationships than are men. This is because: (a) women are more often the victim of sexual abuse, which also occurs frequently in intimate relationships; (b) women suffer more severe consequences, both physically and emotionally; (c) women are more likely to assume responsibility for maintaining the peace and avoiding future conflicts in intimate relationships even to the point of giving up their own power and interests to achieve that end (this is due to economic and social sexism and the strict gender- differentiated nature of family roles which result in different courses of psychological development for men and women); (d) women in conflictful relationships will be especially likely to lose power and influence over their own independent decisions because men generally desire more control over these areas of decision-making (especially abusive men), thus the couple's conflicts will revolve 322 strongly around the man's attempts to dominate her independent activity, and, over time -- to the extent that she attempts to keep the peace and desires to maintain the relationship —- she will lose and her partner will gain more and more power in this area. These expansions on the Feminist/Social Control Model do not include any new or ground-breaking theoretical developments. However, they do take the step of incorporating current feminist theories of psychological development into the model, addressing the impact of a sexist social organization on individual psychology and dyadic dynamics more explicitly. The propositions put forth above clearly go beyond what is justified by the results of the present study. However, they are more consistent with the results -- especially the lack of significant gender differences in the rates of inflicting and sustaining NsPA. The revisions also result in a model of abuse in intimate relationships which is less dependent upon issues regarding the resource distribution between intimate partners. It does not necessarily follow that resource imbalances will predict couples' conflicts about male dominance of their partners. Variables that would be expected to be more predictive of DA are the two partners' expectations about a man's right to dominate his intimate partner and, more importantly, the mismatch between the two partners' attitudes in this regard. (Note that this suggests that attitudes which pertain specifically to male dominance, 323 rather than the more general set of attitudes included in overall sex-role attitudes are the crucial variables.) Also especially germane would be the two partners' attitudes toward the appropriateness (or inevitability) of using violence in intimate relationships and the frequency with which disagreements about dominance come up. The proposed additions to the model have the advantage of containing specific, explicit predictions concerning dyadic interactions and resulting power distributions in conflictful, and especially violent, intimate relationships between men and women. These predictions can and should be tested in future investigations of dating and domestic violence. In conclusion, in regard to the specific predictions of the current formulation of the Feminist/Social Control Model regarding the role of attitudes, resource distribution and patterns of dominance in dating abuse, the results of this study were mixed. The full set of these variables which were predicted to be associated with inflicting dating abuse did, in fact, significantly differentiate between abusive and non- abusive individuals among this sample of college men and women. In addition, substantial evidence indicated that the occurrence of dating abuse was associated with patterns of male dominance in dating relationships, as the model predicts. However, the model did not differentiate as well between the women who do and do not inflict dating abuse as 324 it did between the different groups of men. Also, a number of important, specific predictions made by the model in regard to gender differences in the rates of NsPA and to resource balance and attitudes were not borne out by the current data. The suggested refinements to the Feminist/Social Control Model are more consistent with the results of this study. Conclusions The results of this study demonstrated very clearly that abuse and violence in dating relationships should be of major concern‘on all college campuses. The prevalence of these experiences, the strength of the reported emotional reactions, the risk of mild and serious injury, and the observed strong relationships between these experiences and a wide range of psychological symptoms and measures of social adjustment all indicate that this is a problem that campus communities cannot afford to ignore. The university and college community should be addressing the issue of dating abuse vigorously. Increasing the awareness of the problem throughout the university community is an important first step which can subsequently lead to concentrated efforts to provide educational and intervention services in residential housing, counseling and health centers. Since late adolescence and early adulthood are times when people are developing relationship skills and learning about intimacy, the college years are an excellent time for intervention. 325 Educational programs could be designed to help young people develop assertiveness and negotiation skills and to explore their unexamined expectations about dominance and control and their attitudes about the appropriateness of violence as a method of dealing with conflict and stress in intimate relationships. In addition to this primary conclusion, this study has resulted in some intriguing results. It will be remembered that the study had three main goals: (a) to explore the benefits of adopting a comprehensive View of dating abuse; (b) tocompare the performance of two different measuring strategies for NsPA, especially as they related to gender differences, and; (c) to assess the relevance of the Feminist/Social Control Theory to the study of dating abuse. In regard to the first goal, the results indicate that sexual, non-sexual and psychological abuse are related phenomenon with important similarities and differences and that dating abuse as a whole is probably more validly conceived of as a combination of these three different types of abuse. It was concluded that, at a minimum, future studies should adopt an expanded view of dating abuse which always includes assessment of pp§p_NsPA and SA. The historical equating of "conflict tactics" with intimate violence is unfortunate and results in an incomplete and inaccurate picture which overlooks the devastation of sexual abuse. 326 In regard to the second goal, that of comparing the performance of the CTS-like measure of NsPA to that of The New Violence Measure, it was found that the latter measure performed better in regard to predicting psychological and social adjustment outcomes, as predicted by the critics of the Conflict Tactic Scale. It was therefore concluded that future investigations of NsPA should employ measures, like The New Violence Measure, which address: (a) the forms of violence, like physical restraint, which are often employed in sexual assaults, and; (b) essential contextual features of the experience, like the occurrence of injuries and perceived dangerousness. Sugarman and Hotaling (1989) have also advocated for such an expanded, contextual perspective in studying dating abuse. Contrary to expectations, however, The New Violence Measure did not reveal the predicted gender differences in the rates of inflicting or sustaining NsPA. A number of possibilities were discussed which might account for this failure, including the unrepresentative nature of the current sample and possible problems with the measurement techniques employed. But finding an absence of such differences is consistent with a number of previous studies. It was concluded that it is probable that women inflict significantly more non-self-defensive, non-sexual, physically abusive behaviors against their dating partners than predicted by the Feminist/Social Control Model. 327 This absence of gender differences in the rates of inflicting and sustaining NsPA, if valid, presents a major challenge to the Feminist/Social Control Model as it is currently formulated. However, results pertaining to the immediate emotional and physical consequences of NsPA and (especially) of SA incidents indicate that women are more severely victimized by these experiences than are men. Also, the rates of inflicting and sustaining SA ala reflect the gender differences predicted by the model. The six discriminant function analyses attempted to assess the relevance of the Feminist/Social Control Model's specific predictions about the relationship of attitudes, resource balance variables, and dominance and power issues to the study of dating abuse. The results of these analyses were mixed. In general, the predictions concerning resource distributions were not supported by this data; the results concerning attitudes were mixed; and those pertaining to power and dominance in decision making were very much in line with the model's predictions. Important gender differences were found. Men's rates of DA were significantly related to: (a) increases in their reported dominance goals; (b) increases in their actual power over their partners' independent decisions, and; (c) decreases in the amount of power they report having over their own independent decisions. For women, only the latter was significantly associated with rates of DA. 328 So, all three forms of abuse were found to be associated with male dominance in decision-making, as predicted by the feminist/social control model. However, the arena of dominance was found to be very specific. Male dominance over more mutual, relationship decisions was not related to DA. (And, as stated above, abusive men reported asserting la§§_ control and independence over their own personal decisions than did non-abusive men.) But, as men report engaging in increasing levels of abusive behavior, they also report increasing levels of influence over their partners' independent decisions. Several recommendations were made concerning additions and/or revisions to The Feminist/Social Control Model to account for the results of the current study, especially in regard to the finding of substantial amounts of non-self- defensive NsPA inflicted by women against their intimate partners. Previously the absence of significant differences in the rates of inflicted and sustained non-sexual violent behaviors between males and females have been used to justify gender neutral models of intimate violence such as that proposed by Riggs and O'Leary (1989), based on Social Learning Theory. Yet the results of this study suggest that this approach is misguided on several counts. First of all, it ignores the role and prevalence of sexual assault as a form of abuse in intimate relationships which clearly victimizes women more often and more severely than men. 329 Secondly, the results of the present study suggest that, even when men and women do not report significantly different levels of involvement in DA, they do describe different emotional and physical consequences and different relationship dynamics associated with increased levels of abuse. It appears that the important differences between men's and women's experiences with NsPA probably lie not so much in the behavioral descriptions of what behaviors occurred during the fight but, rather, in what the man and the woman are fighting about when the incident occurs, and in how the fight alters their interactions in the long run. Of the existing theoretical treatments of intimate violence, only The Feminist/Social Control Model sheds any light on understanding these gender differences. The results of this study indicate that issues of control and power are central to experiences of dating abuse, and, further, that the relationship implications of abuse, in terms of the balance of power, are very different for men and women, as the model predicts. The revisions to the Feminist/Social Conflict Model of intimate violence which were suggested here address these gender differences. At the same time they address the dyadic interactions in a violent relationship and allow for the fact that both men and women inflict non-sexual violence in intimate relationships. Yet, they do so without assuming a "systems" perspective. This latter perspective, adopted by a number of previous authors in response to the finding of 330 equal involvement in inflicting NsPA (Shupe, Stacey & Hazlewood, 1987), tends to equate the victimization experiences of men and women and trivializes battered women's need for supportive services resulting from the severe victimization and domination which they experience in violent relationships. The system perspective also usually entails a recommendation for a dyadic approach to treating intimate violence. But such an approach is misguided and dangerous. Since women and men are not equally victimized by the violence they experience, and since violent relationships are associated with increases in male-dominance with the concurrent loss of power for the female which that implies, the two partners are app on an equal footing in their relationship in terms of their ability to communicate and negotiate openly. Moreover, their conflicts are often "settled” violently behind closed doors, which, while it may endanger and victimize both partners, clearly tends to victimize the woman more severely. Given those conditions, couple counseling is not advisable. It can actually result in the couple experiencing EELS violence by bringing out areas of disagreement and conflict, thereby unnecessarily endangering both partners, especially the woman. Instead, the violence must be addressed first in individual counseling and constructive conflict and stress management habits must be firmly established before any joint sessions are warranted. The results of the current study additionally 331 suggest that special attention must be given to the man's desire for and attempts to dominate his partner -- especially in the area of wanting to control her independent activity and decision-making. It may be especially important, as Dutton & Browning's (1988) work suggests, to focus on the man's equation of his partner's expressions of autonomy with his fear of being abandoned. Of course, before any treatment recommendations based on the results of the present investigation are put into practice, the results should be replicated with other populations and samples. This is especially imperative given that this sample was found to be unrepresentative of the larger campus community. Thus, the extent to which these results can be generalized is limited. The observations and results from this study -- as intriguing as they are -- need to be tested with more representative samples of men and women in dating relationships both within and outside of the college community, as well as with community samples of married and co-habitating couples, and with clinical samples of battered women and their partners. Research with non- clinical samples of couples would be especially revealing. However, one conclusion that is warranted from the present study is that the dating abuse literature, in its overemphasis on comparing papa; of non-sexual physically abusive behaviors, has too often failed to explore the important sex differences which do exist in these experiences and, so, has prematurely and unfortunately dismissed the 332 Feminist/Social control perspective. As discussed in the literature review, while the social learning theory can account for any of the findings that are predicted by the Feminist/Social Control Model, only the latter model makes explicit predictions concerning gender differences. While the results of this study do not support these predictions as they pertain to expectations of different rates of inflicting and sustaining NsPA, they a9 support the predictions of different levels of involvement in SA experiences, different victimization experiences vis a vis emotional and physical consequences, and of the association between involvement in DA and patterns of male dominance in decision making. These differences between men's and women's experiences of NsPA are best understood with reference to the cultural context within which they occur -- which includes social and economic inequality between the sexes, differential gender-role socialization experiences, and widespread expectations of male dominance in intimate relationships. The Feminist/Social Control Model makes important contributions toward understanding the gender differences which these students reported and, with the revisions outlined above, the model can be helpful in guiding further research into these issues. LIST OF REFERENCES LIST OF REFERENCES Adams, D. (1988). Treatment models of men who batter: A profeminist analysis. In K. Yllo & M. Bograd (Eds.), Faminist perspectives on wife abuse (pp.l76-199). Newbury Park, CA: Sage Publications. Albee, G.W. (1986). Toward a just society: Lessons from observations on the primary prevention of psychopathology. Amarican Psychologist, ll, 891—898. Allen, C. M., & Straus, M.A. (1980). Resources, power and husband-wife violence. In M. A. Straus, & G.T. Hotaling (Eds.), The social causes of husbandtyife violence (pp.l88-208). Minneapolis: The University of Minnesota Press. American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington,D.C.: Author. Arias, I., & Beach, S. (1987). Validity of self-reports of marital violence. Journal of Family Violenca, 1, 139-149. Arias, I., Samios, M., & O'Leary, K.D. (1987). Prevalence and correlates of physical aggression during courtship. Journal of Interpersonal Violence, l, 82-90. Bandura, A. (1977). Social learning theory. Englewood Cliffs, New Jersey: Prentice-Hall. Bandura, A. (1983) Psychological mechanisms of aggression. in R. G. Geen & E. I. Donnerstein (Eds.), Aggression: Theoretical and empirical reviews (pp 1 - 40). New York: Academic Press. Becker, C. (1987). The invisible drama: Woman and tha anxiety of change. New York: Macmillan Publishing Company. Belknap, J. (1989). The sexual victimization of unmarried women by nonrelative acquaintances. In M.A. Pirog-Good, & J.E., Stets (Eds.), Violence in dating relationships: Emerqinq social issues (pp. 205-218). New York: Praeger Publications. 333 334 Berk, R. A., Berk, S.F., Loseke, D.R., & Rauma, D. (1983). Mutual combat and other family violence myths. In D. Finkelhor, R.J.Gelles, G.T. Hotaling, & M.A. Straus (Eds.), The dark side of families: Current family violence research (pp. 197-212). Beverly Hills, CA: Sage Publications. Bernard, J.L., Bernard, S.L., & Bernard, M.L. (1985). Courtship violence and sex-typing. Family Relationa, ll, 573-576. Billingham, R.E. (1986, November). Gender differences in the use of conflict tactics across eight levels of emotional commitment. Paper presented at the Annual Meeting of the National Council on Family Relations, Dearborn, Michigan. Billingham, R.E., & Sack, A. R. (1986). Courtship violence and the interactive status of the relationship. Journal of Adolescent Research, l(3), 315-325. Bleier, R. (Ed.). (1988). Feminist approaches to acianca. New York: Pergamon Press. Block, J.H. (1983). Differential premises arising from differential socialization of the sexes: Some conjectures. Child Development, ll, 1335-1354. Bograd, M. (1988). Feminist perspectives on wife abuse: An introduction. In K. Yllo & M. Bograd (Eds.), Feminist perspectivaa,on wife abuse, (pp. 11-26). Newbury Park, CA: Sage Publications. Brodsky, A. M. & Hare-Mustin, R. (1980). Psychotherapy and women: Priorities for research. In A. M. Brodsky & R. Hare-Mustin (Eds.), flgmen andypsychotherapy: An assessment of research and practice (pp. 385-409). New York: The Guilford Press. Brown, B.W. (1980). Wife employment, marital equality and husband-wife violence. In M.A. Straus, & G.T. Hotaling (Eds.), The social causes of husband-wife violence (pp.l76-187). Minneapolis: The University of Minnesota Press. Browning, J. & Dutton, D. (1986). Assessment of wife assault with the Conflict Tactics Scale: Using couple data to quantify the differential reporting effect. Journal of Marriage and the Famil , 18, 375-379. Burgess, A. W., & Holstrom, L.L. (1974). Rape trauma syndrome. American Journal of Psychiatry, 131, 981-986. 335 Burke, P.J., Stets, J.E., & Pirog-Good, M.A. (1988). Gender identity, self-esteem and physical and sexual abuse in dating relationships. Social Psychology Quarterly, ll, 272-285. Burt, M.R. (1980). Cultural myths and supports for rape. Journal of Personality and Social Ps cholo , ll, 217-230. Cate, R.M., Henton, J.M., Koval, J., Christopher, F.S., & Lloyd, S. (1982). Premarital abuse: A social psychological perspective. Journal of Family Issues, 1, 79-90. Check, J.V., & Malamuth, N.M. (1983). Sex role stereotyping and reactions to depictions of stranger vs. acquaintance rape. Jourpal of Personality and Social Ps cholo , 1;, 344-356. Chesler, P. (1972). Women and madness. New York: Doubleday & Company. Chodorow, N. (1978). The reproduction ofymothering: Psychoanalysis and tha:sociology of gander. Berkeley, CA: University of California Press. Deal, J.E. & Wampler, K.S. (1986). Dating violence: The primacy of previous experience. Journal of Social and Personal Relationships, l, 457-471. Deaux, K. (1984). From individual differences to social categories: Analysis of a decade's research on gender. American Psychologist, aa, 105-116. Deaux, K. (1985). Sex and gender. Annual Reviaw of Psychology, aa, 49-81. DeKeseredy, W.S. (1988). Woman abuse in dating relationships: The relevance of social support theory. Journal of Family Violence, 1, 1-13. DeMaris, A. (1987). The efficacy of a spouse abuse model in accounting for courtship violence. Journal of Family Issues, a, 291-305. Derogatis, L.R. & Spencer, P.M. (1982). The Brief Symptom_ Inventory: Administration, scoring and procedures manual. Baltimore: John Hopkins University Press. Dobash, R.E., & Dobash, R.P. (1979). Violence against wives: A case aqainat the patriarchy. New York: The Free Press 336 Dobash, R.P., & Dobash, R.E. (1983). The context specific approach. In D. Finkelhor, R.J.Gelles, G.T. Hotaling, & M.A. Straus (Eds.), The dark side of families: Current family violencafiresearch (pp. 261-276). Beverly Hills, CA: Sage Publications. Dobash, R.P., & Dobash, R.E. (1988). Research as social action: The struggle for battered women. In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse, (pp. 51-74). Newbury Park, CA: Sage Publications. Donelson, E. (1977). Social responsiveness and sense of separateness. In E. Donelson & J.E. Gullahorn (Eds.), Women: A paychological perspective (pp. 140 - 153). New York: John Wiley & Sons. Dutton, D. G. & Browning, J. J. (1988). Power struggles and intimacy anxieties as causative factors of wife assault. In G. W. Russell (Ed.), Violenca in intimate relationahips (163-175). New York: PMA Publishing Corporation. Ellis, E.M. (1983). A review of empirical rape research: Victim reactions and response to treatment. Clinical Psychology Review. 1, 473-490. Emery, B. C. (1987). Factors affecting attitudes toward premarital violence. (Doctoral dissertation, Oregon State University, 1987). Dissertation Abstracts International. (University Microfilms. Order #8724473.) Farrinqton, K.M. (1980). Stress and family violence. In M.A. Straus & G. T. Hotaling (Eds.), The social causes of husband-wife violance (pp. 94-114). Minneapolis: University of Minnesota Press. Ferguson, L.R. (1977). The woman in the family. In E. Donelson & J.E. Gullahorn (Eds.), Women: A psychological perspective (pp. 214 - 227). New York: John Wiley & Sons. Finkelhor, D., Gelles, R., Hotaling, G. & Straus, M. (Eds.). (1983). The dark side of families: Current family violence research. Beverly Hills, CA.: Sage Publications. Finn, J. (1986). The relationship between sex role attitudes and attitudes supporting marital abuse. Sex Roles, ll, 235-244. Firestein, B.A. (1987). The relationship of abuse and social support to the psychological health of college women in dating relationships. Doctoral dissertation, University of Texas at Austin. 337 Firestein, B.A. (1989). The relationship of abuse and social support to the psychological health of college women in dating relationships. Paper presented at the Annual Conference of the Association for Women in Psychology. New Port, R.I. Fischer, G. J. (1986). College student attitudes toward forcible date rape: I. Cognitive predictors. Archives of Sexual Behavior, lg, 457 - 466. Foss, J.E. (1980). The paradoxical nature of family relationships and family conflict. In M.A. Straus & G. T. Hotaling (Eds.), The social causes of husband-wife violence (pp. 94-114). Minneapolis: University of Minnesota Press. Foucault, M. (1972). The archaaology of knowledga. New York: Pantheon Books. Friedan, B. (1963). The feminine mystigue. New York: Dell Publishing Company. Frieze, I.H., Parsons, J.E., Johnson, P.B., Ruble, D.N., & Zellman, G. L. (1978). Women and sex roles: A social psychological perapectiva. New York: W.W. Norton & Co. Ganley, N. (1981). Participants' manual: Court mandated couseling for man who batter. Washington, D.C.: Center for Women Policy Studies. Gelles, R. G. & Straus, M.A. (1988). Intimate violence: The definitive_atudy of the causas and consequences of abuse in the American family. New York: Simon & Schuster. Gilligan, C. (1982). In a diffarent voice. Cambridge, Mass.: Harvard University Press. Gondolf, E.W. (1988a). Research on men who batter: An overview, bibliography and resource guide. Bradenton, Fla.: Human Services Institute. Gondolf, E.W. (1988b). The state of the debate: A review essay on woman battering. Response to the Victimization of Women and children. ll(3), 3-8. Goodchilds, J.D. & Zellman, G. L. (1984). Sexual signaling and sexual aggression in adolescent relationships. In N.M. Malamuth & E. Donnerstein (Eds.) Pornography and sexual aggression (pp. 233-243). Orlando, Fla.: Academic Press, Inc. 338 Gwartney-Gibbs, P.A. & Stockard, J (1989). Courtship aggression and mixed-sex peer groups. In M. A. Pirog-Good & J. E. Stets (Eds.) Violence in dating relapionahips: Emerging apcial issues (pp. 185-204). New York: Praeger Publishers. Gwartney-Gibbs, P.A., Stockard, J. & Bohmer, S. (1987). Learning courtship aggression: The influence of parents, peers and personal experiences. Family Relations, 36, 276- 282. Hare-Mustin, R.T. & Maracek, J. (1988). The meaning of difference: Gender theory, postmodernism & psychology. American Psychologist, ll, 455—464. Hare-Mustin, R.T. & Maracek, J. (1989). Thinking about postmodernism and gender theory. American Ps cholo ist, AA, 1333-1334. Heller, K., Price, R. H., Reinharz, S., Riger, S., & Wandersman, A. (1984). Psychology and community change (2nd ed.). Homewood, Ill.: Dorsey Press. Henton, J., Cate, R., Koval, J., Lloyd, 8. & Christopher, S. (1983). Romance and violence in dating relationships. Journal of Family Isaues. A, 467-482. Hoffman, P. (1984). Psychological abuse of women by spouses and live-in lovers. Women and Thera , A, 37-47. Jackson, D. N. (1970). A sequential system for personality scale development. In C. Spielberger (Ed.), Current topics in clinical and community psychology (pp. 61-95). New York: Academic Press. Jouriles, E. N. & O'Leary, K.D. (1985). Interspousal reliability of reports of marital violence. Journal of Consulting and Clinical Psychology. él, 419-421. Kanin, E. J. (1984). Date rape: Unofficial criminals and victims. Victimology: An International Journal, a, 95- 108. Kelly, L. (1988). How women define their experiences of violence. In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse, (pp. 114-132). Newbury Park, CA: Sage Publications. Kennedy, F. (1970). Institutionalized oppression vs. the female. In R. Morgan (Ed.), Sisterhood is powerful (pp 492 - 501). New York: Random House. 339 Koss, M.P. (1985). The hidden rape victim: Personality, attitudinal and situational characteristics. Psychology of Women Quarterly, 2, 193-212. Koss, M.P., & Burkhart, B.R. (1989). A conceptual analysis of rape victimization: Long—term effects and implications for treatment. gaychology of Women Quarterly, ll, 27-40. Koss, M.P., Dinero, T.E., Seibel, C.A., & Cox, S.L. (1988). Stranger and acqaintance rape: Are there differences in the victim's experience? gaychology of Women Quarterly, ll, l-24. Koss, M.P., Gidycz, C.A., & Wisniewski, N. (1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Ps cholo , pa, 162-170. Koss, M.P. & Leonard, K.E. (1984). Sexually aggressive men: Empirical findings and theoreticql implications. In N.M. Malamuth, & E. Donnerstein (Eds.), Pornography and sexual aggression (pp. 213-232). Orlando, Fla.: Academic Press, Inc. Koss, M.P. & Oros, C.J. (1982). Sexual Experiences Survey: A research instrument investigating sexual aggression and victimization. Journal of Consulting and Clinical Psychology, pg, 455-457. Kuhn, R.S. (1970). The structure of sciantific ravolutions (2nd ed.). Chicago: University of Chicago Press. Lane, K.E. & Gwartney-Gibbs, P.A. (1985). Violence in the context of dating and sex. Journal of Family Iaaues. 5, 45-59. Lloyd, S.A., Koval, J.E., & Cate, R.M. (1989). Conflict and violence in dating relationships. In M.A. Pirog-Good, & J.E., Stets (Eds.), Violence in dating relationships: Emerging social issues (pp. 126-142). New York: Praeger Publications. Makepeace, J.M. (1981). Courtship violence among college students. Family Relations, 19, 97-102. Makepeace, J.M. (1983). Life events stress and courtship violence. Famil Relations, 1;, 101-109. Makepeace, J.M. (1986). Gender differences in courtship violence victimization. Famil Relations, 1;, 383-388. 340 Makepeace, J.M. (1989). Dating, living together and courtship violence. In M.A. Pirog-Good, & J.E., Stets (Eds.), Violence in dating relationships: Emerging social issues (pp. 94-107). New York: Praeger Publications. Malamuth, N.M. (1984). Aggression against women: cultural and individual causes. In N.M. Malamuth & E. Donnerstein (Eds.), Pornography and sexual aggression. Orlando, Fla.: Academic Press, Inc. Malamuth, N.M. (1986). Predictions of naturalistic sexual aggression. Journal of Personality and Social Psychology, pp, 953-962. Marshall, L.L., & Rose, P. (1988). Family of origin violence and courtship abuse. Journal of Counseling and Development, aa, 414-418. McCann, I.L., Sakheim, K.S., Abrahamson, D.J. (1988). Trauma i and victimization: A model of psychological adaptation. The Counseling Psychologist, lg, 531-594. . Miller, J.B. (1976). Toward a new psychology of women. E Boston: Beacon Press. Mowbray, C.T., Lanir, S., & Hulce, M. (Eds.). (1986). Women and mental health: New directions for change. New York: The Haworth Press. Muehlenhard, C.L. & Linton, M.A. (1987). Date rape and sexual aggression in dating situations: Incidence and risk factors. Journal of Counseling Psychology, AA, 186-196. Murphy, J.E. (1988). Date abuse and forced intercourse among college students. In G.T. Hotaling, D. Finkelhor, J.T. Kirkpatrick, & M.A. Straus (Eds.), Family abuse and its consequences: New directions in research (pp. 285-296). Newbury Park, CA: Sage Publications. Murphy, S.M., Amick-McMullan, A.E., Kilpatrick, D.G., Haskett, M.E., Veronen, L.J., Best, C.L., & Saunders, B.E. (1988). Rape victims' self—esteem: A longitudinal analysis. Jougpal of Inter ersonal Violence, A, 355-370. Nadelson, C., Notman, M., Jackson, H., & Gornick, J. (1982). A follow-up study of rape victims. American Journal of Béyghiatry. 133, 408—413. NiCarthy, G. (1982). Getting free: A handbook for women in abusive relationships. Seattle, Washington: The Seal Press. 341 Norusis, M.J. (1988). SPSS/PC+ advanced statistics (Version 2.0.). Chicago: SPSS, Inc. Okun, L. (1986). Woman abuse: Facts replacing myths. Albany, NY: State University of New York Press. Pagelow, M.D. (1981). Woman-battering: Victims and their experiences. Beverly Hills, CA: Sage Publications Pirog-Good, M.A., & Stets, J.E. (Eds.). (1989). Violence in dating relationships: Emergingysocial issues. New York: Praeger Publishers. Polk, B.B. (1974). Male power and the women's movement. The Journal of Applied Behavioral Science, lg, 415-431. ‘06 Ptacek, J. (1988). Why do men batter their wives? In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse, (pp. 114-132). Newbury Park, CA: Sage Publications. .r AX Rappaport, J. (1977). Community psychology: Values, research & action. New York: Holt, Rinehart & Winston. L Rappaport, J. (1981). In praise of paradox: A social policy of empowerment over prevention. American Journal of Community Paychology, a, 1-25. Riggs, D.S., & O'Leary, K.D. (1989). A theoretical model of courtship aggression. In M.A. Pirog-Good, & J.E., Stets (Eds.), Violence in dating relationships: Emerging socyal issues (pp. 53—71). New York: Praeger Publications. Roscoe, B. & Benaske, N. (1985). Courtship violence experienced by abused wives: Similarities in patterns of abuse. Family Relations, l1, 419-424. Rouse, L.P., Breen, R., & Howell, M. (1988). Abuse in intimate relationships: A comparison of married and dating college students. Journal of Interpersonal Violenca, A, 414-429. Roy, M. (Ed.). (1977). Battered women: A paychoaociological atudy of domestic violence. New York: Van Nostrand Reinhold Company. Russell, D.H. (1982). Rape in marriage. New York: Macmillan Publishing Co. Russell, D.H. (1984). Sexual exploitation: Rape, child sexual abuse and workplace harassment. Beverly Hills, CA: Sage Publications. 342 Russell, D.H. (1986). The secret trauma: Incest in the lives of girls and women. New York: Basic Books. Russianoff, P. (1981). Why do I think I am nothing yithout a man? New York: Bantam Books. Ryan, W. (1976). Blaming the victim (2nd ed.). New York: Random House. Sanford, L.T. & Donovan, M.E. (1984). Women and self-esteem. Garden City, New York: Anchor Press/Doubleday. Sarason, S. (1983). Psychology and public policy: Missed opportunity. In R.D. Felner, L.A. Jason, J.N. Moritsugu, & S.S. Farber (Eds.), Preventive psychology: Theory, research é practice (pp. 245-250). New York: Pergamon Press. Saunders, D.G. (1988). Wife-abuse, husband abuse, or mutual combat? A feminist perspective on the empirical findings. In K. Yllo & M. Bograd (Eds.), Eaminist perspectives on wifegabuse, (pp. 90-113). Newbury Park, CA: Sage Publications. Shupe, A., Stacey, W.A., & Hazlewood, L.R. (1987). Violent men, violent couples: The dymanics of dpmestic violence. Lexington, Massachusetts: Lexington Books. Sigelman, C.K., Berry, C.J., Wiles, K.A. (1984). Violence in college students' dating relationships. Journal of Applied Social Pscholoqy, é, 530-548. Sonkin, D.J., Martin, D., & Walker, L.E. (1985). The male battarer: A treatment approach. New York: Springer Publishing Co. Spence, J.T., Helmreich, R.L., & Holahan, C.K. (1979). Negative and positive components of psychological masculinity and femininity and their relationship to self- reports of neurotic and acting out behaviors. Journal of Personality and Social Psychology, ll(lo), 1673-1682. Starhawk (1987). Truth or dare: Encounters withpowerI authority and mystery. San Francisco: Harper and Row. Stark, E., Flitcraft, A., & Frazier, W. (1979). Medicine and patriarchal violence: The social construction of a ‘private' event. International Journal of Health Services, a, 461-493. Stets, J.E., & Pirog—Good, M.A. (1987). Violence in dating relationships. Social Psychology Quarterly, a9, 237-246. 343 Stets, J.E., & Straus, M.A. (1989). The marriage license as a hitting license: A comparison of assaults in dating, cohabiting and married couples. In M.A. Pirog-Good, & J.E., Stets (Eds.), Violence in dating relationships: Emerging social issues (pp. 33-52). New York: Praeger Publications. Straus, M.A. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics Scale. Journal of Marriage and the Family, Al, 75-88. ' Straus, M.A. (1980a). A sociological perspective on the prevention of wife beating. In M.A. Straus & G. T. Hotaling (Eds.), The social causes of husband-wife violence (pp. 211-232). Minneapolis: University of Minnesota Press. Straus, M.A. (1980b). Sexual inequality and wife beating. In M.A. Straus & G. T. Hotaling (Eds.), Ema social causea of husband-wife violence (pp. 86-93). Minneapolis: University of Minnesota Press. Straus, M.A. (1980c). Wife-beating: How common and why? In M.A. Straus 8 G. T. Hotaling (Eds.), The social causes of husband-wife violence (pp. 23-36). Minneapolis: University of Minnesota Press. Straus, M., Gelles, R. & Steinmetz, S. (1980). Behind closed doors: Violence in the American family. Garden City, New York: Doubleday Publishing Co. Sugarman, D.B., & Hotaling, G.T. (1989). Dating violence: Prevalence, context and risk markers. In M.A. Pirog-Good, & J.E., Stets (Eds.), Violence in dating relationshlps: Emerging social issues (pp. 3-32). New York: Praeger Publications. Walker, L.E. (1979). The battered woman. New York: Harper & Row Publishers. Walker, L.E. (1984). The battered woman syndrome. New York: Springer Publishers. Walker, L.E. (1985). What counselors should know about battered women. In D. J. Sonkin, D. Martin, & L.E. Walker (Eds.),The male batterer: A treatment approach (pp 150- 165). New York: Springer Publishers. Walker, L.E. (1989). Psychology and violence against women. American Psychologist, 11, 695—702. 344 Warshaw, R. (1988). I never called it rape: The Ms. report on recognizing, fighting and surviving date and acquaintance rape. New York: Harper & Row Publishers. Weisstein, N. (1971). Psychology constructs the female. In V. Gornick & B. K. Moran (Eds.), Women in sexist society: Studies in power and powerlessness (pp. 207-224). New York: Basic Books. Yllo, K. (1988). Political and methodological debates in wife abuse research. In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse, (pp. 28-50). Newbury Park, CA: Sage Publications. Yllo, K., & Bograd, M. (Eds.) (1988). Feminist perspectives on wife abuse. Newbury Park, CA: Sage Publications. APPENDIX A 23115 RELATIONSHIP SURV Y 1 20M01u.'.’°u?80000.0.0.0088.0.0.09...OOOOOOOOOOOOOOOOOOOOOO O“.‘i"W".‘?(“r...l.'2tnal.)oO880088.00.08I00.000000000000008... 3. Nhot is your pain racial or ethnic background? 1 8 Asian, Pacific Islander 6 8 Native Asorican 2 8 Black . _ , 5 8 Uhiie, Caucasian 3 8 hoxican-Asoricon, Chicano, Latino 6 8 Ulhor (please specify)..----.-_.._._.. . . . . . . . . . 6. Nhal your are you in college? (I8Junior, 28Sonior). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Uhoro are you living nov? 1 8 sorority/fraternity 6 8 at hose vith parents 2 8 co-op 5 8 dors 3 8 private aparteeni or house 011 calpus 6 8 other (please specify) . . . . . . . . 6. Nhal is your host eslisalo 01 your parents' coshined annual incoeo over the past 12 eonlhs? [8y 'psrenls’ vo seen the ea 0 and/or locale adultls) that you lived vilh during cost 01 your high school 1 8 Less than 610 000 6 8 630,001 - 60,000 7 8 660,001 - 70,000 2 . 910,001 - 20,000 5 . s10,001 - 50,000 a . 170,001 - 00,000 3 8 620,001 - 30,000 6 8 650,001 - 60,000 9 8 680,001 - 90 000 corresponding 10 the . . -__-(1) . . ..__-(2) . . _----(3) . . -____(6) O I mn‘s) years.) 10 . Over 190,000. . . . . . . . . . . . (6) 0n the next page is I list 01 rohless and cos lainls that so 10 soeolieos have. lead each one carolulll‘ond select one 01 the nueberod descriptors that as) describes N NUCN DISCO 0 1 THAT PROBLEA NAS CAUSED YOU DURING THE 5) MONTH INCLUDING TODAY. Enter the nuebor corresponding to the 60st ansvor in the spaces at the 1011 01 each 1100. Do poi slip any itees. 1 8 Not at all, 2 8 A little bit 3 8 Moderately 6 8 00110 a 611 5 8 Exlroecly In th 346 I 8 Hot at all 2 8 A little bit 5 3 8 Hoderately e past aonth, hov auch vere you distressed by... __-7. lervousness or shakiness inside. ”I. Faintness or dizziness.- 9. Ihe hidea that soaeone else can control your ”10. Feeling; others are to blaae for east of your troub es _._11. Trouble reaeabering things. 12. Feeling easily annoyed or irritated. 13. Pains in the heart or chest. 16. Feeling afraid in open spaces. 15. Thoughts of ending your life. __.-I6. Feeling that aost people cannot be trusted. -.._I7. Poor appetite. ._.__ID. Suddenly scared for no reason. 19. Tenper outbursts that you could not control. ._-20. Feeling lonely even vhen you are with people. 21. Feeling blocked in getting things done. ..-22. Feeling lonely. 23. 26. 25. 26. Feeling blue. Feeling no interest in things. Feeling fearful. four feelings being easily hurt. Feeling that people are unfriendly or dislike you. Feeling inferior to others. Hausea or upset stoaach. M30. Feeling that you are vatched or talked about by other rs. M31. Trouble falling asleep. 32. Having to check and double check vhat you do. ,,..,33. Difficulty aaking decisions. 36. Feeling afraid to travel on buses, subvays, or trains. 6 8 Duite a bit 8 Estreaely 35. Trouble getting your breath. 36. Hot or cold spells. M37. Having to avoid certain things places, or activities because they frighten you. _33. Your aind going blank. 39. Huabness or tingling in parts of your body. ....-60. The idea that you should be punished for your sins. .._--61. Feeling hopeless about the future. 62. Trouble concentrating. ...__63. Feeling veat in parts of your body. 66. Feeling tense or keyed up. 65. Thoughts of death or dying. 66. Having urges to beat, injure, or hara soaeone. 67. Having urges to break or saash things. _63. Feeling very self-conscious vith others. ..__69. Feeling uneasy in crovds. 5D. lever feeling close to another person. 51. Spells of terror or panic. 52. Getting into frequent arguaents. ~53. Feeling nervous uhen you are left alone. ”56. Others not giving you proper credit for your achieveaents. -..55. Feeling so restless you couldn't sit still. M56. Feelings of uorthlessness. ”57. FeelingH that people uill take advantage of you if et then. ”59. Feelings of guilt. 59. The idea that soaething is urong vith your aind. 347 For the following questions, lease lace the nunber corresponding with the best answer in the space to the right of each itea. (For exanp e, l 8 not a all, 8 noderately, 5 8 very nuch, etc.) 60. How such do you think you are able to trust others?. . . . . . lot at all I. . 2. . 3. . 6. . 5 Very such ___.-(60) 61. How nuch do you think you are you able to nake (riends?. . . . lot at all I. . 2. . 3. . 6. . 5 Very auch .__-_(61) 62. Hov nuch do you think you are able to get close to others?. . .Hot at all 1. . 2. . 3. . 6. . 5 Very such _.--_(62) 63. How nuch do you think you are able to aaintain relationships?. Hot at all I. . 2. . 3. . 6. . 5 Very nuch .._--(63l Indicate, below how physically satisfying you have found the following activities during the past nonth: I 8 I don't engage in this activity 6 8 Duite satisfying 2 8 lot satisfying . 5 8 Very satisfying 3 8 Sonewhat satisfying ‘4I ‘i‘sing and huggingI I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I- (6‘) ‘SI "“in' .n‘ ,trotingI I I I I I I I I I-I I I I I I I I I I I I I I I I I I I I I I I I I ...-(‘5) SSI SIIUII inII"OUFI.n I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I (66} Friendship; and Dating Relationships If you are 26 years old or younger, answer question 667. If you are 25 years old or older, answer 668. a particular tiae - either prearranged or not - with a person w ou were 'interested in ei her roaantica or sexually, or who you think was interested in you in these wa s. you haven't been on a 'date' like this during your college years, skip to question 6296. If you have been, go 0 question 669.) 67. Have you been on a 'date' with anyone since you've been in college? (6y 'date' we seen, arran ing or agreeigg to spend 0 I l I 8 yes, I have been on a 'date', as defined here; 2 8 no, I have not been on a date. . . . . . . . . . . .__.._(67) 63. Have ou been on any kind of date with anyone during the ast four years? (See the definition of ‘date' in question 667. If you haven t been on a 'date' Ii e this during t e last four years, skip to question 6296. I 8 yes, I bave been on a 'date', as defined here; 2 8 no, I have not been on a date. . . . . . . . . . . .__.._(6B) 69. Hhich description best reflects your ggrrggt dating relationship(s)? 1 8 narried 5 8 dating this person nore than others 2 8 engaged 6 8 dating this one person - but very casually 3 8 living together 7 8 dating several people at the sane tine - none nore than the otherfs) 6 8 dating steadily this one person I 8 not dating anyone currently. . . . . . . . . . . . . . . . . . . The questions on the followin page ask you about es eriences ou nay have had with a dating partner durin your college years (or during the past our years, if you're . or older . Ho natter hou well a couple gets along, t ere are tines when the disagree, get annoyed with each 0 her or Just have spats or fights because they're in a bad aood or are tired. They a so have nany different ways of handling argunents and trying to settle their di ferences. below are soae things that you night have experienced with a recent ating partner when you were aczry at each other. Please indicate hou i e y often you've ex erienced any of these things 'n an datin relationshi you've had ou were attending college (not necessarily jus UCD) or during the past 6 ears. Ihey say be thin s you have done or thin s one of your ating partners has done during an argunent. t doesn't na ter who did it, gust in icate how often you've ad any of these experiences. Take a ninute to bring to nind the dating relationships you've had during college, (or in the past 6 years, if you’re over 26). Think about the kinds of interactions you had in t ose relationships during argunents, disagreenents or fights. 348 How often in any of your college dating relationships have either you or a partner done any of the following in an argunent? (Please place the nuaber which corresponds to the appropriate answer in the space on the left.) I 8 never 6 8 six to ten tines 2 8 once or twice 5 8 nore than ten tines 3 8 three to five tines _._70. Calaly discussed the problea and arrived at a solution that helped? .____71. Insulted or swore at the other? ____.72. Stoaped out of the roon or house or yard? ___..73. Cried? ...._76. Expressed anger physically by throwing sonething at or hitting the other in soae way? ._..-75. Grabbed, shoved, pushed, or physically restrained the other? 76. Expressed anger physically toward the other in soae other way? The following uestions ask about situations pertaining to sexual activity in dating and courtship. At this_point we would like to renin you that this questionnaire is confiletely voluntary and any questions below which are obyectionable to you nay be skipped. e would also like to reaind you t at this questionnaire is conpletely anonyaous; no one will ever be able to link your answers to your nane. Again, take a ninute to brin to nind the dating relationships you've had while in college (or during the past four years, if ou're over 26). How 0 ten in any of these relationships ave you had any of the following experiences with a dgting gar ner? (If ou have had any of the experiences described in itens 679 through_91 with a stranger or an acquain ance, ut no with a dating partner, according to the definition of a 'date' given above, then put an '6' here..._._-. Do not include those experiences in your answers below.) Hhich of the following experiences have you had with a dating partner since you've been in college for in the last four years, if you're over 26). Have you ever?... I 8 Yes 2 8 Ho 77. engaged in kissing, hugging or fondling when both you and a partner wanted to? 7 . had sexual intercourse when both you and a partner wanted to? 79. had sexual contact with a partner even thou h one 06 you didn't really uant to but did so because they were threatened, pressured, or coerced by the 0 her partner in soae way? 60. had sexual contact with a {artner even though one of you didn't really uant to but did so because the other person threatened to use physica force otherwise. _....31. had sexual contact with a partner even though one of you didn't really want to but did so because the other person actually did use physical force? How we’d like you to go back over your answers to the nrevious questions. If you answered that you had any of the exieriences in questions 676 throu h 76 or 679 throug ll ( lease check these nunbers carefully , then answnr the to lowing questions about the rela ions i: in which you had hose exteriences, If you reported experiences in response to those questions which hapiened with aore han one partner, focus on he relationship where you had the nost such experiences. (If that sti l describes nore than one FOIIIIOflShIg choose the nost recent one where you FIB—the highest nuaber of these experiences.) If you reported no incidents in _hese itens (76-76, 6 79-31), then describe your current or nost recent dating relationship. 11 of the relationship questions in the reaainder of this questionnaire uill ask you to t sh' . Please take a ninute now to think about this relationship, and the kinds of focus on th s one datin r interactions you a or ave wi is dating partner, in as such detail as you can. I I 32. Uhen did this relationship start - uben, approxinately, use your first date? (daylnonthlyear) _--I___I_._. D3. Hhen did this relationship end - when approxinately, did you last see this person? (daylnonthlyear, or leave blank if still involved with his person) ..... 66. Here you a student at Ut Ierkeley during any 01 the tine you were dating this person? (I 8 yes, 2 8 no) 349 On this page, please enter the nuaber for the appropriate answers in the spaces on the right. 35. Vhat is your best estiaate of 12!; approxinate grade point average during this relationship?. . . . . . . . . -_-_-(65) 36. Hhat is your best astinate of your average annual incone during this relationshi . (Include incone tron work, financial aid scholarships, noney froa parents or others, and all 0 her sources of incone. Also include noney paid by others toward your rooa, board, tuition or other educational expenses.) i . 0 - 05,000 0 - 011,001 - 10,000 - 7 - 020,001 - 23,000 2 . 05,001 - a 000 s . 010,001 - 17,000 0 . 023,001 - 30,000 3-0mwi-dmw 080nmm-2mwo 080mmm-00w0 10 . Over 000,000. . . . . . . . . . . “mum 67. How nuch noney, on the average, would you say you had available each nonth for entertainaent, spending noney and extras during this relationship? I 8 0 8 610 6 8 651 8 75 7 8 6126 8 150 2 8 611 8 25 5 8 676 8 100 I 8 6151 8 200 3 8 626 8 50 6 8 6101 8 125 9 8 6201 8 300 ‘o‘m."300IIIIIIIIIIIII-..“(B?) 33. Your age at the end of this relationship, or now, if still dating. . . . . . . . . . . . . . . . . . . . . . . (85) 99. Last grade in school that you had conpleted at the beginning of your relationship I 8 less than high school graduation 6 8 second year of college 2 8 high school raduation 5 8 third year of college _ 3 8 first year 0 college 6 8 other (please speci y) . . . -__--(897 For the following questions choose the nuaber that best represents how you would have rated yourself on each of these attributes during the tine Ihat you were involved in this relationship. Enter that nuaber in the space to the right. 90. Hy physical health was: . . . . . . . . . . . Poor 1. . . .2. . . .3. . . .6. . . .5 Excellent . . . . . . ._.-_(90) 91. Hy self-confidence was: . . . . . . . . . .Very Low 1. . . .2. . . .3. . . .6. . . .5 Very High . . . . . . (91) (92) (93) 92. Hy social life was: . . . . . . . . . Very Inactive I. . . .2. . . .3. . . .6. . .5 Very Active . . . . 93. Physically, I was: . . . . . . . .Very Unattractive l. . . .2. . . .3. . . .6. . .5 Very Attractive . . The questions below ask for soae inforaation about this dating partner. 94ePIFtnIV'III‘SI‘rIIIII 28”.].IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII...”(9‘) 95. Partner's age at the end of this relationship, or now, if still dating. . . . . . . . . . . . . . . . . . . . .._--(95) 96. Last grade in school that this partner had conpleted at the beginning of your relationship I 8 less than high school graduation 5 8 third year of college 2 8 high school raduation 6 8 gradua ed froa college 3 8 first year 0 college 7 8 soae graduate school 6 8 second year of college 6 8 conpleted a graduate degree. . . . . . . . . . . .--__-(96) 97. Has this partner a student at UC Berkeley at any tine while you were dating? 1 8 yes 2 8 no. . . . . . . (97) 98. If a student, what is your best estiaate of this partner's approxinate grade point average during Ihis'.l.‘lm‘hlp?cIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII“.--(98) 99. Hhat is our best estiaate of this dating‘fiartner's average annual incone during this relationship. (Include incona roe work, financial aid, scholar ips noney fron parents or others and all other sources of incone. Also include noney paid by others toward his/her roon, board, tuition or otIer educational expenses.) 1-0-0mmo 0-0nmm-inw0 7-nmwi-nmm 2=umm-aow s=0umm-1Lm0 38uLWl-wa 3 . 00,001 - 11,000 5 . 017,001 - 20,000 9 . 030,001 - 40 000 10 . Over 040,000. . . . . . . . . . . ,_,,,199) 235() Please place the nuaber which corresponds to the appropriate answer in the space on the left. 100. How nuch noney, on the average, would you say this dating partner had available each nonth for entertainnent, spending noney and extras during this relationship? 1 8 0 8 610 6 8 651 8 75 7 8 6126 8 150 2 8 611 8 25 5 8 676 8 100 I 8 6151 8 200 3 8 626 8 50 6 8 6101 8 125 9 8 6201 8 300 10 8 Over 6300 .,___101. Vhat is your best estiaate of this partner's parents' conbined annual incone while you were dating? (Dy 'garents' we nean the sale and/or feaale adult(s) that helshe lived with during nost of high school.l Even hough you probably won't know, exactly, what is your best guess? i 8 Less than 010 000 0 8 030,001 - 60,000 7 8 000,001 - 70,000 2 8 010,001 - 20,000 5 8 000,001 - 50,000 a 8 070,001 - 00,000 3 8 020,001 - 30,000 0 8 050,001 - 00,000 3 8 000,001 - 90 000 10 . om 090,000 For the following questions, choose the nuaber that best re resents how you would rate this dating partner on r each of these at ibutes during the tine that you were invo ved in this relationship. Enter that nuaber in the space to the left. (For exaaple, 1 8 poor, 3 8 noderate, 5 8 excellent.) 102. 103. 106. 105. 106. 107 I. 0 CD I 111. His/her physical health was: . . . . . . . . . . . Poor 1. . . .2. . . .3. . . .6. . . .5 Excellent His/her self-confidence was: . . . . . . . . . .Very Low 1. . . .2. . . .3. . . .6. . . .5 Very High His/her social life was: . . . . . . . . . Very Inactive I. . . .2. . . .3. . . .6. . . .5 Very Active Physically, helshe was: . . . . . . . .Very Unattractive I. . . .2. . . .3. . . .6. . . .5 Very Attractive Peo le differ in how the view aen’s and wonen's l'proper' roles in societ . How strongly would you say this par ner naintained tradiiional stereotypes about nen and wonen and the diIferences between then? Helshe strongly disagrees Helshe strongly agrees with traditional stereotypes I. . . .2. . . .3. . . .6. . . .5 with traditional stereotypes In general, how such power or influence did your partner think helshe should have over the decisions the two of you nade? Helshe thought I should . Helshe thought they should alnost alnost always sake the decisions I. . . 2. . . 3. . . 6. . . 5 always sake the decisions Vhich of the following best describes this relationship? I 8 I only dated this person once . 2 8 an or was dating several ieople at the sane tine - none aore than the other(s), this was one of hose relationshi s. or was dating only this one person - bu vecz casually or was dating this person nore than others 0 I dated at the sane tiae a or was dating this person axclusively or uas engaged to narry this person or an an an an was narried to this person I 3 8 I 6 8 I 5 8 I a 6 8 I 7 8 I How such of your tree tine would you say you spent (or spend) with this person in an average week during the tine you were dating? 1 8 saw then only once 6 8 saw then several tines a week 2 8 saw then less than once a week 5 8 saw then alaost every day 3 8 sew then approxinately once a week 6 8 saw then daily How often would you say you and this partner had (or have) argunents, fights or heated disagreenents? l 8 never 2 8 rarely 6 8 often 3 8 sonetines 5 8 alnost every tine we were together How satisfied would you say you are (or were) with this relationship, on the whole? I 8 ver dissatisfied 6 8 aostly satisfied 2 8 nos 1 _616666156166 5 8 very satisfied 3 8 satis ied with sone things, dissatisfied with others 351 On this page, please enter the nuaber for the appropriate answers in the spaces on the right. 112. Do you think this relationship was or is 'good' for you? (i.e. does it help you to bring out your best qualities?) 1 8 it is ver bad for ne 2 8 it is nos l bad for ae 6 8 it is aostly good for ne 3 8 sonetiaes i is good for ne, 5 8 it is very good for ae. . . . . . . . . . . . . . . . ...__(112) and soaetiaes bad 113. How nuch fun and enjoynent did you or do you get froa this relationship? I 8 none at all 6 8 quite a bit 2 . V", li‘tl. s . l "..I ‘.alI I I I I I I I I I I I I I I I I I I I I I I I I I (:13) 3 8 soae 88888 Please answer the following questions, as ou think this datin artner would have answered then when th tw 01 ou were atin . Use the following scale to indica e ow s rong y you in s a enent. Enter the appropriate nuaber in the spaces to he right. Strongly disagree I. . . .2. . . .3. . . .6. . . .5. . . .6. . . .7 Strongly agree 116. A nan should fight when the wonan he’s with is insulted by another nan. . . . . . . . . . . . . . . . . . . .._(ll6) __(115) 116. A wonan should be a virgin when she narries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ._, _.(ll6) 115. It is acceptable for the wonan to pay for the date. .6. . . . . . . . . . . . . . . . . . . . . . . . . . . 117. There is sonething wrong with a wonan who doesn't want to narry and raise a fanily. . . . . . . . . . . . . _.._-(117) 116. A wife should never contradict her husband in public. . . . . . . . . . . . . . . . . . . . . . . . . . . . __.__(118) _..(119i 120. It is acceptable for a wonan to have a career, but aarriage and fanily should cone first. . . . . . . . . . _..__(120) 119. It is better for a wonan to use her feninine charn to get what she wants rather than ask for it outright. . 121. It looks worse for a wonan to be drunk than for a nan to be drunk. . . . . . . . . . . . . . . . . . . . . .....-(l211 122. There is nothing wrong with a wonan going to a bar alone. . . . . . . . . . . . . . . . . . . . . . . . . . __.__(122) If you reported no incidents in itens 676 through 76 then skip to question 6 171, otherwise continue on Iron here. The questions on the next pa e ask you nore about the specific interactions that you and this partner had when you were angry at each other. Indita e how often our rtn r did each of these things when you and helshe were having an. argunen . Also, (or each behavior listed indica e ov angerous you feel this action was by indicating how such {hysical here it ould have caused you. It doesn't natter, in rating the dan erousness of an act, whether you were, 1n_fac in ured or hurt. onsider instead what the consequences of this ac ion nigh have been if you'd experienced the full 0 act 0 it. Also, indicate in the stares on the far right how nany tines you actually very inygrgg in any way by each type 0 action. Leave blank any spaces hat do not apply because the ehavior never occurre . 352 How often, in argunents, did this dating partner... 123. 126. 125. 126. 127. 126. I29 I30. I31. I32. I33. 136. I35. 136. I37. .____I33. How severe were the worst injuries you reported above? How den erous was the nost serious incident of this be avior? (Use the ratings at the bottoa of How nany tines? the colunn.) Throwsonethinggtyou?......_.__............... Pushorshoveyou?.........._.____............. Crab you (not around the neck)?. . ._____.. . . . . . . . . . . . . Slapyou?..............._-_............. Physically restrain or hold you forcibly and keep you fron noving?. . . . . . . . Pull your hair?. . . . . . . . . . _____... . . . . . . Twist or bend your are, hand, or fingers?. . . . . . . . . . . . -__..--. . . . . . . Kick you?. . . . . . . . . . . . . ._.._._. . . . . . . Iiteyou?.............__”__....... Hit you with a fist?. . . . . . . ._.._-__. . . . . . . :66e6616'z to hit you with g.. . . . . . . . . . . . . _______ . . . . . . . beat you up?. . . . . . . . . . . .--...__. . . . . . . 32':.?I??'f”.’?“.'f°‘.'"?. . . . . ,,,,,,, . . . . . . . 131113-930"??? 3 Pi”? . . . . .. . . Use a knife or fire a gun?. . . . .--.----. . . . . . . gangergusness Ratings: 1 8 Could ave cause no si 1 a'n or barn 2 8 Could have caused ninor ransi gry gain only, no injuries . . _ . . _ 3 8 Could have caused nild in uri s including redness lingering pain, sna cu s, scrapes, sore points, luaps or bruises . 6 8 Could have caused gedergtg injuries including serious cuts, blac eye, ex ensive bruising or swellin , broken nose or finger, strained nuscles oin s . have been lite threatening or caused nore gerious inyurigg including other broken bones unconsciousness, anything requiring a stay in the hospital or 5 8 Coul I 8 no injury or pain sustained How bany tines were you inyured y this? . -(123) . -(126) . (125) . (126) . ”mum .__._...(l231 . (129) .__.m_11301 . (133) .m”-(136) . (135) . (I36) . (137) 2 8 nild inyury (includes redness lingering pain or soreness, snall cuts, scrapes, sore joints, luaps or bruises) 10 extensive bruising and swelling$ fractured 3 8 noderate injury - (includes cu _ _ nose or tinger, strained nuscles or goin 6 8 severe inyury - (includes required hospitalization required surgery, ternination of a pregnancy, kniI requiring stitches, black eye, 0 reatnent, or visit to doctor etc. broken are or leg, pernanent disabilin or gunshot woun s, , eaergency roon etc.) 139. How upset were you by the worst of the incidents that you reported in questions 6123 to 137? hot at all upset I. . . . .2. . . . .3. . . . .6. . . . .5 Extrenely upset , unconsciousness, 353 I60. Uhich of the following best describes how long you were generally affected by the worst of the incidents you reported? . I 8 fly feelings of anger and hurt assed as soon as the incident(s) was/were over. 2 8 I was upset for up to several ours afterward 3 8 I was upset for up to several weeks 6 8 I was upset for longer than that but got over it 5 8 Ihe experience(sl pernanently aercted ae. . . . . . . . . . . . . . . . . . . . . . . . . . . . . (160) how, with the followin questions indicate how often you did each of these things when you and this dating partner uere haying an argunent.. A so, as before, for each behavior is ed, indicate how dangerous you feel this action was be indicatin how such physical 6376.16 ould have caused your partner. It doesn't natter, in rating the dangerousness 0 an act whet er helshe was in fact inyure or hurt. Consider instead what night the consequences of this action have been if Ihe 'd experienced the full inpact of it?). Also, indicate in Ihe spaces on the far right how nany tines your partner actual y was inyured in any way by each type of action. Leave blanks if the behavior never occurred. How nan How often, in argunents, did you... tines 01.. How nany How dan erous was the nost serious incident of this the! iniured ‘ tines? be avior? (Use the ratings below.) by his. I‘II 1h,“ m.thin' n h.’lhi.?I I I I “oo.--. I I I I I I I I I I I I ‘ -I I I I I I I I I I I I I I I I-..-...(I‘1’ ‘42I “‘h 0' “0'. h.’lbi.?I I I I I I I"..-I I I I I I I I .I I I I I ----...I I I I I I I I I I I I I I I Imm(‘42) I63. Grab her/hie (not around ‘h. n.‘k)?I I I I I I I I I I I I I” ..... I I I I I I I I I I I I I 00...... I I I I I I I I I I I I I I I. -(143) 14‘I $1., ”IF/”6.1% I I I I I I I I I I oo.-o--. I I I I I I I I I I I I ””---I I I I I I I I I I I I I I I I --(‘44) I65. Physically restrain or hold her/hin forcibly and keep then "“ “'ing?I I I I I I I I I I I I-”-”I I I I I I I I I I I I I ..... I I I I I I I I I I I I I I I I (:‘s) “‘I Pull “rihi’ h.i'?I I I I I I I I I“.--..I I I I I I I I I I I I I ““..-I I I I I I I I I I I I I I I In“---(l‘s) 167. Iwist or bend her/his are, hand, 0' ‘inger‘?I I I I I I I I I I I I -------I I I I I I I I I I I I I ....... I I I I I I I I I I I I I I I I“-m(‘47) 14.I ‘i‘t h.’/hi.?. I I I I I I I I I I “--“-I I I I I I I I I I I I I ...... I I I I I I I I I I I I I I I I--.....(!‘a) “,I .i‘. h.’/.i.?I I I I I I I I I I I .- I I I I I I I I I I I I I ”no.-. I I I I I I I I I I I I I I Im(149) 15°I "i‘ ..'Ihi. 'i‘h . 'i"?I I I I I I ”I I I I I I I I I I I I I --....-I I I I I I I I I I I I I I I om----(3507 I51. Hit or try to hit her/hin with M‘hifl’oI I I I I I I I I I I I In“---I I I I I I I I I I I I I --..---I I I I I I I I I I I I I I I I--””-(i51) 152I “I3 h.'/hi. u’?I I I I I I I I I In“---I I I I I I I I I I I I I ““---I I I I I I I I I I I I I I I I- -(152) 153. Choke or grab her/his around I”. a“ ?I I I I I I I I I I I I I -I I I I I I I I I I I I I .. I I I I I I I I I I I I I I I I (153) 156. r t her/his with a knife or 'ua-IIIIIIIIIIIIIIII ”IIIIIIIIIIIII -IIIIIIIIIIIIIIII (‘5‘) ‘SSI ”I. . kni'. " 'i'. . wa?I I I I I”--..I I I I I I I I I I I I I -m..-’ I I I I I I I I I I I I I I I (‘55) Dan erousness Ratin s I 8 Could ave cause no sical a'n or harn 2 8 Could have caused ninor transiIor ain only, no injuries _ _ 3 8 Could have caused nild in ur‘ including redness lingering pain, sna cu s, scrapes, sore joints, luaps or bruises 6 8 Could have caused noderate ‘n'uries including serious cuts, blac eye, ex ensive bruising or swellin , broken nose or finger, strained nuscles or oin s 5 8 CoulI have been life threatening or caused nore serious inyuries including other broken bones, unconsciousness, anything requiring a stay in the hospital 354 hhy do you think you engaged in the behaviors you reported in itens Ildl to l55? Check all that apply. (If you never did any of he behaviors described in these itens, go to question OISE.) ....-l56. uncontrollable anger __._-l57. feelings of jealousy or hurt __.._ISB. to get your partner to do sonething specific ....-159. I don't know, it just happened ._._-l60. to get your partner to stop doing what they had done ..___l6l. to nake you partner sorry for what they’d done, or to nake hin/her 'pay' ..._.l62. to get your partner to listen l63. to defend yourself lid. to scare or intinidate your partner lbs. Please use the space below to describe other reasons or causes you think night have led you to engage in these actions. Please write or print legibly. bow think about the nost regent fight or argunent that you and this partner had in which any of these ph sical ex ressions of anger occurred (t a is, in which either you or your partner did an of the actions described in ques ions 012 through :37 g; ldl through 155 above). Answer the following questions about t at nost recent fight. Please write or print egi y. lbs. lhat started this fight? Uhat was it about? 167. what specific interactions preceded the actions described in questions tl23 through I37 and ldl through 155 that occurred in this nost recen fight? 3555 l68. Describe exactly what happened in this fight. Include especially any inforaation that you think isn’t included in the preceding questions. 169. Uho do you think was the 'aggressor' in this fight, and who the 'victin'? Uhy do you think so? I70. Ihich of the lollowing best describes how you think these incidents affected your relationship? I 8 Ihey had a positive effect. A 8 They had a lasting negative effect 2 8 Ihey had li tle or no effect. 5 8 Ihey ruined the relationship. . . . . . . . . . .__-..(l70) 3 8 they had a tenporary negative effect Different people respond in different wa s when they are angry or upset with a dating partner. Ihe questions on the following page ask you about soae other hings that dating partners any do when they are angry or upset with each other. Please indicate how often the dating partner you've been answering all these questions abou did (or does) these things to annoy or hurt you, or even at other tines. 356 Uhen this)dating partner wanted to annoy or hurt you, how often did she or he... (Place an 'I' under the appropriate response. leeer Raeely Sonegines Often l7l. Refuse to talk to you?. . . . . . . . . . . . . . . . . . . . . . . . . . ____-. . _.._.. . . .__.... . ..__.,. .(lIl) I72. Accuse you of having or wanting other sexual relationships?. . . . . . . .--,._. . ..,... . . ..__._. . .__.... .(l72) I73. Iell you about his/her other sexual relationships?. . . . . . . . . . . . .._... . _..... . . .,____. . ....,-. .(173) I74. Refuse to do things with you that you wanted to do?. . . . . . . . . . . .-___-. . ...... . . .__.... . .__.__. .(l74) I75. forbid you to go out without hin/her?. . . . . . . . . . . . . . . . . . ..__... . ...--. . . ...__-. . ....__. .(l75) I70. Iry to control how you spent your noney?. . . . . . . . . . . . . . . . . __.._. . .__... . . ....... . .___... .(176) I77. Iry to control how you spent your free tine?. . . . . . . . . . . . . . . ...._. . .._... . . .,.,... . .__,... .(l77) l7R. Iithhold approval, appreciation or affection as punishnent?. . . . . . . .__.... . __._-. . . ._...-. . ..__... .(l7i) I79. Lie to you or deliberately nislead you?. . . . . . . . . . . . . . . . . ._...-. . ..._-. . . ._.___. . .___._. .(l79) Ibo. flake contradictory denands or requests?. . . . . . . . . . . . . . . . . ._..--. . _-._.. . . . ..... . . ._..... .(lRO) lBl.CalIyounanes?......................i........._“..___...._"__.....m..(ldl'i IRZ. Iry to huniliate you?. . . . . . . . . . . . . . . . . . . . . . . . . . ._..... . _.___. . . ...-... . ._..... .(l82) l93. Ignore or nake light of your anger?. . . . . . . . . . . . . . . . . . . .._.--. . _..--. . . ._.---. . ._..--. .(183) IDA. Ignore or nake light of your other feelings?. . . . . . . . . . . . . . . ..-__. . _-.._. . . ..._._. . ._..._. .(le) IRS. Ridicule or criticize you in public?. . . . . . . . . . . . . . . . . . . ....,. . _._._. . . ...-... . .__.._. .(IBS) ICC. Ridicule or insult your nost valued beliefs7. . . . . . . . . . . . . . . _.,__. . _,..-. . . ._.._.. . .___... .(ldd) l87. Ridicule or insult your religion, race, heritage or class?. . . . . . . . .__... . ____-. . . .___._. . .-..-_. .(187) ISO. Ridicule or insult your gender as a group?. . . . . . . . . . . . . . . . ._.--. . .._... . . ._.._-. . ...-__. .(lBBl 189. Criticize your strengths, or those parts of yourself which you were proud of?. . . . . . . . . . . . . . . . .__.__. . __.--. . . ...___ . ._.__-. .(IBS) l90. Criticire your intelligence?. . . . . . . . . . . . . . . . . . . . . . . --_... . ,._-.. . . .__,--. . ._..... .(I90) l9l. Criticire you physical appearance and/or sexual attractiveness?. . . . . ....-.. . ...... . . ..._... . ._..... .(l9l) l92. Criticire your fanily or friends?. . . . . . . . . . . . . . . . . . . . .,,.... . _._... . . .-..... . ._..._. .(l92) I93. harass your fanily or friends in sone way?. . . . . . . . . . . . . . . . ___... . _.___. . . ._...-. . .____.. .(l93) ISA. Discourage your contact with fanily or friends?. . . . . . . . . . . . . .__._.. . .._,_. . . .____,. . .____.. .(l94) l95. Ihreaten to hurt your ianily or friends?. . . . . . . . . . . . . . . . . _____. . _..... . . .____.. . .._.... .IISS) I“. breatordestroy sonething inportant to you?. . . . . . . . . . . . . . .__.. .____. . . .____. . ..___. .096) l97. break, destroy, throw, hit, or kick sonething else in your presence?. . . __.... . ..__.. . . .__.._. . ....... .(197) ISO. Abuse or threaten to abuse your pets?. . . . . . . . . . . . . . . . . . .__.__. . _.__.. . . .___... . .____.. .(ISB) l99. Leave yon sonewhere with no way to get hone?. . . . . . . . . . . . . . . ____,. . _.__.. . . ._____. . .___... .(199) 200. Ihreaten to end the relationship if you don't do what slhe wants?. . . . ._..... . ._..-. . . ....._. . .___... .(200) 20I. threaten to connit suicide when slhe is angry at you?. . . . . . . . . . .-..... . _...-. . . ....... . .-..... .(201) 202. Ihreaten to hit or throw sonething at you?. . . . . . . . . . . . . . . . -_.... . -_--,. . . .-,,... . .-..,.. .(202) 203. threaten you with violence verbally? . . . . . . . . . . . . . . . . . . ...._.. . ____.. . . ._._... . .__..,. .(203) 204. lake other threats not included in the itens above?. . . . . . . . . . . .-,._-. . ..__-. . . ._...,. . .-___.. .(204) 357 On the itens below, enter the nunber corresponding to the appropriate answer in the spaces on the right. 205. how upset were you by the incidents that you reported in questions till through 204? hot at all upset I. . . . 2. . . . 3. . . . d. . . . 5 Extrenely upset. . . . . . . . . . . . . (205) 206. hhich of the following best describes how long you were generally allected by the worst of the incidents you reported? I 8 fly feelings of anger and hurt assed as soon as the incidents were over. 2 8 l was upset for up to several ours afterward 3 8 I was upset for up to several weeks 4 8 I was upset for longer than that, but got over it 5 8 Ihe experiences pernanently affected ne. . . . . . . . . . . . . . . . . . . . . . . . . . . ...---(2061 how think about those tines when you wanted to annoy or hurt your dating partner. How often did you... (Place an 'I' under the appropriate response.) leeer Raeely Sonegines Ofeen 207. Refuse to talk to hinlher?. . . . . . . . . . . . . . . . . . . . . . . . ._.--. . ._.._. . . ....... . .__.._. .(207: 208. Accuse your partner of having or wanting other sexual relationships?. . . _.---. . _,__-. . . ...-... . ._,-... .(208) 209. Iell bin/her about your other sexual relationships?. . . . . . . . . . . ...-_.. . -._.-. . . .___--. . .,____. .(205: 2l0. Refuse to do things with your partner that helshe wanted to do?. . . . . .-_-,.. . ._.._. . . ._.._-. . ._..... .(2l0) 2Il. Forbid hinlher to go out without you?. . . . . . . . . . . . . . . . . . .--.._. . _.__.. . . .--._.. . ....__. .(2ll) 212. Try to control how your partner spent his/her noney?. . . . . . . . . . . ._.-_. . ...... . . ...._.. . .,--... .(212) 2l3. Iry to control how your partner spent his/her free tine?. . . . . . . . . ..-_-. . _._-_. . . .._._-. . ....._. .(2l3) 2lt. Iithhold approval, appreciation or affection as punishnent?. . . . . . . ._--_.. . ..... . . . ....,.. . ....... .(2l4) 215. Lie or deliberately nislead hinlher?. . . . . . . . . . . . . . . . . . . -..._. . __-_-. . . ._.___. . .__.._. .(2l5) 2l6. lake contradictory denands or requests?. . . . . . . . . . . . . . . . . ._..... . _.,.-. . . .__..-. . ._____. .(216) 2l7. Call your partner nanes?. . . . . . . . . . . . . . . . . . . . . . . . . ...... . __-__. . . .-__.-. . .__.._. .(217) 2lB. Iry to huniliate hinlher?. . . . . . . . . . . . . . . . . . . . . . . . ...-_-. . -,-._. . . .__-_.. . ._..__. .(2l8) 219. Ignore or nake light of his/her anger?. . . . . . . . . . . . . . . . . . ._._.. . -_---. . . ._.__.. . ....._. .(2l9i 220. Ignore or nake light of his/her other feelings?. . . . . . . . . . . . . ..._._. . ,._--. . . .-...-. . .._..-. .(220) 22l. Ridicule or criticize your partner in public?. . . . . . . . . . . . . . ..__... . _._... . . ._,.__. . ._.___. .(221) 222. Ridicule or insult his/her nost valued beliefs?. . . . . . . . . . . . . .__.... . __,.-. . . .____.. . .__._-. .(222) 223. Ridicule or insult your partner's religion, race, heritnge or class?. . . ,.___. . _._... . . .,,.... . ._._... .(223) 224. Ridicule or insult his/her gender as a group?. . . . . . . . . . . . . . ._-..-. . .._... . . .._._-. . ..._... .(224) 225. Criticize your partner's strengths, or those things which helshe is or once was proud of?. . . . . . . . . . . . _..__. . ...__. . . .___._. . ._____. .(225) 225. Criticire his/her intelligence?. . . . . . . . . . . . . . . . . . . . . .__.__. . _..... . . ..._.-. . .__.._. .(226) 227. Criticize your partner's ph sical nppearance and/or sexual at ractiweness?. . . . . . . . . . . . . . . . . __.,,. . .._.-. . . ._,,__. . ....... .(227) 229. Criticize his/her fanily or friends?. . . . . . . . . . . . . . . . . . . _..-_. . __._-. . . ....... . ...._-. .(228) 229. harass his/her fanily or friends in none way?. . . . . . . . . . . . . . .-.._-. . _._--. . . ._..... . .__._.. .(229) 230. Discourage your partner's contact with fanily or friends?. . . . . . . . ._..-,. . _.-,-. . . .---_-. . .__.... .(230) 231. 7hrenten to hurt his/her fanily or friends?. . . . . . . . . . . . . . . .-.._-. . ---__. . . ....-_. . ....... .(23l) 232. break or destroy sonething inportant to hinlher?. . . . . . . . . . . . . . . . . . . . . . . .(232) 355E! bhen you wanted to annoy or hurt your dating partner, how often did you... (Place an 'I' under the appropriate response.) lever Rarely Sonetines Often I 2 3 233. break, destroy, throw, hit, or kick sonething else in his/her presence?. ....... . . . . . . . . . .(233) 234. Abuse or threnten to abuse your partner’s pets?. . . . . . . . . . . . . . . . . . . . . . . . .(234) 235. Leave hinlher sonewhere with no way to get hone?. . . . . . . . . . . . . .___,. . . . . .,_._.. . .__.... .(235) 230. Ibreaten to end the relationship if helshe didn't do what you wanted?. . ....._. . . . . . . . . . .(236) 237. Ihreaten to connit suicide when you were angry at hinlher?. . . . . . . . __.... . . . . .____.. . ._..... .(237) 23c. Ihreaten to hit or throw sonething ii hinlher?. . . . . . . . . . . . . . ___,,. . . . . .__,,.. . .____. .(239) IIII III II(239) 240. lake other threats not included in the itens above?. . . . . . . . . . . . . . . . . . . . . . .(240) 239. Ihreaten hinlher with violence verbally?. . . . . . . . . . . . . . . . . . . 241. Uhich of the following best describes how you think all of the incidents described in itens bl7l through 204 and 207 through 240 affec ed this relationship. I 8 They had a ositive effect. 4 8 Ihey bed a lasting negative effect 2 8 Ihey had li tle or no effect. 5 8 Ihey ruined the relationship 3 8 Ihey had tenporary negative effect below are specific questions about your sexual activity in this dating relationship.. In these questions the l al definition of sexual intercourse should be used. Ihat is, yexgal intercourse is.defined as 'penetration, no na ter how slight, of one erson b another; ejaculation is not require n a. i ion, it includes contact between the sex organs of one erson and he nout or anus of another. Penetration of the vagina or anus nay occur with any object. Indicate whet er or not each of the following occurred in this dating relationship by placing an 'I' under the approprinte answer. In this relationship, have you ever?... lo (I) Yes (2) 242. engaged in kissing, hugging or fondling when both you and your partner wanted to?. . . . . . . . . . . ____..(242) 243. bad sexual intercourse when both you and your partner wanted to?. . . . . . . . . . . . . . . . . _.__-. . -(243) (If you previously reported no incidents in questions 079, 30 or bl, skip now to iten 273 244. had sexual intercourse with your partner even though you didn't. really want to because helshe threatened to end your relationship otherwise?. . . . . . . . . . . _._.-. . _(244) 245. had sexual intercourse with your partner when ou didn’t really want to because you felt pressured by t eir continual argunents?. . . . . . . . . . . . . . . . (245) 245. found out that your partner had obtained sexual intercourse wit you by lying or saying things helshe didn't really nean?. . . . . . . . . . . . .__--,. . _____-(246) 247. had sexual intercourse with your partner unknowingly because they had gotten you very drunk or ’stoned or that purpose?. . . . . . . . . . . . . . . _____. . (247) 240. been in a situation where your artner used sone degree of physical force (twisting your arn, holding you down, etc.) to try o nake you engage in kissing or petting when you didn't want to?. . . (248) 24!. been in a situation where your partner threatened to use soae sort of ysical force to try to nake you engage in sexual contact (other than intercourse) en you didn't want to?. . . . . ___.._(249) 250. been in a situation where your partner tried to have sexual intercourse with you when you didn't want to by threatening to use physical force if you didn’t cooperate, but for various reasons sexual in ercourse did not occur?. . . . . . . . . . A ~ 0! O v 251. been in a situation where your partner actually used soae degree of physical force (twisting your arn, holding eou down, etc.) to get you to have sexua intercourse with hinlher, u for various reasons sexunl intercourse did not occur?. . . . . . . .___.,. . (25l) 359 In this relationshi have ever?... ’7 you no (I) Yes (2) 252. had sexunl intercourse with your partner when you didn't want to because helshe threatened to use physical force if you didn't cooperate?. . . . . . . . . . . . . . . . . . . (252) 253. had sexual intercourse with your partner even though you didn't really want to because helshe nade other threats (not covered in previous questions)?. . . . . . .,_ . . (253) 254 had sexual intercourse with your artner even thouuh you didn't want to because helshe used sone degree o physical force twisting your arn, holding you down, etc.)?. . . . (254) 255. been raped by your partner?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .___. . . _(255) 256 below indicate how severe the worst injuries you sustained as a result of the experiences reported in questions 0244 to 255 above were. I 8 nu injury or nin sustained . . 2 8 nild injury includes redness lingering pain or soreness, snall cuts, Iunps or bruises etc.) 3 8 noderate injury - (includes cuis requiring stitches blauk eye, extensive bruising and swellin , fractured nose or fin er, strnined nuscles or joinls, visit to eaergency roon or doctor, etc. 4 8 severe injury - (inclu es required hospitalization, broken arn or leg, pernanent disability, unconsciousness, required surgery, ternination of a pregnancy, knife or gunshot wounds, e c.). . . . . (256) 257. how upset were you by the incidents that you reported in questions 0244 through 255? hot at all upset I. . . . 2. . . . 3. . . . 4. . . . 5 Extrenely upset. . . . . . . . . . . . . (257) 258. hhiuh of the followin: best describes how long you were generally affected by the worst of the incidents you reporte ? I 8 by feelings of anger and hurt assed as soon as the incidents were over. 2 8 I was upset for up to several ours afterward 3 8 l was upset for up to several weeks 4 8 l was upset for longer than that but got over it 5 8 The experience(s) pernanently aflected ne. . . . . . . . . . . . . . . . . . . . . . . . . . ._...-(258) The questions below ask about whether or not you have initiated any of these types of interactions in this dating relationship. In this relationship, have you ever?... I (l) l (2) 0 es _(259) 260. threatened to end your relationship if your partner did not have sexual intercourse with you?. . . . . __-(260) 253. becone so sexuelly aroused that you could not stop yourself, even though your partner did not want to have sexual intercourse?. . . . . . . . . . . . . . . . . . . . . . . . 26l. had sexual intercourse with nour partner when helshe didn't really want to because you continua ly pressured then with verbal argunents?. . . . . . . . . . . . . . ...,... . (26l) 262. lied or said things that you didn't rally nean in order to have sexual intercourse with your partner?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (262) 263. gotten your partner very drunk or 'stoned' to have sexual intercourse with then?. . . . . . . . . ,.__,. . (263) 264. used soae degree of physical force (twistinu an arn holding the person down, etc.) to try to nake your partner engage in sexual contact other than intercourse) when helshe didn't want to?. . . . ___..-(264) 265. thyggtgned to use soae sort of h sical force to tr to nake our artner en a e in sexua contact (other than i:t:rcowrse) when helzhe didn't’want'to?. . . 3 3 . . . . . . . . . . . __..__(265) 266. tried to have sexual intercourse with your partner when helshe didn't want to by threatening to use p ysical force if they didn't cooperate, but for various reasons sexual intercourse did not occur?. . . . . . . . . _,_,. . ”““(266) 267. actually used none degree of physical force (twisting an arn, holding then down, etc.) to get your partner to have sexual intercourse with you, but for various reasons sexual intercourse did not occur?. . . . . . . . . . . . . . . . . . __,.._(267) 263. had sexual intercourse with your partner when helshe didn't want to because you threatened to use physical force if they didn't cooperate?. . . . . . . . . . . . . . . . . . . . (268) 360 In this relationship, have you ever?... lo (I) Yes (2) 263. had sexual intercourse with your partner even thou h helshe . . didn't really want to because you nade other threa s (not covered in previous questions)?. . . . .- . . (269) 270. had sexual intercourse with your partner even though helshe didn't want to because you used sone degree or physical force (twisting your arn, holding you down, etc.)?. . . . . . . .--_.,. . .,._-.(270) 27l. rnped your partner?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ._._... . _._._-(27l) __,.,272. Ihich of the following best describes how you think these incidents affected your relationship? I 8 They had a ositive effect. 4 8 They had a lasting negative effect 2 8 They had li tle or no effect. 5 8 They ruined the relationship. . . . . . . . . . ._..._(272) 3 8 They had tenporary negative effect In this neut sectiun you are asked to describe the decision naking in this relationship. becisions in relationships are ne otiated in a varietu of ways, depending on the peo le involved and on the issue to be decided. below are a nunber of dif erent decisions tha a couple of en na es while da ing. .Tor each one, circle the nunber that corresponds to the best description of how that kind of decision islwas nade in this relationship. I 8 I alnost always decide . 4 8 fly partner usually hns nore influence 2 8 I usually have nore influence over the over the decision than I do, decision than ny partner does 5 8 by partner alnost always decides 3 8 by partner and I usually have equal influence how did (or do) you and this dating partner decide... 273. bhen to see each other?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(273) 274. bho to spend tine with when you are together?. . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(274) 275. bhat you wear when you are going out with your partner?. . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(275) 27b. bhat your partner wears when you are going out together?. . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5.(276) 277. bhat clothing and hair styles are best for you, in general?. . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(277) 27b. hhat clothing and hair styles are best for your partner, in general?. . . . . . . . . . . . . . . .I..2..3..4..5.(27b) 279. bhether or not to engage in sexual activities?. . . . . . . . . . . . . . . . . . . . . . . . . . .i..2..3..4..5.(279) 230. How to engage in sexual activities (i.e. 'how far to go’, what kind of protection to use, etc.). . l..2..3..4..5.(280) 281. How to spend your tine when you are together?. . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(281) 282. how you spend your noney?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(282) 283. how your partner spends his/her noney?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5.(2b3) 234. how to deal with relationship probleas?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(284) 285. how nuch tine to spend together?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5.(235) 236. Ibo you spend tine with when you are not with your partner?. . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(286) 237. bho your partner spends tine with when not with you?. . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5.(237) 230. Ihat you do in your free tine away froa your partner (not counting who you spend it with)?. . . . .l..2..3..4..5.(288) 289. bhat your partner does in his/her free tine away fron you (not counting who they spend it with)?. .l..2..3..4..5.(289) 290. Ihether or not you have contact with your previous dating partner(sl?. . . . . . . . . . . . . . . i..2..3..4..5.(290) 291. Ihether or not your pnrtner has contact with his/her previous doting partner(s)?. . . . . . . . . .l..2..3..4..5.(29l) 292. lhat classes, choices of najor or career plans are best for you, in general?. . . . . . . . . . . .I..2..3..4..5.(292) 293. bhat classes, choices of najor or career plans are best for your partner, in general?. . . . . . . I..2..3..4..5.(293) 361 Next indicate how you think these decisions would be nadu in your idggi relationship - i.e. one in which the decisions were alwnys nade exactly the '.K'y°:h‘:‘nt they should be. Hr to answer ese uestions as ou th' k t th i a r e innin of the relations ou have been describ_ng. ven oug a nig e i , y a your re a ions ip i ea s we e a a ine an answer ese questions (run that perspective. Take a nonent now to try to renenber exactly uhen that wouid have been and to bring to nind the details abou your life at that tine. bon't consider how things were with this particular dating partner when you answer, just describe how ou thought these decisions should be nade in an ideal relationship at that tine. If you haven’t had an dating relationn ips in colle e, or in the past four years if youfre oyer 24 answer the following questions according to ow you think decisions should nude in an ideal ating re ationship that you any be involved in in the future. for each one, circle the nunber that corresponds to the best description of how that kind of decision would be nude in an idggl,relationship. l 8 I would alnost always decide 4 8 he nartner would usually have sure 2 8 I would usually have nore influence over the n luence over the decision than I decision than ny partner would 5 8 by partner would alnost always decide 3 8 by partner and I would usually have equal influence In an ideal relationship, how would you and a dating partner decide... 29‘Iu.nt°‘....‘hogh.'?IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIZII3II4II5I‘29‘) 295I “I ‘0 ”'0‘ II“ 'i‘h ”I“ m .r. I”.Ih.’?I I I I I I I I I I I I I I I I I I I I I I I I I I I IIIZIISII‘II5I(295) 296. bhat you wear when you are going out with your partner?. . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(296) 297. what your partner wears when you are going out together?. . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5.(297) 290 299 300 30l. how to engage in sexual activities (i.e. 'how far to go', what kind of protection to use, etc.). . l..2..3..4..5.(301) what clothing and hair styles are best for you, in general?. . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(293) bhat clothing and hair styles are best for your partner, in general?. . . . . . . . . . . . . . . .l..2..3..4..5.(299) bhether or not to engage in sexual activities?. . . . . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..3.(300l 302. how to spend your tine when you are together?. . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(302) 303. how you spend your noney?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(303) 304. how your partner spends his/her noney?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5.(304) 305. How to deal with relationship probleas?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(305) 306. How nuch tine to spend together?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5.(306) 307. Io you spend tine with when you are not with your partner?. . . . . . . . . . . . . . . . . . . . l..2..3..4..5.(307) 300. Iho your partner spends tine with when not with you?. . . . . . . . . . . . . . . . . . . . . . . .I..2..3..4..5.(303) 303. Ihat you do in your free tine away (run your partner (not counting who you spend it with)?. . . . .l..2..3..4..5.(309) 3l0. Ihat your partner does in hinlher free tine away froa you (not counting who they spend it with)?. .I..2..3..4..5.(3i0) 3i). Ihether or not you have contact with your previous dating partner(s)?. . . . . . . . . . . . . . . l..2..3..4..5.(3ll) 3l2. Ihether or not your partner has contact with his/her previous dating partner(s)?. . . . . . . . . .l..2..3..4..5.(312) 3(3. Ihat classes, choices of najor or career plans are best for you, in general?. . . . . . . . . . . .I..2..3..4..5.(3i3) 3l4. bhat classes, choices of najor or career plans are best for your partner, in general?. . . . . . . l..2..3..4..5.(3ld) A ' d s . best are none questions concerning your attitudes and feelings about the behavior of non and wonen toward each other in their everyda lives and also in their ronantic and sexual behavior. A ain tr nswer these uestions as ou th nk ou would have a the be innin of the relationshi that ou have een descri n . r‘ o renen er ow you e a co ings a e ine an answer ese ques ions roa a perspe ou aven't ad any dating relationships in college (or in the past 4 years, if you're over 24) then answer these ques ions as ou see t ings now. Please indicate how :irongly you would ave agreed or disagreed with each statenent by circling the nu er corresponding to the appropriate swer. 362 lndicate below how strongly you would have agreed or disagreed with each statenent by circling the nunber corresponding to the appropriate answer. 315. 316. 317. 318. 3l9. 320. 32l. 322. 323. 324. 325. 325. 327. 320. 329. 330. 33l. 332. 333. 334. 335. 336. 337. 338. 339. l 8 Strongly agree . 2 8 Agree 5 8 disagree soaewhat 3 8 Agree soaewhat h 8 Disagree . 4 8 leither agree nor disagree 7 8 Strongly disagree A nan should fight when the wonan he's with is insulted by another nan. . . . . . . . . . . .l..2..3..4..5..h..7.(315) it is acceptable for the wonan to pay for the date. . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5..h..7.(316) A wonan should be a virgin when she aarries. . . . . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5..h..7.(3i?) there is sonething wrong with a wonan who doesn't want to aarry and raise a fanily. . . . . .l..2..3..d..5..b..7.(3l8) A wife should never contradict her husband in public. . . . . . . . . . . . . . . . . . . . .l..2..3..4..5..6..7.(319) it is better for a wonan to use her feninine chars to get what she wants rather than ask for it outright. . . . . . . . . . . . . . . . . . . . l..2..3..4..5..b..7.(320) it is acceptable for a wonan to have a career, but aarriage and faaily should cone first. . .l..2..3..4..5..h..7.(321) it looks worse for a wonan to be drunk than for a nan to be drunk. . . . . . . . . . . . . . l..2..3..4..5..5..7.(322) lhere is nothing wrong with a wonan going to a bar alone. . . . . . . . . . . . . . . . . . .l..2..3..4..5..6..7.(323) A wonan will only respect a nan who will lay down the law to her. . . . . . . . . . . . . . .l..2..3..4..5..h..7.(324) hany wonen are so deaanding sexually that a nun just can't satisfy then. . . . . . . . . . . f..2..3..4..5..h..7.(325) A nan's got to show the wonan who's boss right froa the start or he'll end up henpecked. . . l..2..3..4..5..h..7.(326) doaen are usually sweet until they've caught a nan, but then they let their true self show. .l..2..3..4..5..h..7.(327) A lot of nen talk big, but when it cones down to it, they can't perforn well sexually. . . . l..2..3..4..5..h..7.(328) In a dating relationship a wonan is largely out to take advantage of a nan. . . . . . . . . .l..2..3..4..5..h..7.(329) hen are out for only one thing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5..h..1.(330) host wonen are sly and aanipulating when they are out to attract a nan. . . . . . . . . . . .l..2..3..4..$..h..7.(33f) A lot of wonen seen to get pleasure in putting aen down. . . . . . . . . . . . . . . . . . . l..2..3..4..5..b..7.(332) People today shouldn't use 'an eye for an eye and a tooth for a tooth' as a rule for living. l..2..3..4..5..h..7.(333) being roughed up is sexually stiaulating to nany wonen. . . . . . . . . . . . . . . . . . . .l..2..3..4..5..h..7.(334) hany tines a wonan will pretend she doesn't want to have intercourse because she doesn't want to seen loose, but she's really hoping the non will force her. . . . . . . . . .l..2..3..4..5..h..7.(335) A wife should uove out of the house if her husband hits her. . . . . . . . . . . . . . . . . l..2..3..4..5..6..7.(336) Sonetines the only way a nan can get a cold wonan turned on is to use force. . . . . . . . . l..2..3..4..5..h..7.(337) A nan is never justified in hitting his wife. . . . . . . . . . . . . . . . . . . . . . . . .l..2..3..4..5..b..7.(338l A wonan is never justified in hitting her husband. . . . . . . . . . . . . . . . . . . . . . l..2..3..4..5..h..7.(339) MW 363 below is a list of different services on berkeley's caapus that are available to students. lndicate (by placing an 'l' under the arpropriate coluan) those (1) that you are faniliar with (i.e. you know what they do and have a good idea about a how you won d contact then if you wanted to), (2) those that you have been to, (3) those t t friends of yours have been to, and, (4) those that you think you night use in the future, if a need arose. Check as nany as apply. I an faniliar with this service. 1. Career Planning and Placenent. . . . ._ . . 2. Student Advising. . . . . . . . . . . . . 3. The Counseling Center. . . . . . . . ..._. . . 4. Rape Prevention. . . . . . . . . . . . . . 5. The Student Learning center. . . . . . . . ‘I P." ‘dVi‘ingI I I I I I I I I I I I I ...... I I 7. Psychiatric Services. . . . . . . . . ...... . . 8. lhe anen's Center. . . . . . . . . . . . Hhile at UC Ierkeley, have you ever gone to a counselor or 9. probleas you experienced in handling acadeaic stress?. . . . . . . . . . . . . . . . . . . . . 10. l . 12. b. Uhile at l have been to this service. People I know have been to this service. l eight use this service in the future. I I I I I I I I I I I I I I I I I I I‘I) I I I I I I(2) I I I I I I(3) I I I I I I I I I I I I I I I I I I It‘) I I I I I I I I I I I I I I I I I I I‘S) I I I I I 0(6) I I I I I I(7) I I I I I I‘a) therapist (other than an acadeaic counselor) to talk about. . Yes I I I I ‘9) choosing a najor or deciding on a career?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .(l0) probleas you experienced in your relationships with a dating partner?. . . . . . . . . . ._ . . . . . .(ll) 0th., PIfsonIl P'ODI..‘?I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I‘lZ) UC berkeley, have you ever gone to a counselor or therapist off caapus? Yes lo 13. private therapist in or around Perkeley. . . . . . . . . . . . . . . . . . . . . . . _._.-. .._.... . . .(l3) 14. private therapist in your hone town (if not local). . . . . . . . . . . . . . . . . ._..... ..._... . . .(l4) IS. a local clinic (specify). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .___... .__.._. . . .(lS) 16. a self-help prograa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ____.. .____.. . . .(16) 17. other (please specify) . . . . . . . . . . . . . . . . -._._. .____.. . . .(l7) froa what you know about (he Counseling Center, both first-hand_and (row what on have heard, lease rate it on the following diaensions. Circle the nuaber between the two adjectives that you t ink best descri es the Counseling Center. ‘al I I I I I I I I I I I I I I I I.“..’ibl.I IlII2II3II‘IISII‘II7I 19. 20. 2f. 22. 23. 24. . . . . . . . . .responsive. I I I I I I I I I I ..""II . . . sensitive to needs of people like nyself. .helpful with probleas in choosing a najor or career. . helpful in dealing with acadeaic stress. helpful in dealing with personal probleas. n‘no2aa’on‘aa5an‘oo7o a‘an2ao’no‘ao’nu‘an7n I’II2II3II‘IISII‘II7I I'II2II3II‘IISII‘II7I I‘II2II3II‘IISII‘II7I IIII2II3II‘II5II‘II7I .inaccessible. . . . . . . . . . . . . . . . ((8) I.n'."°n‘i'.I I I I I I I I I I I I I I I I (‘9) InOI “.I'tulI I I I I I I I I I I I I I I I l(ZO) insensitive to the .needs of people like nyself. . . . . . . . .(Zl) unhelpful uuth probleas in .choosing a najor or career. . . . . . . . . (22) .unhelpful in dealing with acadeaic stress. .(23) .unhelpful in dealing with personal probleas.(24) 364 _..._25. How nuch of what you know about the Counseling Center is based on your personal experiences, and how nuch on what you've heard frou others? 2 8 based coo letely upon by own experiences 8 based nos ly on by own experiences 3 8 based partly on :h own experiences and partly on what l've heard 4 8 based aostly on at l've heard 5 8 based conpletely upon what l've heard ..___2h. how likely is it that you would go to talk to a counselor about eroblens or experiences ou have in your gating rglgtrggshigg, if the probleas were really bothering you and ta king with friends or wi h others i n e p. Very unlikely l. . . .2. . . .3. . . .4. . . .5 Very likely .._27. How likely is it that you would go to talk to a counselor about other probleas you say have if they were really bothering you and talking with friends or others didn't help? Very unlikely l. . . .2. . . .3. . . .4. . . .5 Very likely (HANK YOU PO! YOUR llfli AID CONSIDERAIIOI ll flLLlIi OUT TRIS VUESTIUIIAIRE!!!! If you would like to talk with soueone about any reactions or feelings you aay be having in response to soae the uestions and content areas covered here, reaeaber t at you nay call no ulia Parisian, at he Counselin Center (642- 366). You say also nake an aBpointuent to talk with an of t e counselors there, if you would like. I friedaan at the Rape Prevention and Education rograa is also available 0 talk with you. her phone nuaber is h42-73lo. Reaeaber to fill in the inforaation on the participation agreenent that was included with the questionnaire and return it, and this conpleted uestionnaire (in separate enveloges to protect your anonyaity) to the Counseling Center so that your nane can be entere in the two drawings for the 450 ift certificates. If you get it there by beceaber 0th at 4:00 IPAV {gu'll be eligible for [gth drawings; by the 15th at :00 p.n and you'll nake it for the final one. -6009 LUCK ll (HE l 5!! And thanks again. 365 UNIVERSITY OF CALIFORNIA, BERKELEY SANTA 3M ° META CRUZ COUNSELING AND PSYCHOLOGICAL SERVICES COUNSELING CENTER 200 BUILDING T-S BERKELEY. CALIFORNIA 91720 November, 1989 Dear Student, I would like to Invite you to participate in an important study that is being conducted. It is a study of college students' dating relationships. You are among 800 juniors and seniors at Berkeley who have been randomly chosen to participate. If you decide to do so, you will be eligible to win a $500 gift certificate. I am conducting the study in cooperation with The Counseling Center, in Building T-5. If you were to agree to participate, you would be asked to fill out the enclosed anonymous questionnaire. It will take about an hour. It includes questions about your dating relationships, your mood, health, and some of your views about men and women. It also includes some questions about arguments and sexual experiences that you may have had with dating partners. All of your answers to these questions will be kept completely anonymous and confidential. You will notice that neither your name, nor any other identifiers appear on the questionnaire. This ensures your anonymity. No one will ever be able to identify the answers of any of the study participants. Participation in this study is completely voluntary. If you do agree to answer the questionnaire, you are free to skip any question that you would rather not answer. If you decide to participate, you will be eligible to enter two drawings for $500 gift certificates from Macy's. The first drawing will be held on December 8th, and the second on December 15th. People who return their questionnaires by the 8th will be entered in both drawings. Your completed questionnaire and your signed consent form must be returned to The Counseling Center in Building T5 by December 8th or 15th at 4:00 p.m. in order to be entered in the drawings. If you have any questions about this research project, or if you want to talk about any feelings or reactions that come up for you about the questionnaire or specific experiences it asks you about, you can reach me, Julia Parisian, at The Counseling Center. The phone number there is 642-2366. Or you can call me at home at 526-1400. You can also call Rob Friedman, at The Rape services accredited by the International Association of Counseling Services internship program accredited by the American Psychological Association 366 Prevention and Education Program. Her phone number is 642-7310. We both welcome your calls. Enclosed are two copies of a consent form which, when you sign it, indicates your willingness to participate in this study. One of these is for you to keep. If you decide to participate, sign one and print your name where indicated. Fill in the information about how to reach you in case you win one of the drawings and return the consent form in one of the two enclosed, postage-paid envelopes. It is important that you return your consent form and your questionnaire in two separate envelopes in order to protect the anonymity of your answers. You may also drop your questionnaire and consent form off at the front desk of The Counseling Center in Building T5. Remember to enclose your questionnaire and consent form in the two different envelopes either way. I hope that you decide to participate in this study because I think it is an important one. If you do, thanks in advance for your time and cooperation. If you decide not to, thank you anyway for your consideration. Remember, if you have any further questions, you can give me a call. Sincerely, . 4% Julia A. Parisian, M.A. The Counseling Center 642-2366 or 526-1400 367 UNIVERSITY OF CALIFORNIA, BERKELEY IEIIEELEY ° DAVIS ' IRVINE ° LOS ANGELES ' RIVERSIDE 0 SAN DIEGO ' SAN MCISCO IAN'I‘A SAHARA ° SANTA CRUZ COUNSELING AND PSYCHOLOGICAL SERVICES COUNSELING CENTER 200 BUILDING T-S BERKELEY. CALIFORNIA 94780 Participation Agreement You are invited to participate in a study of college students’ dating relationships. Participation entails filling out the enclosed anonymous questionnaire. Students who complete and return it will be eligible to win a $500 gift certificate from Hacy's. This study is being conducted by Julia Parisian, in cooperation with The Counseling Center. It will take about an hour to fill out the questionnaire. It includes questions about your relationships, and about arguments and sexual experiences that you might have had with dating partners. It also includes questions about your mood and health, and about some of your views about men and women. All of your answers to these questions will be kept completely anonymous and confidential. Your name can never be linked with your answers by anyone connected to the study or outside of it. Participation in this study is completely voluntary. Some of the questions concern personal and, perhaps, upsetting events that have occurred in your life. If you decide to participate, you are free to skip any question that you would rather not answer. Also, if you have any questions about this research project which are not answered here, or if you want to talk about any feelings . or reactions that come up for you about the questionnaire or specific experiences it focuses on, you can reach me, Julia Parisian, at The Counseling Center. The phone number there is 642-2366. Or you may call me at home. My home phone number is 526-1400. You can also call Rob Friedman, at The Rape Prevention and Education Program. Her phone number is 642-7310. If you decide to participate, your name will be entered in two drawings for $500 gift certificates from Macy's. One of the drawings will be held on December 8th, and the other on December 15th. People who return their questionnaires and signed'consent forms by December 8th will be eligible for both drawings. Your questionnaires must be returned to the'counseling center by December 8th, or by the 15th at 4 p.m., in order to be entered in the drawing. Further instructions are in the enclosed letter. If you are interested in participating in this study, please read and sign the back of this form. Fill in the information requested so that we can contact you if you win one of the gift certificates, and return this form in the small, postage paid envelope which is included. Thanks! services accredited by the International Association of Counseling Services internship program accredited by the American Psychological Association 368 1. I understand that my participation in this study is completely voluntary and that I am free to skip any questions that I do not wish to answer. 2. I have read and understood the information explaining this study on the other side of this form and in the letter which accompanied this Participation Agreement. I understand that my participation in this study will involve filling out the enclosed questionnaire and returning it and this form in the two separate postage-paid envelopes to The Counseling Center address. 3. I understand that my name will be entered in drawings for two $500 gift certificate from Macy's, if I return my questionnaire and consent forms to the Counseling Center by December 8th at 4:00 p.m., and one such drawing if I return them by December 15th at 4:00 p.m. 4. I understand that all of the information on the questionnaire will be handled confidentially. I further understand that all the information that I share will be anonymous. It will be impossible for anyone to link my responses with my name. 5. I understand that there may be no direct benefits to me as a result of my participation other than eligibility in the drawings. However, I also understand that the information that I share may be useful in helping The Counseling Center to improve their services to students in the future. 6. I am at least 18 years of age. 7. I agree to participate in this study, as it is described herein. Signature Date Print name here Address where my gift certificate can be sent after December 15th if I am awarded one of them. (Please print.) Phone number Dec. 15th through 18th Check here if you would like to receive information about the results of this study when they are available (next spring). Address where that information should be sent. "ITilll'lilllllllllli