:Z:t.;::. : ... _ l . L. . i :3 3:53.»... .. I. .12.; .3 3.. tr. .4; .1. ._ .31.? 7317.55)“, a...» 2.51:}; .5. L: .1.“ : 1 52:15:: .1 I 30.3.53... 3:“ : 2.2.}...4 2...... 3.1:... a. 5...... ’33:. \ .5... 25.3.1)?! 3.)).(4. 4 G .Y 3.3.. .333. 4;) .. .5353 1): 34:1,»)! In}... 1... 3):.)1L133L l £111.35 I:)....i>|$.. ‘33.; 4 .7. :23) .3121. .I: T... :3. . 5‘ . t. . J: SILTY \\\l\l\l\\\Jl\\\\\\l LIBRARY Michigan State University \llllllllllllllllll 131293 This is to certify that the dissertation entitled Wife Abuse Attitudes and Attributions: The Role of Fear of Crime and Sex-Role Ideology presented by Helene Lensky Finke has been accepted towards fulfillment of the requirements for PhD degree in Psycholqu C7 @Z 5* . 0/ Major pttofessor-f DateQ/fzi Wfl MSUis an Affirmative Action/Equal Opportunity Institution 0—12771 ________...._._ . . i.._ V _ _. V ,7 fl , if 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 ram iii-M . with 082535? l1 [ml/5&4": [mo We 021412 MSU Is An Affirmative Action/Equal Opportunity Institution WMWMHJ WIFE ABUSE ATTITUDES AND ATTRIBUTIONS: THE ROLE OF FEAR OF CRIME AND SEX-ROLE IDEOLOGY BY HELENE LENSKY FINKE A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1994 ABSTRACT WIFE ABUSE ATTITUDES AND ATTRIBUTIONS: THE ROLE OF FEAR OF CRIME AND SEX-ROLE IDEOLOGY BY HELENE LENSKY FINKE Three theories--Defensive Attribution, Just-World, and Balance Theories--were reviewed and applied to an examination of the largely atheoretical research on attitudes about wife abuse. Using Saunders et al.’s Inventory of Beliefs About Wife Beating, the wife abuse attitudes of 173 female and 98 male undergraduates were compared with their sex-role beliefs and perceptions of vulnerability to crime to test the Defensive Attribution Theory hypothesis that perceived vulnerability would be instrumental in attributions about wife abuse. It was hypothesized that perceived vulnerability to crime would relate to wife abuse attitudes curvilinearly and also mediate the relationship between wife abuse attitudes and sex-role ideology. Moderately vulnerable participants were expected to perceive of wife abuse as most justified and to assign the least responsibility to wife abuse perpetrators. Consistent with Defensive Attribution Theory, males ‘moderately vulnerable to crime reported perceiving wife abuse perpetrators as least responsible. Unexpectedly, Ihcwever, it was the least vulnerable men who attributed more responsibility to perpetrators, while women did not show a <2urvilinear pattern. Thus, men’s perceived vulnerability to crime predicted attitudes toward wife abuse better than women's perceived vulnerability levels. Sex-Role attitudes strongly predicted all wife abuse measures, supporting Balance Theory and conflicting with Defensive Attribution Theory. More egalitarian sex-role ideology was consistently associated with disapproval of wife abuse. These findings were explored within the context of gender differences in both the relevance of wife abuse and in the social acceptance of perceived vulnerability. Limitations of this study were discussed, including the truncated distributions yielded by some of the measures. In addition, directions for future research were suggested. AGAIN, TO DAVE iv ACKNOWLEDGMENTS I wish to express my gratitude to Dr. John Hurley who helped to mold me, often times despite myself, in to a researcher. Our working relationship has developed over several years and he has been supportive and encouraging throughout while also being gently confrontational with me when I needed it. My graduate school experience was much enriched through my association with him. I would also like to thank Dr. Norman Abeles, Dr. Elaine Donnelson, and Dr. Michael Lindell for their assistance and their flexibility in the completion of this research. TABLE OF CONTENTS List of Tables ..................... ... .................. viii List of Figures ........................................... ix Introduction ....................... ................. ....... 1 Attribution Theory .................................... 2 Just—World Theory......................... ............ 5 Walster’s Theory of Attribution versus Just-World Theory ........................................... 6 Defensive Attribution Theory .......................... 8 Defensive Attribution Theory Versus Just-World Theory .......................................... 11 Balance Theory ....................................... l7 Attitudes Toward Woman Abuse/Rape and Attributional Theory ............................ 21 Demographic Correlates ............................... 32 Attitudes Toward Wife Abuse and Sex-Role Attitudes...33 Perceived Vulnerability to Victimization ............. 39 Hypotheses ................................................ 45 Method .................................................... 48 Participants ......................................... 48 Measures ............................................. 48 Inventory of Beliefs about Wife Beating ......... 48 Fear of Crime Scale ............................. 52 Sex-Role Ideology Scale ......................... 54 vi Procedures .......... . ..... .......... ..... .. .......... 56 Results ...... . ......... ..... ............ .. ....... .........58 Discussion. ........................ . ...... ..... ........... 69 Fear of Crime. ............. ...... ...... ......... ..... 69 Fear of Crime and Wife Abuse Attitudes....... ...... ..70 Sex-role Attitudes, Fear of Crime, and Wife Abuse Attitudes ............................... . ............ 76 Future Research ...................................... 78 Appendix A: Fear of Crime.......................... ...... 81 Appendix B: Inventory of Beliefs About Wife Beating (IBWB) .......... ................ ........... 84 Appendix C: Sex-Role Ideology Scale (SRIS) ............... 87 Appendix D: Chronbach’s Alpha Coefficients for Men and Women for the Inventory of Beliefs About Wife Wife Beating, Fear of Crime, Victimization Experience, and Sex-Role Ideology Scales ..... 89 List of References ........................................ 90 Table Table Table Table Table LIST OF TABLES Means, Standard Deviations and t-tests of Differences Between Men and Women on Fear and IBWB scaleSOOOO0.00....0.000.000.0000000000059 Multiple Regression Analyses of Fear with IBWB Justification and Responsibility Scales.........62 Multiple Regression Analyses of Fear of Crime and Sex-Role with IBWB Scales...................64 Correlations of IBWB Attitudes with SRIS Attitudes in High and Low Fear of Crime... ...... 65 Correlations of IBWB, SRIS, Fear of Crime, Property Crime, and Personal Crime Scales ....... 68 viii LIST OF FIGURES Figure l. The Influence of Sex-role on Perceptions of Victims as Predicted by Defensive Attribution Theory Versus Balance Theory ..... . ........ . ..... 34 ix INTRODUCTION Evidence of woman battering appears at least 3000 years old (Dickstein, 1988). American law and history shows that the practice of socially condoned hitting of women by men is longstanding. For example, the "Rule of Thumb Law", in effect in the United States as late as the end of the 19th century, stipulated that a man could beat his wife with a stick, provided it was not larger than the circumference of his thumb (Dickstein, 1988). Violence within the family is a problem of staggering proportions. According to Hilberman (1980), "The family is the most frequent single locus for violence of all types..." (p. 1337). Estimates based on FBI statistics approximate that a woman is beaten every 18 seconds (Dickstein 1988). And a 1986 study by the National Institute of Justice approximated that 2 million people are beaten by a spouse each year; 95% of these victims are women (Green, 1988). Prevalence statistics paint an even more dire picture. It has been estimated that between 25% (Pagelow, 1984) and 50% (Koslof, 1984) of all marriages include battering at l 2 some point. Straus (1980) found that 28% of his nationally representative sample of 2,143 couples reported at least one violent episode. The actual numbers of victims can never be known, but it has become increasingly clear that the victimization of women within the family is a widespread problem. How do people perceive this problem? The present work concerns the factors affecting attitudes toward woman abuse, focusing specifically on violence in marriage. One question is whether woman battering is abnormal or merely is a part of mainstream culture. If abnormal is defined as "not normal, average or typical; irregular" (Webster’s New World Dictionary, 1984), then wife abuse is abnormal only to the extent that it is culturally defined as deviant. Does our culture define wife abuse as deviant? Second, what factors predict an observer’s judgments about wife abuse? How are attributions about violence against women formed? The literature has focused primarily on attributions of blame to either victim or perpetrator and the tendency toward victim derogation. These areas will be reviewed within the framework of attribution theory. Attribution Theory As defined by Frieze (1979), attribution theory is "a theoretical perspective for understanding people’s ideas about why things happen". In measuring attitudes about wife 3 abuse, it is clear that some understanding of the way in which individuals make sense of events in their environment is necessary. The present paper reviews three main theories that have been employed to explain the way in which attributions about victims and perpetrators of woman abuse are made: Defensive Attribution Theory, Just-World Theory, and Balance Theory. Walster (1966), as a precursor to Defensive Attribution Theory, theorized that when a negative event occurs, the typical response on the part of victims and observers is an attempt to hold someone responsible. She argued that the more severe the negative event, the more likely we are to engage in this behavior of assigning responsibility, since minor events can be more easily dismissed. As a means of self-protection, Walster held that observers would attribute responsibility to an available target (often the victim) and view the victim as different than themselves, e.g. "things like that only happen to other people". A clear demonstration of this process has been observed in the wake of the catastrophic hurricanes that swept through Florida in the summer of 1992. Several news programs broadcast stories questioning whether building contractors and inspectors were to blame for the destruction of many of the homes in the storm’s path. According to Walster’s theory, these questions are part of the typical response in which individuals who observe or experience a 4 serious tragedy try to cognitively organize it by attempting to figure out to whom responsibility should be assigned. Walster (1966) reasoned that we engage in this process as a way to defend against the notion that catastrophes are often uncontrollable and can happen to anyone. To allay our anxieties, we believe that the world is ordered. In an empirical test of her theory, Walster (1966) presented 88 undergraduates (44 women, 44 men) with vignettes describing one of four different conditions: an accident in which a car owner suffers either minor (a small dent) or serious (car is totaled) consequences, and an accident in which the owner and bystanders suffer a minor (car misses bystanders and has only a dent) or severe (someone is hurt) outcome. Descriptions of the individual and his actual behavior were held constant across conditions. In addition, all scenarios included the potential for serious consequences. As predicted, subjects assigned significantly more responsibility to the individual in the two serious conditions than in the trivial conditions. In addition, participants in the severe conditions were significantly more likely to report that drivers have a moral responsibility to ensure that their cars are safe. Thus, in the more severe accidents, observers appear more likely to perceive that the responsible party had violated a moral contract. However, contrary to predictions that observers 5 would try to differentiate themselves from the individual in the more severe scenarios, participants were not more likely to see the individual in the severe conditions as more careless. Just-World Theorv Just-World Theory (Lerner, 1965, 1971; Lerner & Simmons, 1966) assumes, as does Walster’s theory (1966), that individuals have a need to believe that the world is orderly and that significant events are not random. Lerner extended this notion by hypothesizing that most people believe the world is just, and individuals get what they deserve. As a way to reconcile perceptions with their just- world beliefs, observers would be expected to see victims as responsible for their fate, either because of their behavior or their personal characteristics. In this way, people can maintain feelings of security because they preserve their perception that the world is an orderly and just place in ‘which good people and good deeds are compensated. In a test of Just—World Theory, Lerner and Simmons (1966) examined the perceptions held by 75 female llndergraduates of a fellow student whom they believed was Iteceiving painful electric shocks. Subjects were Siignificantly more likely to see the victim as less Eittractive when they believed that her suffering would <=<>ntinue. The victim received the lowest attractiveness 6 ratings when she was believed to be a "martyr", a condition which the researchers believed is most threatening to the observers’ belief that the world is just. Using a similar method, Lerner (1971) found that subjects who believed that the shock victim received $30 perceived the victim to be significantly more attractive than did subjects who believed the victim received no compensation. According to Just-World Theory, individuals tend to derogate recipients of unjust treatment (uncompensated suffering) so that they may see the event as deserved and therefore fair. In this way, observers can feel secure that the negative event would not happen to more attractive persons like themselves. Walster’s theorv of attribution versus Just-World Theory How do Walster’s (1966) and Lerner’s (1965) theories compare? Both theories assume that individuals have a need 'to believe that significant events do not happen in a :random, haphazard manner. So that they may feel safe and tsecure, individuals tend to see events as having :identifiable causes. Thus, they attribute responsibility thor events to some source other than chance. Where the tuneories diverge is Lerner’s (1965) proposal that assignment Crf'responsibility is not the only factor in the cognitive <>13ganization of negative events (Chaikin & Darley, 1973). chzcording to Lerner, people actively attempt to restore justice in their minds. Using the case of the Florida hurricanes as an example, assigning responsibility for the disaster to building contractors would not restore justice if the victims were not either derogated or compensated. Just-World theory predicts that the uncompensated victim would be derogated. In contrast, Walster’s theory predicts that observers will assign responsibility to a target other than chance, while also differentiating themselves from the victim. Thus the response of an observer on the east coast might be to blame the building contractors while believing that hurricanes happen predominantly on the west coast. Clearly, this response is not motivated by a need for justice, but rather a need to believe that one's environment is predictable and orderly. Neither theory has been definitively supported. 'Walster (1967) was unable to confirm the positive and significant correlation between severity and attribution of :responsibility reported in her original study (Walster, .1966). The scenario in the 1967 study described a home <3wner who either serendipitously lost or gained differing amounts of money. Shaver (1970) posited that the failure to Iiind the severity-responsibility relationship may be due to ‘tlle fact that the stimulus character may have been too <3tifferent from that of the college student participants. UPIlus, the participants may not have had the concern that the 8 outcomes described could actually happen to them (are "relevant"), a necessary condition for "defensive attribution" (Shaver coined the term) to occur. Defensive Attribution Theory Shaver (1970) differentiated two types of similarity: situational and personal. Situational similarity refers to the fact that observers believe they could encounter the same situation as the stimulus person; this type of similarity is necessary for defensive attribution to occur. Personal similarity refers to shared beliefs and values between the observer and stimulus person. Walster’s theory states that, in the face of situational similarity, observers would try to differentiate themselves from the victim. Shaver (1970) hypothesized that, when both situational and personal similarity are high, it would be difficult for observers to attribute responsibility to the victim and avoid the realization that the catastrophe could befall them. In this situation, despite the necessary relevance to the observers, they would be less likely to hold the victim accountable and more likely to attribute responsibility to environmental stimuli or chance. In this case, observers are motivated primarily by a desire to avoid liability should they find themselves in the same predicament as the stimulus person, rather than by a need to defend against the possibility that the event might happen 9 to them. To test this hypothesis, Shaver (1970) virtually replicated Walster's (1966) study with one modification. Similarity to the stimulus person was manipulated by asking the subject to imagine that the personal characteristics of the stimulus persons were either "very similar to your own" or "not at all like your own". As predicted, individuals were significantly less likely to attribute responsibility to a similar person. In addition, when asked to evaluate conscientiousness and carefulness, participants evaluated similar targets more leniently than dissimilar ones. Thus, defensive attributions appear two-sided, with individuals wanting to believe they can avoid a negative situation, but also wanting to believe that they would be perceived as blameless should they end up in one. As Lerner (1965, 1970) and Walster (1966) predicted, individuals tended to attribute responsibility to those who are sufficiently dissimilar from themselves, so that they might believe that negative events tend to happen to persons different than themselves. In the face of personal similarity, however, observers appear more inclined to think that they would not be held accountable in such situations. Avoidance of blame thus supersedes the avoidance negative outcomes. And, in this latter condition, observers would attribute the outcome to chance, in contrast to Lerner and Walster’s hypotheses. 10 Defensive Attribution Theory presumes that individuals may alter their perceptions of events to reduce their perceptions of threat. However, Thornton (1984) noted that there had been no empirical evidence actually demonstrating the role of negative arousal in the defensive attribution process and performed two experiments designed to investigate the impact of negative arousal on attributions of responsibility to a sexual assault victim. Before reading an account of a sexual assault, 24 of the 48 female undergraduates in study I were given a special explanation for anxiety or apprehension they might feel during the study. They were told that psychological experiments often cause participants to feel nervous and uncomfortable. The other 24 received no such information. Thornton (1984) found that those given the special explanation for negative arousal were significantly less likely to attribute responsibility to the sexual assault victim than the others. Thornton (1984) reasoned that participants in the special condition had attributed their anxiety to the experiment rather than to the victimization. Having done so, they had a lesser need to see the victim as responsible for her fate, since holding a victim responsible is one way to reduce feelings of threat and anxiety. Thornton’s (1984) second experiment demonstrated that increasing a subject’s self-awareness, through the use of mirrors, led to increased attribution of responsibility to 11 the sexual assault victim. It appears that the heightened self-awareness made individuals more sensitive to their emotional reaction, causing an enhancement of the defensive attribution process. Taken together, these two works provide strong support for the defensive attributional hypothesis that enhanced feelings of threat induce the altered perceptions that have been found in studies investigating response to victimization. Defensive Attribution Theory Versus Just-World Theory Chaikin and Darley (1973) designed a study to empirically compare Defensive Attribution and Just-World Theories. Forty male undergraduate students viewed a videotape in which a "worker" assembled building blocks on a table while being observed and evaluated by a "supervisor". In the tape, the supervisor bumps into the table as he stands up, knocking over the blocks that the worker had assembled. The severity of this episode was manipulated by leading the subjects to believe that the worker would lose either a large or small portion of what they were to be paid as a result of the accident. Situational similarity was manipulated by telling subjects either that they would later assume the role of the worker or the role of the supervisor. As predicted by both Walster's (1966) and Lerner’s (1965, 1971) theories, participants were significantly less likely to attribute responsibility for the accident to 12 chance in the severe condition than in the minor condition. As Shaver’s Defensive Attribution Theory predicts, people chose to assign responsibility to the target that allowed them to feel the least threatened. Thus, "future supervisors" attributed significantly less responsibility to the supervisor than did "future workers". Future supervisors in the severe condition were significantly more likely than subjects in other conditions to assign responsibility to the equipment, an attribution that does not allow for justice. Also inconsistent with Just-World Theory, future workers in the severe condition were significantly more likely to hold the supervisor accountable, another attribution that does not provide for justice. Justice would be achieved if the worker/victim were seen as responsible for his fate, as this would permit the perception that the worker's reduction in pay was deserved. However, this would threaten the future worker who believes that he is likely to encounter the same fate. Thus, while a person appears to have the need to believe that the world is just, "he will occasionally deny justice for the sake of his own security" (Chaikin & Darley, 1973; p. 274). Examining mothers’ attributions of responsibility for their child’s illness, Tennen, Affleck, and Gershman (1986) investigated the applicability of Defensive Attribution Theory. As predicted, maternal self-blame correlated 13 positively and significantly with severity of illness and similarly with improved mood and beliefs that future illness could be avoided. These mothers appeared to view themselves as responsible as a means to bolster their perceptions of control. Path analysis revealed that this increased control, in turn, led to positive mood. This finding is consistent with the research of Janoff-Bulman and Wortman (1977) who found that self-blame was significantly correlated with good coping skills among 29 paralysis victims. Somewhat counter-intuitively, the self was the least threatening target to whom responsibility could be attributed. Consistent with Defensive Attribution Theory, but clearly contradicting Just-World Theory, accountability was attributed to this least threatening target. Thus, as in Chaikin and Darley’s study (1973), these mother appear to have forsaken the justice motive for feelings of increased security. Also comparing Defensive Attribution Theory and Just- World Theory, Shaw and Skolnick (1971) examined whether a "happy accident" was more likely to be attributed to a target person or to chance. Just-World theory predicts that subjects would attribute an event (whether positive or negative) to the target person, since "people get what they deserve". In contrast, Defensive Attribution Theory predicts that subjects would attribute a positive event to chance, while attributing a negative event to the target. 14 In this way the observers can believe that the good fortune could happen to them, while believing that the negative event could not. The findings were consistent with Defensive Attribution Theory. Also supporting this theory was the finding that males were significantly more likely to see the positive event as a chance occurrence and as something that could happen to them than was true for the negative event. Also supporting Defensive Attribution Theory was the finding that male subjects were significantly more likely to attribute responsibility to the target when the negative event was serious rather than minor. The reverse was true for positive events; attribution of responsibility to the target individual was significantly less likely in the case of a more serious positive event. Thus, subjects appeared to feel more secure believing that ‘very negative outcomes happen largely to victims that deserve it, whereas very positive outcomes might happen to anyone. That these findings were true only for males was explained as due to fact that the use of a male target individual rendered the situations less relevant to females. However, Schiavo’s (1973) later replication using a female target also failed to find a severity-responsibility link among females, suggesting a systematic gender difference. Sloan and Gruman (1983) investigated the impact of Observers' perceived vulnerability on ratings of victim 15 attractiveness among 73 undergraduates. Using the case of illness, they unexpectedly found that vulnerability, as measured by estimations of risk for disease, was unrelated to attractiveness ratings. Thus, increased risk did not lead to victim derogation as Defensive Attribution Theory predicted. Just-World Theory also failed to fit, as subjects rated potentially culpable patients (those with preventable diseases) as significantly less attractive than innocent patients (those with unpreventable diseases). Thus, neither theory readily explained Sloan and Gruman’s (1983) findings. However, subjects who actually knew someone afflicted with the disease at issue, a relationship likely to increase perceived threat, did tend to rate illness victims as less attractive (p = .07). Although somewhat short of statistical significance, this finding suggested that more vulnerable subjects did demonstrate an increased likelihood to differentiate themselves from victims via derogation as predicted by both theories. Coates, Wortman, Abbey and Holland (1979) investigated observers’ reactions to rape victims as a function of their own perceived vulnerability and the emotional reaction of the victim. Pertinent to the present study is their finding that observers who reported feeling more vulnerable to rape (females versus males) were more likely to empathize with the victim but were also more likely to see her as responsible and derogate her. These findings, in contrast 16 to Sloan and Gruman’s (1983) findings, are consistent with Lerner’s (1967, 1971) and Walster’s (1966) hypothesis that increased vulnerability leads to attributions of responsibility to, and differentiation from, the victim. Coates et al. (1979) noted that only two levels of vulnerability had been used in their study, and the same is true for Sloan and Gruman (1983). Shaver (1975) had suggested a curvilinear relationship between observers’ perceptions of vulnerability and reactions to victims. In accordance with Walster’s and Lerner’s positions, Shaver hypothesized that subjects would be indifferent toward victims if they felt minimal threat and would attribute responsibility to them and derogate them only if they experienced some degree of vulnerability. However, Shaver hypothesized a vulnerability threshold, which, once exceeded, would tend to focus the observers on their own <:oncerns about victimization. In turn, observers would then 'tend to be more lenient in their judgments about other victims. Burger (1981) performed a meta-analysis of Defensive Attribution Theory studies encompassing works conducted within the 12 years following Walster’s pivotal 1966 study. After examining 22 studies, Burger concluded that " although it is probably true that the occurrence of the Walsterr (1966) main effect for accident severity on Perpetln1tor responsibility measures is more frequent than 17 that expected by chance alone, the strength of the relationship may still be less than some investigators find comfortable" (p. 501). Burger also investigated the other main tenet of Defensive Attribution Theory, that observers are motivated toward greatest self-protection. Thus, studies investigating the variables of personal and situational similarity (i.e., relevance) were also examined. Analyses revealed that studies which investigated both severity and similarity demonstrated virtually universal (one exception) consistency with Defensive Attribution Theory. Balance Theory Studies (e.g., Gentemann, 1984, Lavoie, Jacob, Hardy, & Martin, 1989) have shown that seeming responsibility on the part of the victim largely influences the perception of ;perpetrator culpability. Heider (1958) hypothesized that attributions of responsibility were based on the observer’s estimation of the extent to which the person or the environment affected the outcome. For example, one’s ability to achieve an outcome is weighed against the Opposing environmental forces. In addition to ability, motivation is necessary for the completion of a desired outcome; and this is also considered in attributions. Pheider (1958) identified several different components in attributions of responsibility: the degree to which the 18 individual is associated with the outcome and directly caused it, the degree to which the outcome was intentional, and the degree to which the outcome was justifiable. Sulzer (1971; cited in Shaver 1970) organized Heider’s levels of attribution in the following manner: 1. Association - responsibility is attributed to an individual simply because she is in some way, no matter how remote, associated with an outcome. 2. Causality - responsibility is attributed to an individual because she was a necessary factor in its occurrence, intention and motivation are not considered. 3. Foreseeability - responsibility is attributed to an individual because she was a necessary factor and could foresee the outcome; motivation and intention are not considered. 4. Intentionality - responsibility is attributed to an individual because she was motivated to cause the outcome. 5. Justifiability - responsibility is apt attributed to tflme person despite the motivation to cause the outcome because something in the environment necessitated the act. (fidhn and Sugarman (1982) studied the extent to which 19 observers used Heider’s (1957) attributions in their judgments about victims and perpetrators of wife abuse. Seventy-two volunteers read vignettes depicting either physical or sexual abuse of a wife by her husband and then responded to a Likert-formatted questionnaire soliciting their attributions of responsibility. In the area of offender responsibility, male and female subjects reported significantly differing importance for the various levels of attribution in their judgments for all levels except justification. That is, men and women appeared to base their attributions of responsibility on different factors. For example, with regard to victim responsibility, female subjects perceived the victim’s intentions as playing a larger role in responsibility for sexual than physical assault. Heider (1958) had argued that our "sentiments" about (others influence our attributions about them. The noted tgender differences may be partially explained by gender- dependent differences in such sentiments toward a female victim. In our perceptions of others, we tend toward a balanced state. That is, if persona feels positively toward a stimulus target, they will tend to interpret that individual’s behavior positively to maintain a balance. If the valence (e.g., positive or negative) of the target indivixlual and behavior do not match, then the persons will reinterpret their perceptions of either the person or the 20 event. Thus, the observers’ sentiments about an individual strongly affect their interpretation of stimuli that impinge on that individual. The observation that negative events happened to disliked individuals would preserve balance and be perceived as deserved, while negative occurrences to liked individuals would tend to be perceived as miscarriages of justice. In the situation of wife abuse, Heider’s theory predicts that the observers’ identification with either the perpetrator or the victim will affect to whom they would attribute responsibility and/or derogation. From this perspective, one would anticipate that the extent to which observers identify with women’s issues would influence their perceptions of women who are abused. Studies examining attitudes about victims have tended to use potentially responsible victims. For example, Walster’s (1966) study depicted an individual whose car rolls down a hill, and Shaver’s (1970) experiments portrayed a researcher called away for a phone call while leaving machines running. Although these scenarios are quite distinct from the wife abuse situation, it is true that a battered woman, like the stimulus characters used in the earlier studies, is rarely seen as completely free from responsibility. It is implied that she had some choice in selecting her partner and has typically engaged in verbal arguments with him. This is not to say that a woman is 21 responsible for her abuse. It must be mentioned, however, that responsibility in domestic assaults is often perceived as more ambiguous than stranger assaults (Hilton, 1989), with the victim rarely being viewed as wholly free from accountability. It is expected that studies using wife abuse scenarios would likely yield findings similar to the earlier Studies on attribution that involved different forms of victimization. Attitudes Toyard Woman AbusezRape and Attributional Theory Do the predictions of the various attribution theories apply to perceptions of wife abuse? There has been much recent interest in the area of attitudes about wife abuse, with several studies investigating the types of attributions made by "observers" as they respond to questions about wife abuse. In this context, the term observer will be used to mean persons not involved in wife abuse. Thus, observers as defined here have not necessarily seen incidents of actual abuse, but have merely learned of it through vignettes given to them by researchers or by other means in their daily lives. Attitudes about wife abuse are important because our perceptions of this problem affect the way we respond to it. For example, Gentemann (1984) drew attention to the fact that observers often hold victims responsible, which promotes silence on the part of victims. Thus, the 22 attitudes held among the persons to whom victims turn (friends, family, mental health professionals, police, lawyers) has a great impact on whether or not victims obtain help and also whether offenders will be punished. Legislation designed to aid victims and punish offenders is not likely without the support of public opinion. Legislators would be unwise to act in a manner wholly discrepant with the general attitudes and values of the majority of their constituency. In addition, attitudes among the general public are of the utmost importance when jurors are required to render judgments on cases involving wife abuse (Ewing & Aubrey (1987). Few researchers (e.g., Kristiansen & Giulietti, 1990; Miller, Smith, Ferree, & Taylor, 1976) have incorporated attributional theory in interpretations concerning wife abuse data. The present review of the attitudinal literature evaluates the usefulness of the various theories in explaining the findings wherever information is sufficient to do so. It provides an application of attribution theory to the research on attitudes about wife abuse. Hilton (1989) compared the attitudes about assaults on strangers versus wives. After reading vignettes describing an assault, 240 participants (120 women, 120 men) rated how serious and dangerous the incident was and to whom they attributed responsibility. From a list of ten options, they 23 also chose what actions they believed the police should take. The police record of the perpetrator, the location of the attack, and the relationship of the attacker and victim (spouse versus stranger) were varied. Judged seriousness of the assault was significantly and positively correlated with the attacker’s record, with attacks by repeat offenders being viewed as the more serious. However, only when the victim was a stranger were repeat offenders seen as more responsible than first-time offenders. Furthermore, repeat non-domestic assailants were also significantly more likely to be viewed as warranting legal punishment than were domestic offenders with the same characteristics. When asked whether a crime had been committed, 81% responded affirmatively regardless of whether the victim was a stranger or wife. Thus, even though there was agreement that a serious law violation had occurred, there was a significantly greater likelihood that wife abusers, as compared to stranger abusers, would be seen as less responsible for their actions and less appropriate for criminal charges. Also noteworthy was that Walster’s (1966) severity- responsibility relationship held only in the case of stranger assaults. Both Just-World and Defensive Attribution Theories predict that more serious events are less likely to be attributed to chance because to do so would be threatening. The fact that wife abusers who were 24 perceived to have committed a more serious offense were not seen as more responsible suggests that wife abusers may be seen as less threatening to observers than stranger abusers. However, no data on perceived vulnerability to the two types of crime were obtained. Edwards (1987) conducted a statewide telephone poll of 1,018 people in rural, urban, and suburban areas of New Jersey. When asked what the phrase "domestic violence" refers to, more than 80% of respondents expressed some knowledge of domestic violence, defining it as some type of family conflict. Responses indicated that most people defined it as physical rather than emotional abuse. Child abuse was specified as the most common type of domestic violence, with wife abuse following closely. However, only 20% of those polled perceived wife abuse as a "very common problem"; and even fewer believed that violence in relationships outside of marriage was a major problem. Thus, as in Hilton’s (1989) study, perceived vulnerability to this problem appears to be low. About two-thirds of those New Jersey residents preferred counselling to punishment for assailants, a finding similar to Hilton’s (1989). Correlational analyses revealed that those who preferred counseling were most likely to be male, less educated, or older. While most respondents viewed wife abuse as a problem belonging outside of the criminal justice system, less than 15% considered domestic violence to be a 25 "private problem" warranting no intervention. Koski and Mangold’s (1998) telephone interviews of 176 Arkansas residents found that females were significantly more likely than males to see public intervention (e.g., police, crisis lines, shelters) as an effective response to family violence. Although both males and females preferred private responses (e.g., family, clergy, etc.), males were significantly less likely to view family violence as a serious problem and to endorse public responses. Gentemann (1984) investigated attitudes toward wife abuse among 422 women over 18 in North Carolina. Respondents were contacted through random digit dialing and surveyed by telephone as part of a larger study on a broad range of women’s issues (e.g., menstruation, Equal Rights Amendment, etc.). Participants were asked whether the following behaviors on the part of a husband toward his wife are "okay ... once in a while": swearing, slapping, punching, kicking. Although swearing was perceived as justified by 40% of respondents, the other behaviors received virtually no approval (range of 1%-3%). These North Carolina women appeared to disapprove of physical force in marriage and View it as deviant. When participants asked whether "a man has a right to beat his wife" in response to four different behaviors on the part of the wife (flirting with other men, sexual infidelity, drunkenness, nagging), Gentemann’s results changed significantly. In 26 contrast to the almost complete disapproval when no context was provided, almost 20% of the respondents reported believing that violence was justified in at least one of the situations. This finding is strikingly similar to the results of a study done more than one decade earlier (Stark & McEvoy, 1970) in which approximately 20% of a sample of 1,176 Americans reported approval of slapping a spouse in certain circumstances. Similarly, Lavoie et al. (1989) found that police officers were significantly more likely to hold a victim responsible for her attack if she had verbally antagonized her husband. Thus, nearly 20% of Americans appear to believe that hitting one’s spouse is appropriate if that spouse has committed certain seemingly provocative behaviors. These findings contradict Just-World Theory which predicts that an innocent victim would be held more responsible and/or derogated more than a potentially culpable victim so that her attack may appear justified. Similarly, Sloan and Gruman (1983) found that observers rated patients with preventable diseases as significantly less attractive than seemingly "innocent" patients who had diseases they could have done nothing to avoid. Shaver (1975) hypothesized that people will try to attribute responsibility to an appropriate target; and provocative victims may appear more appropriate than innocent ones. As Walster (1966) stated, "... we do not expect the subjects to 27 surrender all standards of judgement and habits of reasonableness in order to satisfy their desires to assign responsibility." (p. 74). Regardless of the explanation for attributions, it is clear that attributions of responsibility for wife abuse can apparently shift from the husband to the wife under certain conditions. One may assume that the decreased responsibility attributed to wife assailants as compared to stranger assailants in Hilton’s (1989) study may be explained by an attribution of some accountability to the wife. Investigating public perceptions of wife abuse, Ewing and Moss (1987) presented a vignette depicting marital violence to 216 subjects whose participation was solicited via a mass mailing. After reading the vignette, subjects were asked for true or false responses to eight statements about woman abuse. About 35% endorsed statements that a battered woman is "masochistic", "emotionally disturbed" or "bears at least some responsibility". Approximately two- thirds of the sample endorsed the belief that a battered woman "could simply leave", with females being significantly more likely than males to endorse this viewpoint. Women were also significantly more likely to report believing that the battered woman in the scenario was masochistic while males were significantly more likely to attribute responsibility to the victim and to believe that the assault 28 was not part of a pattern of abuse. These gender differences can be readily explained according to Shaver’s (1970, 1975) Defensive Attribution Theory. That female participants derogated the victims and saw them as choosing to be abused may be interpreted as an attempt by the female subjects to reduce their feelings of vulnerability by differentiating themselves from the victims and believing that they are in a negative situation only because of choice. These findings could also be explained by the Just-World view that victims deserve their fate because they are unlikable women who make poor judgments. However, when the responses of the female participants are compared to those of their male counterparts, Just-World Theory does not adequately explain the pattern of gender differences. The tendency of men to see the perpetrators as less responsible does not restore justice. However, it can be argued that the males formed these perceptions as an attempt to avoid responsibility should they commit an offense of this type. Similarly, the females may have attributed more responsibility to the perpetrators so that they could avoid feeling responsible should they ever be victimized. Thus, both the male and female participants appear to have perceived wife-battering in a manner that would make them feel least threatened, although this study did not obtain perceived vulnerability data. Thus, we can only assume that 29 female respondents felt more vulnerable to the possibility of being victimized, and that these men were more threatened by the potential for negative consequences if they were to perpetrate a similar offense. Janoff-Bulman (1982) examined college students’ perceptions of rape victims using actual police accounts. In the "observer" condition, participants were simply instructed to read the account, while in the "victim" condition, participants were instructed to read the account and imagine themselves to be the victim of the described rape. Through this role-play design, the author tried to assess the reactions of observers versus victims of violent crimes. This manipulation allows for an examination of the effects of relevance on observers attributions regarding victims. Janoff-Bulman (1982) identified two types of attributions of responsibility: characterological and behavioral. In the first type, the person’s personality and character are maligned (e.g., she is stupid); in the latter case, the victim’s behaviors are seen as causal (e.g., she should not have gone out alone after dark). Shaver and Drown (1986) referred to characterological and behavioral attributions as attributions of responsibility and causality respectively, preserving Heider’s differentiation of the different levels of attribution. According to Shaver and Drown (1986), causality does not necessarily imply 30 responsibility; thus an individual may attribute causality to the self without feeling responsible for the rape. As hypothesized, self-derogation or characterological self-attribution among victims was correlated with low self- esteem, while behavioral self-attribution was not. This is consistent with other recent findings (Andrews & Brewin, 1990; Sweeny, Anderson, & Bailey, 1986) demonstrating that internal, global, andstable attributions for one’s misfortunes are linked with depression. Clearly, characterological self-attributions are relatively global and stable. In contrast, behavioral self-attributions (characterized as more specific and unstable) were not correlated with low self-esteem and were, in fact, correlated with high self-esteem and also with perceived ability to avoid future rape. Janoff-Bulman reasoned, in accord with several other researchers (e.g., Shaver & Drown, 1986; Tennen, Affleck, & Gershman, 1986), that behavioral self-attribution is an adaptive strategy for victims because it allows them to restore feelings of power and control in their lives, an idea similar to Shaver’s (1970) Defensive Attribution Theory. Rather than simply trying to restore justice by maligning their own character, victims attribute responsibility to the target that is the least threatening. In this case, attributing the responsibility for their misfortune to their own behavior allows victims to resume a 31 sense of control and orderliness in their lives; and this occurs without a sense of justice. For observers, however, both characterological and behavioral attributions of responsibility were related to high self-esteem. Thus, for the crime of rape, it appears that the derogation of the victim’s character or behavior is an adaptive strategy for observers. In contrast to the experience of victims, characterological and behavioral attributions of responsibility are equally effective in providing a sense of security for observers. What this means then, is that derogating the victim appears to have harmful effects for the victim, while fostering positive effects for the observer. Also investigating observers’ reactions to rape, Yarmey (1985) compared responses of 192 male and female college students with an equal number of persons over 50. After learning about a rape via slides, photographs or vignettes, these students attributed significantly more responsibility to the victim than did their elders. Consistent with similar research (e.g., Kristiansen & Giulietti, 1990; Miller et al., 1976), "provocative" victims were regarded as significantly more responsible than "innocent" victims by both age groups. This finding was inconsistent with the Just-World Theory prediction that observers’ increased responsibility attributions to, and derogation of, innocent victims serves as a means to preserve the belief that the 32 world is just and victims deserve their fate. Yarmey (1985) hypothesized, consistent with Defensive Attribution Theory, that the students may have attributed more responsibility to the victims because they identified more with them and thus felt more at risk of victimization themselves. In turn, to reduce their sense of vulnerability, they perceived the victims as more responsible. This hypothesis is notably speculative, however, since no identification or perceived vulnerability measures were obtained. Taken together, these empirical studies regarding both criminal and noncriminal victimization tend to provide more support for Defensive Attribution than Just-World Theory. Although some aspects of the research fit either theory, Defensive Attribution Theory appears more accurate in situations in which these two theories predict opposing outcomes. Demographic Correlates Addressing different populations, Aubrey and Ewing (1989) compared the attitudes toward battered women held by college students (mean age 18) with those held by registered voters (mean age 42). Students endorsed significantly fewer Stereotypes, such as "If she remains, she is emotionally disturbed" or "If she remains, she is masochistic". Persons over age 50 were significantly more likely to believe that 33 victims had some responsibility and were masochistic. To ascertain who is more likely to approve of marital violence, Gentemann (1984) correlated attitudes with various demographic variables. Older, non-Caucasian women with less education and lower incomes were found_to be more likely to approve of violence, a finding consistent with other studies in which older persons were shown to be more likely to adhere to stereotypes regarding wife abuse (Aubrey & Ewing 1989; Ewing & Aubrey, 1987) and less likely to advocate arresting offenders (Edwards, 1987). Unlike those studies, however, the age effects found in Gentemann’s study dissipated when education level was controlled. Only race and income maintained their relationship to justification. Thus, consistent with Yarmey (1985), this study provides little support for the hypothesis that younger women would be less likely to see the victim as responsible and more likely to disapprove of violence than would older women. Attitudes Toward Wife Abuse And Sex-Role Attitudes Several studies have investigated whether sex-role attitudes predict approval of wife abuse and attributions of responsibility for this type of violence. Heider’s (1958) balance theory predicts that those who identify with women’s causes would be more likely to see male perpetrators as responsible and female victims as innocent. Extending Heider’s theory, more feminist observers would be expected 34 to endorse punishment of offenders. In contrast, Shaver (1970, 1975) hypothesized that moderate relevance to the observer, via personal similarity (e.g., feminist attitudes) will elevate the observer’s sense of vulnerability. This, in turn, will cause the observer to derogate similar persons and hold them responsible. The two figures below illustrate how these two theories differ. Defensive Attribution Theory Low< -------------- Victim Responsibility ---------------- >High high relevance low relevance moderate relevance (high feminism) (low feminism) (moderate feminism) Balance Theory Low< -------------- Victim Responsibility ---------------- >High high relevance moderate relevance low relevance (high feminism) (moderate feminism) (low feminism) Figure 1. The Influence of Sex-Role on Perceptions of Victims as Predicted by Defensive Attribution Theory Versus Balance Theory. As can be seen, Balance Theory predicts a direct linear relationship between attitudes toward women and attitudes about victims, while Defensive Attribution Theory predicts a curvilinear relationship. The research literature has focused predominantly on Heider’s hypothesis of a linear relationship, predicting that sex-role attitudes will be positively and significantly correlated with attitudes toward wife abuse. Several researchers have hypothesized that woman battering is, at least in part, an outgrowth of the 35 inequality of women. Finn (1986) empirically examined the relationship between attitudes regarding sex-roles and those regarding woman battering. Among 300 college students, he found that the males endorsed traditional sex-roles significantly more than did the females, and were also more likely to endorse the use of force in marriage. However, gender itself had no direct relationship with approval of wife abuse. According to Finn (1986), "... a traditional sex role orientation is the strongest predictor of attitudes supporting marital violence, while sex and race per se play a relatively small role" (p. 241). Thus, these findings supported Heider’s (1958) Balance Theory. Kristiansen and Giulietti (1990) studied the interaction of several variables in college students’ perceptions of wife abuse: attitudes toward women, belief in a just world, and "provocation" by the victim. After completing written measures of attitudes toward women and just-world beliefs, participants read one of two domestic dispute scenarios presented as excerpts of police records. One scenario included "provocation" by the victim; the other did not. Consistent with prior research, men had significantly more traditional attitudes toward women and significantly stronger Just World beliefs. As Balance Theory predicts, males’ perceptions of wife abuse were consistent with their attitudes toward women. As in Finn’s (1986) study, 36 however, it was attitudes regarding sex-roles, rather than gender, that predicted derogation of the victim and the degree to which she was seen as responsible. Females were more likely to attribute responsibility to the "provocative" victim than were males. In addition, women with more progressive attitudes were more likely to attribute responsibility to the "provoking" victims as their Just-World beliefs increased. Thus, perceptions of wife abuse appeared to be contingent both upon attitudes toward women’s roles and their beliefs about the world as fair and just. Men’s perceptions of wife abuse, however, appeared to be primarily motivated by a desire for cognitive balance and thus by attitudes about the equality of women. Kristiansen and Giulietti (1990) suggested a "controllability" hypothesis as a means to explain the interaction found between female participants’ Just-World beliefs and attitudes toward women’s roles as predictors of their attitudes about wife abuse. They explained this interaction in the following manner. Strong just-world beliefs have been shown to correlate positively with a more internal locus of control. This control need may render them less willing to tolerate feelings of vulnerability, since vulnerability involves an inability to control one’s fate. More egalitarian females would feel most vulnerable in response to the provocation scenario because they would ‘be most likely to fight back if abused. Thus, egalitarian 37 females with a high need for control would be more likely to see a victim as responsible for her fate. Kristiansen and Giulietti (1990) hypothesized that the attributions of responsibility to the victims were due primarily to the need for control, giving only minor attention to the fact that the gender difference is linked to differences in perceived vulnerability. The control hypothesis is not a new one; an aversion to lack of control has been addressed by several theorists (e.g., Lerner, 1966, 1971; Walster, 1966; Shaver, 1970, 1975). However I know of no studies investigating the hypothesis that vulnerability interacts with one’s identification with the victim to alter perceptions of that victim. Defensive Attribution Theory predicts that perceived vulnerability would impact the observer’s perceptions of victims; and this vulnerability is associated with the observers’ degree of identification (relevance) with such victims. An earlier study by Miller et al. (1976) examined the same variables studied by Kristiansen and Giulietti (1990): attitudes toward women’s roles, belief in a just world, and victim responsibility (rather than provocation). They used one-minute videotapes depicting a female patient telling a physician about her physical injuries. The variable of responsibility for the injury was manipulated by presenting scenarios of fabricated case histories in which the patient was described either as: a "culpable driver", an "innocent 38 pedestrian", or a rape victim (intended as ambiguous on the continuum of responsibility). Judgements of the victim’s "responsibility" for her injuries correlated negatively with the rater’s perceptions of her likability and intelligence, contradicting Just-World Theory. Contrasting later studies (Finn, 1986; Kristiansen and Giulietti, 1990) which found that attitudes toward women’s roles (rather than gender) predicted response to victims, Miller et al. found that feminist ideology related to identification with the victim only for women. For men, feminist beliefs did not appear to influence their attitudes toward the victims. Miller et al. reasoned that women were more likely than men to identify with female victims and react with sympathy. They argued that it was identification with either the victim or the perpetrator that best predicted observers’ reactions, as Heider’s (1958) theory would predict, not Just-World beliefs. In their view, observers simply try to make a situation appear just in their mind as they look back on it; and this process is interwoven with identification. However, as in Kristiansen and Giulietti’s (1990) study, those with stronger Just-World beliefs were more likely to derogate the victim. Also consistent with Kristiansen and Giulietti’s results, Just-World beliefs correlated negatively with feminist values. This relationship is not surprising, because feminist values are related to one’s 39 awareness of gender inequalities. How can the inconsistent findings regarding the relationship between attitudes toward women and attitudes toward wife abuse be explained? One hypothesis that might account for these discrepancies is the construct of perceived vulnerability. Defensive Attribution Theory predicts that perceived vulnerability will mediate the identification-perception link described by Heider (1958). This variable has not yet been explored in studies examining sex-role ideology and attitudes toward wife abuse. Perceived Vulnerabilitv to Victimization Numerous studies have shown that women express significantly more heightened feelings of vulnerability to crime than men, despite the fact that they seem less at risk for most crimes (Gomme, 1988; Gordon, Riger, LeBailey, & Heath, 1980; Junger, 1987; M.D. Smith, 1988; G. Smith, & Lab, 1991; L. Smith, & Hill, 1991). Studying 640 subjects in three major Canadian cities, Gomme (1988) found that gender surpassed prior experience with victimization (either first hand or "vicarious") as a predictor of feelings of vulnerability to crime. Thus, risk of victimization, as measured by crime rates, appears to be less predictive of perceived vulnerability to victimization than one might expect. Researchers investigating perceived vulnerability to 40 crime have identified two distinct aspects: the cognitive assessment of risk and the emotional reaction of fear (Gordon et al., 1980; Junger, 1987; Warr, 1987). As stated above, the relationship between these two factors is not necessarily linear. While individuals may believe they are at high risk of being pick-pocketed, they may not report a related high level of fear. Warr (1987) described a "sensitivity to risk" factor as "the relation between fear of a particular offense . . . and the perceived risk of that offense (i.e., the subjective probability that it will occur)". (p. 30). Using an inventive design, he compared the sensitivity to risk for different types of crime among 339 Seattle residents. The respondents’ fear of various crimes was plotted on a graph against their perceived likelihood of victimization for each crime, allowing for the slope of the relationship to be used as an indicator of sensitivity to risk. Sensitivity to risk, as measured by slope, correlated highly (; = .80) with the perceived seriousness of crime. Rape, perceived as among the most serious of crimes, was shown to elicit the highest sensitivity to risk of the eight crimes investigated. These findings regarding rape may explain women’s higher levels of perceived vulnerability. Rape is one crime for 'which women are undoubtedly at greater risk and it is a crime that instills more fear than most others (Warr, 1987). Are women’s feelings of vulnerability regarding 41 victimization predominantly a fear of rape? Gordon et al. (1980) investigated perceived risk and fear of rape and also perceived risk and fear of other crime as a whole among 299 women and 68 men in Chicago, Philadelphia, and San Francisco. Males were asked to rate the risk of rape for women but not for themselves. All respondents were also asked to assess their own strength and ability to defend themselves. Results indicated that perceived risk of rape was a significant predictor of fear of crime for all respondents. In addition, physical vulnerability was an "independent and additive" contributor to the women’s fear response. Thus, lower levels of physical strength combined with the fear of rape appeared to account for much of the increased fear of crime demonstrated in women as compared to men.1 Other researchers have posited that other crimes largely directed at women, such as sexual harassment (Junger, 1987) and domestic violence (M.D. Smith, 1988), are also partially responsible for women’s elevated perceptions of vulnerability to crime. M. D. Smith (1988) found limited support for this hypothesis in a study of the relationship 1It is puzzling that males’ estimated risk of rape for women was also found to contribute to increased fear of crime in men. This study’s requirement that men focus on risk and fear of rape undoubtedly made these concerns more salient to men than is generally the case in their daily lives. Branscombe and Deaux (1991) demonstrated that making an attitude more accessible can increaserthe impact of that attitude on behavior; The relationship between.fear of rape and fear of crime among the male subjects may have been caused by such increased accessibility. 42 between domestic victimization and fear of crime as a whole. Women who had been severely abused by a male partner reported significantly more vulnerability to crime than other women. However, less severe domestic victimization was not correlated with increased perceptions of vulnerability. Junger (1987) measured sexual harassment, defined as "violation of the physical integrity and/or autonomy which is related to the primary and secondary sexual characteristics of persons" (p. 364) among 279 women recruited from a list of those who had participated in a study of violence several years earlier. To discern the effects that relationship to the offender had on the victim’s response, data on harassment both within and outside the family were collected. The association between prior victimization and perceived vulnerability was positive and significant (r = .50) for harassment within the family but somewhat lower for outside the family (r = .34). Intrafamilial victimization was also significantly correlated with such behavioral responses as taking precautions and avoidance. In contrast, extrafamilial harassment did not predict perceptions of vulnerability as well. The authors posited that it is more difficult to "neutralize" victimization occurring with someone who was previously seen as safe, thus heightening both feelings of vulnerability and avoidance. 43 In summary, incongruities between risk and perceived vulnerability to crime appear to be at least partially explained by two factors. First, women are typically more physically vulnerable than men. Second, women are targeted for certain kinds of crime (e.g., rape, wife abuse, sexual harassment); and these types of victimization appear to create a climate that fosters feelings of vulnerability to crime. Typifying crimes directed specifically at women is the fact that they often involve trusted others. Thus, the crimes targeted at women occur in situations in which they expect to feel safe. As Perloff (1983) explicated, victims who feel particularly invulnerable before a crime are most likely to experience post-victimization emotional distress. It is hypothesized that coping is more difficult because the victim’s view of the world is severely shaken (Junger, 1987; Perloff, 1983). Thus, crime at the hands of an intimate partner or family member appears to be particularly likely to lead to decreased coping and elevated levels of vulnerability. It follows, therefore, that high levels of vulnerability may well indicate concerns about violence of this type. The present study will examine perceptions of 'vulnerability to crime and the relationship of these perceptions to both sex-role ideology and attitudes toward wife abuse. Gender differences will be examined; and it is 44 hypothesized that women will report more vulnerability to crime as well as more egalitarian sex-role beliefs. Consistent with Defensive Attribution Theory which proposes that feelings of threat affect perceptions of significant events, perceived vulnerability is expected to be significantly related to wife abuse attitudes. Also expected is a significant interaction between vulnerability and sex-role ideology as predictors of wife abuse attitudes, such that wife abuse attitudes will be most consistent with sex-role beliefs when vulnerability is in its lowest range. HYPOTHESES Hypothesis 1. Women will report a significantly stronger sense of vulnerability to overall crime than their male counterparts. Hypothesis 2. There will be a curvilinear relationship between perceived vulnerability and both amount of perpetrator responsibility and justification of wife-beating. Those perceiving moderate vulnerability will attribute the least responsibility to the perpetrator and will be more likely than others to view wife-beating as justified. Those perceiving the greatest levels of vulnerability will attribute the most responsibility to the perpetrator, instead of the victim, and engage in the lowest amounts of justification of wife- beating. Subjects perceiving themselves as least vulnerable to crime will be moderate on the continuum for both perpetrator blame 45 46 and justification. Hypothesis 3. Sex-Role attitudes (as measured by the Sex- role Ideology Scale) and vulnerability will interact as predictors of attitudes about wife abuse as measured by the five scales of the Inventory of Beliefs About Wife Beating (IBWB). In the case of LOW VULNERABILITY only, progressive sex-role beliefs will have a linear negative relationship (correlation) with the justification of abuse and also with beliefs that a wife gains from abuse. In addition, progressive sex-role beliefs will have a positive linear relationship with perpetrator responsibility, approval of perpetrator punishment, and endorsement of aid for victims. Among more threatened persons, however, attitudes toward sex-roles and attitudes toward wife abuse will no longer correlate significantly. 47 Hypothesis 4 Feminist sex-role beliefs will correlate positively and significantly with perceived vulnerability. METHOD Participants Undergraduates (173 female, 98 male) from the Psychology Department’s subject pool at a large midwestern university volunteered to participate in this study in exchange for credit toward one requirement in an introductory psychology course. Their mean age was 19. Measures The present study examines the impact of two variables, perceived vulnerability to crime and attitudes toward sex- roles, on attitudes toward the physical abuse of women. While there are many well-validated inventories of attitudes toward sex-roles, there seems a dearth of psychometrically sound measures of both perceived vulnerability to crime and attitudes toward woman abuse. Inventory of Beliefs about Wife Beating A large portion of the studies conducted on attitudes toward victims have used written vignettes or newspaper/police reports, as is the case in the rape literature (Ward, 1988). Participants have typically been asked questions regarding attributions of responsibility. Other studies employing questionnaires have simply focused on the "normality or approval" of wife-beating (Saunders, Lynch, Grayson, & Linz, 1987), finding that most respondents 48 49 disapprove. As Saunders et al. noted, several important aspects of such observers’ attitudes were unaddressed and the measures used to assess attitudes toward wife abuse were not well-validated. Saunders et al. (1987) have developed a new measure designed to assess diverse aspects of attitudes toward wife abuse, The Inventory of Beliefs about Wife Beating (IBWB). The authors began with a pool of 119 items drawn from such varied sources as rape-attitude scales, wife abuse literature, clinical experience, and consultation with advocates for battered women. A pilot study eliminated approximately two-thirds of these items due to ambiguity or low correlations with other items, resulting in a 41-item measure. A subsequent factor analysis identified five reliable subscales entitled: Wife Beating Is Justified (alpha coefficient = .86), Wives Gain from Beatings (alpha = .77), Help Should Be Given (alpha = .67), Offender Should Be Punished (alpha = .61), and Offender Is Responsible (alpha = .62). They found that all IBWB subscales were significantly correlated at the .001 level (r’s ranging from .20 to .62) with Burt’s (1980) Rape Myth Acceptance Scale for a sample of 675 undergraduate students. They also administered the IBWB with a variety of other measures to more diverse samples: 145 nurses and 86 physicians at a teaching hospital/clinic, 21 male batterers receiving treatment, 70 50 female advocates for battered women, 97 undergraduates at a New England university, and 94 subjects who were part of a study investigating ethnic differences in wife abuse. Saunders et al. (1987) also found that, among college students, all IBWB subscales were significantly correlated (r’s from .18 to -.45) with the short form of the Attitudes Toward Women Scale (AWS; Spence & Helmreich, 1973). Except for the Offender Is Responsible subscale, all IBWB subscales were also significantly correlated with the Sex-Role Stereotyping (SRS) measure (range of r’s from -.15 to .32) and the Hostility Toward Women Scale (HTW) (r’s from .14 to .34), Check & Malamuth, 1983). The IBWB subscales of Wife Beating is Justified, Wife Gains From Beatings, and Help Should Be Given had the strongest correlations with the AWS, the SRS, and the HTW. Thus, college students with more progressive ideologies about women were less likely to view beatings as justified or to think that women benefitted from abuse, and were more likely to believe that victims should be helped. Suggesting that advocates for battered women tend to base their attitudes about abuse on variables other than their attitudes about women, Saunders et al. (1987) found that advocates’ scores on the AWS, SRS, and HTW were not significantly correlated with IBWB responses (except on the Wife Beating is Justified subscale). Treated abusers’ IBWB scores were similarly unpredictable from their attitudes 51 about women. One exception was the significant correlation between AWS scores and the IBWB’s Offender is Responsible subscale, a finding that may have been a product of this sample’s treatment. Virtually all listed correlations of IBWB measures with the AWS, SRS, and HTW were statistically significant, although not all subscale scores were available for the other populations studied (Saunders et al., 1987). The ethnic research sample was the only exception, as their IBWB Offender is Responsible scores did not correlate significantly with AWS responses. All other IBWB subscales yielded significant gender differences. Women were less likely than men to report believing that wife-beating is justified and that women gain from beatings, but more likely than men to state that offenders should be punished and that victims should be helped. Although the Offender Should Be Punished and Wives Gains From Beatings subscales were correlated significantly with the Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1964), Saunders et al. reported that this shared variance was modest and ranged only from 5.8% to 9%. Also suggesting limited response bias was a further finding that no IBWB subscale correlated significantly with the Lie Scale of the Eysenck Personality Inventory (Eysenck & Eysenck, 1969). Comparisons between the IBWB responses of 70 advocates for battered women and 71 batterers yielded the anticipated significant differences on 52 each subscale. Thus, the IBWB appears a promising measure of attitudes about wife abuse. Its construct validity was supported by significant correlations of IBWB scores with gender, sex- roles, and attitudes toward rape. In addition, known-groups validity was demonstrated and there were only mild indications of a social desirability response bias. However, some caution is warranted. The IBWB’s Offender Is Responsible subscale did not relate as expected to several other measures in initial validation studies. The IBWB’s authors also noted that its subscales of Help Should Be Given, Offender Is Responsible, and Offender Should Be Punished had only marginal internal consistency. _§ar of Crime Scale The extant literature has commonly used two items from the National Crime Survey: "How safe would you feel being out alone in your neighborhood during the day?" and the same question for ". . . at night" to assess vulnerability. Because these highly ambiguous questions allow for no specificity with regard to type of crime this methodology is problematic and only using two items makes for questionable reliability. Data on vulnerability to crime in the present study includes a measure of fear/vulnerability as well as an index of prior victimization. Perceived vulnerability was 53 measured using L. N. Smith and Hill’s (1991) Fear of Crime index. This eight-item scale includes questions about concerns for personal and property safety in addition to questions about safety behaviors to which subjects respond in one of four ways: strongly disagree, disagree, agree, or strongly agree. Item-to-total correlations ranged from .43 to .67. Cronbach’s alpha was reported at .82. Factor analysis revealed only one reliable factor that also had face validity. Data on prior victimization were collected via L .N. Smith and Hill’s (1991) Victimization Experience (VE) scale. Respondents were requested to indicate the number of times each of 16 types of victimization was experienced by anyone in their household in the past year, allowing for specificity of crime. In addition, this VE scale includes a weighting system, in which crimes with higher perceived seriousness contribute proportionally to the overall victimization score. Weightings for the present study were based on the mean scores of a separate sample of 130 undergraduate students who rated the seriousness of each offense on a 1 to 10 scale, with 10 indicating extreme seriousness (Smith & Hill, 1991). Each item’s weighting was determined by obtaining the difference between the target item’s mean and the lowest mean and then adding one. The highest weight was assigned to murder (4.4), the lowest to stealing a car or truck part 54 (1.0). The Fear of Crime index was significantly and positively correlated with the Victimization EXperience scale (r = .10, p s .05) and with age (r = .05, p g .05), but inversely and significantly associated with both education (r = -.22, p g .05) and income (r = .11, p g .05, Smith & Hill, 1991). Thus, as in prior studies, more fear was reported by older participants (Box, Hale, & Andrews, 1988) and those with less education and income (Covington & Taylor, 1991). Also consistent with prior findings (Gomme, 1988; G. Smith & Lab, 1991), L.N. Smith and Hill (1991) found that age and income correlated negatively correlated with victimization experiences. Sex-Role Ideology Scale Attitudes about sex—roles were measured using Kalin and Tilby’s (1978) Sex-Role Ideology Scale (SRIS). This measure’s original set of 82 items was based on questions from Kirkpatrick’s (1936) earlier inventory on feminism and also included items addressing many recent societal concerns (e.g., abortion). Known-groups validity was measured via Kalin and Tilby’s (1978) administration to 27 members of women’s liberation organization, women from church or gardening groups (believed to be more traditional), and 324 college students from psychology classes. They omitted items that did not discriminate between 55 groups or which did not have high item-total correlations, resulting in a 30-item measure. Questions were phrased equally in a traditional or nontraditional manner to which subjects were asked to indicate agreement or disagreement on a seven point scale. Kalin and Tilby then administered these 30 items to three groups similar in composition to earlier samples: 29 women from a profeminist organization, 60 women recruited from church and other traditionally-oriented women’s groups, and 162 male and female undergraduates. Each item significantly discriminated between the traditional and nontraditional groups with the mean item scores differing by approximately two standard deviations. Gender differences were in the expected direction, with women scoring as significantly more feminist than men. Three-week test- retest reliability among their student sample was .87. Construct validity was further supported by Buhrke’s (1988) study of 180 undergraduates in which the SRIS was shown to correlate significantly with The Attitude Toward Women Scale (Spence & Helmreich, 1972), the Attitude Toward Masculinity Transcendence Scale (Moreland & Van Tuinen, 1978), and the Sexist Attitudes Toward Women Scale (Benson & Vincent, 1980). Burhke’s (1988) factor analysis of all items from these inventories identified a single reliable factor, indicating that sex-role ideology is a unidimensional variable appropriately measured by a single- 56 scale index such as the SRIS. Similar studies on sex-role ideology (Larsen & Long, 1988; Rombough & Ventimiglia, 1981) have also identified only one reliable factor that accounted for a substantial amount of the variance. In summary, the SRIS appears a valid and reliable measure of attitudes about the appropriateness of varying roles for men and women. Although there are several statistically adequate measures of sex-role ideology, the SRIS was selected chosen for the present purposes for two reasons. Unlike some other inventories (e.g., The Attitudes Toward Women Scale), the SRIS includes items particular to women (e.g., changing name at marriage, abortion, etc.). It also appears less time-dated than earlier measures such as The Attitudes Toward Women Scale, as the latter included such items as "Women should not be bosses in important jobs in business and industry" and "Sons in a family should be given more encouragement to go to college than daughters". Procedures All participants were recruited via a standard Departmental sign-up form soliciting participation in a study entitled "Social Perception and Aggression". After signing an informed consent form, subjects completed all measures in a group setting in the author’s presence. Measures were administered as a set in the following order: the Fear of Crime scale (FOC), the Victimization Experience 57 scale (VE), the Inventory of Beliefs About Wife Beating (IBWB), and the Sex-Role Ideology Scale (SRIS). Forty-five minutes were allotted for the completion of all instruments. At the close of the session, participants were given a "debriefing" form, informing them of this study’s general purpose. Results Data were screened for omissions and irrational (e.g., out of range) responses prior to inclusion in analyses. In total, 11 data sets were excluded, leaving 271 complete sets. Mean scores for all IBWB scales were within one standard deviation of those reported by Saunders et al. (1987) among their sample of 675 undergraduates. Similarly, Fear of Crime mean scores were consistent with those of 3,109 North Carolina residents (L. N. Smith & Hill, 1991) after adjusting for this study’s larger range of seven versus four options on the Likert-formatted measure (i.e., multiplying the present sample’s means by 4/7). Mean scores on the SRIS were also within one standard deviation of Kalin and Tilby’s (1978) sample of 162 male and female college students. Thus, the present sample’s scores appear reasonably representative of other United States college samples. Scale reliabilities are presented separately for men and women in Appendix D. Means and standard deviations for all measures are presented separately for men and women in Table 1. Approval of marital violence in the present sample appeared generally low. Mean item scores indicated that, when responding on a 7-point Likert scale ranging between "strongly agree" and "strongly disagree", women tended to "strongly agree" that women do not gain from abuse and to "agree" that wife abuse is not justified and that help 58 59 should be provided for victims. Women tended to "slightly agree" that offenders are responsible and should be punished. Men tended to "agree" that help should be provided for victims; "slightly agree" that wife abuse is not justified, women do not gain from abuse, and offenders should be punished; and "neither agree nor disagree" that offenders are responsible. Confirming prior gender differences (Saunders et al., 1987), t-tests presented in Table 1 revealed that women’s attitudes differed significantly (p g .01) from men’s in the Table 1. Means, Standard Deviations and t-tests of Differences Between Men and Women on Fear and IBWB Scales. g .511 _1__3 WOMEN Variables M SD M SD t-value IBWB JUSTIFY 53.6 27.8 32.0 12.1 -7.30** IBWB GAINS 16.8 10.0 9.8 5.1 -6.46** IBWB PUNISH 23.1 5.7 27.6 5.2 6.57** IBWB RESPONSIBLE 24.5 6.0 27.7 5.4 4.34** IBWB HELP 28.0 6.6 32.3 3.5 6.09** FEAR OF CRIME 27.3 6.7 30.8 8.0 3.86** PROPERTY CRIME 2.8 3.2 2.4 3.2 -.88 PERSONAL CRIME 1.4 2.9 1.4 3.5 .17 TOTAL VICTIMIZATION 4.1 5.3 3.8 5.6 -.42 SEX-ROLE IDEOLOGY 126.5 19.7 145.3 18.0 7.81** AGE 19.7 1.3 19.3 1.3 ** Q g .01, two-tailed test 60 expected directions for all IBWB scales. Thus, these women were generally much more disapproving of wife abuse than were men. Women’s IBWB scores were also less variable than men’s, as their responses tended to cluster more closely around total disapproval of marital violence while men’s spanned the entire range of responses. This was most pronounced on the IBWB Justification scale on which 13% of women versus 4% of men endorsed complete lack of justification for wife abuse. In addition, 86% of women reported agreeing that wife abuse is not justified (IBWB Justify scores of 42 or less) versus 44% of men. Clearly supporting the hypothesis that women would report more vulnerability to crime than would men, their mean Fear of Crime score exceeded men’s (t = 3.86, p g .001) and women’s scores were also somewhat more variable (SD of 8.0 vs. 6.7). Men’s scores were significantly more variable than women’s on the IBWB scales of: Wife Beating is Justified, Wives Gains from Beatings, and Help Should be Given. Multiple regression analyses were performed on Fear of Crime and IBWB Justification scores, and separately for Fear of Crime with the IBWB Responsibility scale. To test the hypothesized curvilinear model, Fear of Crime scores were squared and added to each regression equation. Thus, each regression equation included two distinct variables: Fear of Crime (to assess the linear relationship between Fear and 61 the variables Justification and Responsibility) and Fear of Crime2 (to assess the quadratic/curvilinear relationship). Tests of significance were performed on the change in the squared multiple correlation coefficient (R2 change) for each of these variables as they were added to the regression equation predicting Justification and Responsibility, so that the variance accounted for by each could be ascertained. These findings are presented in Table 2. As predicted, Fear of Crime associated significantly and negatively with Responsibility scores for both men (r = -.34, p < .01) and women (r = -.18, p < .05), accounting for approximately 11% and 3%, respectively, of the latter measure’s variance. Unexpectedly, Fear correlated significantly with Justification only among men (r = .26, p <.01; women’s r = .12). Thus, there was a clear linear component to the relationship between Fear and Responsibility for all participants but only for males between Fear and Justification. This significant linear relationship did not necessarily preclude a significant quadratic element. Adding the quadratic component (Fearz) to the regression equations after the linear Fear variable provided an index of the quadratic relationship from which the linear effect of Fear was removed. As is shown in Table 2, this curvilinear component of the Fear and Justification relationship was trivial; moreover the curvilinear component 62 of the Fear and Responsibility relationship was significant only for males. For them, the linear aspect of Fear accounted for 11% of the variance in Responsibility scores while the quadratic aspect of Fear accounted for an Table 2. Multiple Regression Analyses of Fear of Crime with IBWB Justification and Responsibility Scales. Spep Dependent Var. Indep. Var. F8 Change Mult. R 98 Men 1. Justification Fear .068** .261 2. Justification Fear2 .004 .268 1. Responsibility Fear .114** .338 2. Responsibility Fearz .035* .386 173 Women 1. Justification Fear .015 .124 2. Justification Fear2 .016 .178 1. Responsibility Fear .033* .182 2. Responsibility Fear2 .000 .183 *p g .05. **p g .01. additional 3%. This significant enhancement identifies a curve in the relationship between males’ Responsibility and Fear scores. Examination of the variables Fear and Responsibility plotted against each other revealed that Responsibility scores decreased as Fear of Crime increased, Thareling off when the Fear score approached its mean value (27.3). 63 To test the hypothesized interaction between Sex-Role and Fear of Crime as predictors of IBWB scales, their composite (Sex-Role x Fear) was computed and then included in a multiple regression equation following its two components. As is shown in Table 3, Sex-Role significantly predicted all IBWB scales for both men and women. Weakly supporting the hypothesis, there was a significant interaction between Fear of Crime and Sex-Role only for men’s Justification scores. No other regression analyses indicated that the relationship between Sex-Role and IBWB scales was contingent upon Fear of Crime scores. Contrary to expectations, it was in cases of high Fear of Crime that men’s sex-role attitudes were more highly related to beliefs about the justification of wife abuse. To scrutinize this trend, post—hoc analyses examined the correlations between sex-role attitudes and wife abuse attitudes for subgroups high and low fear in Fear of Crime, defined as above and below the latter scale’s midpoint. As is shown in Table 4, the relationship between Fear of Crime and Sex-Role Ideology tended to be stronger among the more fearful women and among the less fearful men. That is, with the exception of IBWB Justification, beliefs about wife abuse attitudes were more consistent with Sex-Role Ideology for women who reported more Fear of Crime and for men who reported less. However, none of these differences 64 Table 3. Multiple Regression Analyses of Fear of Crime and Sex-Role with IBWB Scales. Step Dependent var. Indep. Var. I? Change Mult. R 98 Men 1. Justification Fear of Crime .068** .261 2. Justification Sex-Role .292** .600 3. Justification Fear x Sex-Role .026* .621 1. Responsibility Fear of Crime .114** .338 2. Responsibility Sex—Role .181** .543 3. Responsibility Fear x Sex-Role .015 .556 1. Gains Fear of Crime .078** .279 2. Gains Sex-Role .154** .482 3. Gains Fear x Sex-Role .009 .491 1. Punish Fear of Crime .078** .279 2. Punish Sex-Role .170** .498 3. Punish Fear x Sex-Role .025 .522 1. Help Fear of Crime .032 .178 2. Help Sex-Role .374** .637 3. Help Fear x Sex-Role .002 .638 173 Women 1. Justification Fear of Crime .015 .124 2. Justification Sex-Role .152** .409 3. Justification Fear x Sex-Role .012 .424 1. Responsibility Fear of Crime .033* .182 2. Responsibility Sex-Role .046** .280 3. Responsibility Fear x Sex-Role .007 .292 1. Gains Fear of Crime .002 .048 2. Gains Sex-Role .091** .305 3. Gains Fear x Sex-Role .007 .316 1. Punish Fear of Crime .004 .065 2. Punish Sex-Role .049** .231 3. Punish Fear x Sex-Role .000 .231 1. Help Fear of Crime .000 .002 2. Help Sex-Role .090** .301 3. Help Fear x Sex-Role .002 .304 *9 S ’05. **E S 001. 65 between correlations was significant using Fisher’s r to z transformation. Also surprising was the finding that Fear of Crime was a stronger predictor of men’s than of women’s IBWB scores. Fear of Crime was significantly associated with four of the five IBWB scales for men but only one of these five for women. Table 4. Correlations of IBWB Attitudes with SRIS Attitudes in High and Low Fear of Crime‘. Justif. Resp. Gains Punish Help Women Low Fear(n = 83) -.36** .10 -.19 .15 .22* High Fear(n = 90) -.41** .32** -.38** .29** .36** Men LOW Fear(n = 52) -.54** .54** -.49** .56** .68** High Fear(n = 46) -.62** .28 -.37** .24 .57** + Fear scores were divided at the midpoint; men’s scores above 27 and women’s scores above 31 were categorized as high fear. *9 S .05. **p g .01. Correlations of all measures are presented separately for men and women in Table 5. Correlations among the five IBWB scales as well as the inter-correlations of IBWB, Fear of Crime, and SRIS for men were consistently higher than for women. In 9 of the possible 36 comparisons, men’s correlations were significantly higher. For men there appeared to be a general acceptance or rejection of wife abuse, with all attitudes about wife abuse tending to correspond more strongly. In contrast, women’s wife abuse 66 attitudes were appreciably less associated. The significantly higher variance in men’s scores on the Wife Beating is Justified, Help Should be Given, and Wives Gain From Beatings scales may partially explain higher correlations between these scores and other measures. Analyses of all regression equations revealed that the residuals were not appreciably correlated with the independent variables. To investigate the possibility of multicollinearity between independent variables in the regression analyses, correlations between all independent variables were calculated. No such correlation exceeded .34, precluding significant multicollinearity. Clearly refuting the hypothesis that sex-role attitudes would correlate positively with perceived Fear of Crime, sex-role ideology was significantly negatively correlated only with men’s (p = -.30, p g .01), but not with women’s (3 = -.09) Fear of Crime scores. Although not part of the proposed hypotheses, ‘Victimization rates were also explored as such experiences are likely to influence attitudes regarding crime. Caution must be exercised in interpreting any differences in Victimization rates, as this sample is comprised solely of Cellege students rather than various age groups. In Contrast to the 23% household victimization rate reported by the U.S. Bureau of Justice for 1992 (Rand 1993) and the 38% victindzation rate reported by Smith and Hill (1991) for 67 their sample of 3,049 North Carolina residents, 48% of the present sample’s households reported experiencing a crime. The rates for individual crimes were: larceny 65%, robbery 5%, assault 16%, rape 7% and murder 2%. These figures are consistently three to four times higher than those reported by the U.S. Bureau of Justice in their annual National Crime Victimization Survey report (Rand, 1993), an index of both reported and unreported crime. 68 Table 5. Correlations of IBWB, SRIS, Fear of Crime, Property Crime, and Personal Crime Scales. Justify Gains Punish Respons. Help SRIS Fear Prop.Cr. Per. Cr. Justify .750 -._62£ 5 -.6_0 ;.7_9§ -.590 .26b .20a .16 Gains .620 -.400 -.410 -.510 -.460 Jib .15 .19 Punish -.22b -.21b .810 Q 3L0 -.28b -. l6 -. 12 Resp. -.340 -.350 .750 580 file -.340 -.18 -.08 Help -.440 -.410 .380 .360 1641: -.18 -.22a -.22a SRIS -.400 -.300 .23b .23b .300 -.30b -.16 -.16 Fear .12 .05 -.07 -. 18a .00 -.09 .213 .03 Prop.Cr. .17a .06 -.07 -.13 -.21b .04 .16a .520 Per.Cr. .240 .18a -.02 -.12 -.13 -.03 .07 .430 Prop.Cr. = Property Crime Per.Cr. = Personal Crime a=p§.05 b=p_5.01 0=p_§.001 Note: Correlations for 98 men given above the diagonal; correlations for 173 women given below it. Note: Underlined correlations significantly exceeded their cross-gender comparison. Discussion Prior research has demonstrated that age and education may significantly affect the perceptions of both victims (Hilton, 1989) and perpetrators of crime (Aubrey & Ewing, 1989; Gentemann, 1984; Edwards, 1987). Thus, caution must be exercised in interpreting data derived from this wholly college-student sample. The additional fact that this sample was youthful and largely unmarried likely influenced their attitudes regarding violence in marriage. Shaver (1975) hypothesized that a minimal level of relevance is necessary for defensive attribution to occur. Thus, this study’s general finding of minimal defensive attribution may be associated with the uncertain relevance of wife abuse to this sample of late-adolescent college students. Fear of Crime Consistent with hypothesis 1 and prior research (e.g., Gomme, 1988; Gordon et al., 1980; Junger, 1987; M.D. Smith, 1988), these college student women reported significantly more Fear of Crime than did the men. Also corroborating prior research, women’s elevated Fear of Crime scores were not based on higher rates of prior victimization for either personal or property crimes. The slightly older male sample (.4 years) reported a slightly higher amount of property victimization (28 versus .24), although this difference was statistically insignificant. Thus, the notion that women 69 70 have a higher sensitivity to risk of crime received support. This age group is at a higher risk for crime than the national average (Bastian, 1993). In addition, their households are likely larger in size as many college students have several roommates, suitemates, and housemates. The amount of household victimization reported by the present sample did significantly exceed the findings reported by the U.S. Bureau of Justice in their annual National Crime Victimization Survey report (Rand, 1993). Possible methodological variation can be largely ruled out as a factor in this elevation, as Smith and Hill’s 1991 study employed the same victimization measure and found overall victimization rates that were also slightly but significantly lower (38% vs. 44%) than the present sample’s. The youth of the participants and the likelihood of larger household size are probable contributors to the higher household victimization rates. Fear of Crime and Wife Abuse Attitudes Consistent with Defensive Attribution Theory, hypothesis 2 predicted that Fear of Crime would be a significant factor in individual’s attributions regarding wife abuse. Briefly stated, this theory posits that, in response to feelings of threat, we cognitively organize our perceptions of serious events to reduce feelings of vulnerability. Fear of Crime was significantly correlated 71 with attributions of responsibility for wife abuse among both men (p = —.34) and women (; = -.18). However, attitudes about justification of wife abuse related significantly only to Fear of Crime among men. It is puzzling that perceived vulnerability to crime played a more pivotal role in men’s attitudes, since victimization via wife abuse is clearly more relevant to women. Women were markedly less likely than men to view wife abuse as justified (p = -7.30, p 5 .01). In addition, women’s Justification scores tended to cluster near a total lack of justification and were markedly less variable than men’s (women’s variance = 12.1; men’s variance = 27.8). Women’s slightly lower correlation than men’s (p’s of .12 vs .26) between Fear of Crime and Justification was likely related to their narrower set of Justification scores. To test the Defensive Attribution Theory hypothesis regarding the relationship between perceived vulnerability to crime and wife abuse attitudes, regression analyses were required. Partially supporting the hypothesis of a curvilinear relationship between perceived vulnerability to crime and wife abuse attitudes, men’s attitudes regarding responsibility for wife abuse related curvilinearly to Fear of Crime scores. As Fear of Crime increased from its low to midpoint, men attributed less responsibility to perpetrators of wife abuse. Thus, as predicted, moderate levels of perceived vulnerability to crime among males were associated 72 with decreased attributions of responsibility to male wife abusers. Perceiving a perpetrator to be responsible for the violence would not seem to allow a moderately vulnerable observer to feel safe, as such persons, by definition, believe that they are also vulnerable to crime. Thus, they would tend to attribute responsibility to other sources so that they may feel safe that there are causes of victimization other than brutal criminals. One may presume that some of the responsibility for her abuse was assigned to the wife, as she appeared the most likely target for such attributions. This would allow observers to feel more secure, believing that there was something about the victim or her situation, rather than about the perpetrators to whom they feel vulnerable, that was responsible for the victimization. Surprisingly, it was this sample’s men who were lowest in Fear of Crime that attributed the most responsibility to wife abuse perpetrators. It had been hypothesized that those perceiving themselves as less vulnerable to crime would make more moderate attributions because crime does not appear sufficiently relevant to them to warrant a strong attributional stance. It had also been anticipated that the most vulnerable participants would hold wife abusers more liable because they would identify most with victims and want to avoid being seen as responsible should they be 73 victimized. A clear complicating factor is that men are overwhelmingly the perpetrators, rather than the victims, of marital violence. Thus, heightened vulnerability to crime might not necessarily predict men’s attributions about marital assault in the same way it might predict attitudes about other crimes for which men have more pertinent fears of victimization. It is quite interesting to note, however, that the Defensive Attribution Theory hypothesis was supported, albeit marginally, only for men. Women’s attributions regarding the justification of, and responsibility for, wife abuse did not demonstrate the curvilinear relationship to Fear of Crime predicted by Defensive Attribution Theory. In fact, Fear of Crime was not significantly associated with women’s beliefs regarding the justification of wife abuse. Similarly, Fear of Crime correlated even more faintly with women’s beliefs regarding whether a woman gains from abuse, whether offenders should be punished, or whether help should be given to victims. Among the men, however, Fear of Crime correlated significantly with all attitudes toward wife abuse, except whether victims should be helped. Prior research has also demonstrated that men have stronger tendencies toward defensive attribution than women (e.g., Schiavo, 1973; Shaw & Skolnick, 1971). Shaw and Skolnick (1979) similarly posited that differential gender 74 relevance was likely responsible for women’s decreased defensive attributional reactions. When their study was replicated with an adjustment for relevance (Schiavo, 1971), however, the gender difference persisted. Thus, there are indications of important gender differences in defensive attributions. The present male sample reported less Fear of Crime than these college women. However these men’s Fear levels were more predictive of their wife abuse attitudes, suggesting that Fear of Crime may have a pivotal role in their perspective about crime of this type. The minimal correspondence between women’s Fear of Crime and their wife abuse attitudes was unexpected. Violence in the home is certainly a category of crime for which women are particularly at high risk. Thus, it was expected that Fear of Crime in general would predict attitudes about the specific crime of wife abuse. One possible explanation is that the general Fear of Crime measure employed was not sufficiently suited to tap in to perceptions of vulnerability about wife abuse. Another hypothesis is that these women failed to recognize their vulnerability to this type of crime. This lack of awareness may be at least partially responsible for the lack of association between women’s perceived vulnerability to crime and their attitudes about wife abuse. Generally, women’s wife abuse attitudes were also less associated with each other as well as less 75 consistent with their sex-role ideology than were men’s. The research of Kristiansen and & Giulietti (1990) similarly demonstrated a less direct link for women than for men between wife abuse attitudes and their sex-role ideology and Just-World beliefs. It was hypothesized in the present study that the variable of fear was the mediating factor. However, this was not demonstrated. In summary, the Defensive Attribution hypothesis was partially supported in that Fear of Crime was clearly appreciably associated with men’s attitudes toward wife abuse. However, perceived vulnerability did not demonstrate the expected associations with women’s wife abuse attitudes. Methodological issues may have been a factor. The Fear of Crime scale does not include items that measure fear of specific types of crime. Thus, participants may have been using different kinds of crime as their reference point when responding to the scale’s eight items, and this may have been partially gender-determined. This susceptibility to individual interpretation may have served to attenuate the findings. Consistent with Balance Theory, women were much more disapproving of wife abuse than men, suggesting that gender differences in identification with female victims is a relevant factor in the formation of related attitudes about wife abuse. Indeed, women were significantly more egalitarian in their sex-role ideology than were men. Thus, 76 women reported attitudes more favorable toward women and also reported disapproval of the abuse of women, supporting Heider’s (1958) hypothesis that people tend toward a balanced state in their perceptions of others. Sex-Role Attitudes. Fear of Crime. and Wife Abuse Attitudes Contrary to expectations, Fear of Crime was not found pivotal in the relationship between sex-role attitudes and wife abuse attitudes. With one exception, sex-role attitudes consistently displayed the expected associations with wife abuse attitudes regardless of Fear of Crime level. Hypothesis 3 predicted that, when observers feel vulnerable to crime in general, their wife abuse attitudes would not necessarily be consistent with their sex-role attitudes, as issues of safety would supersede cognitive concordance in attitudes. However, sex-role attitudes related fairly strongly to all facets of attitudes toward wife abuse, especially among men, such that Fear of Crime did not appreciably affect this relationship. Thus, these findings appear more consistent with the Balance Theory tenet that our "sentiments" about victims affect our perceptions of them and events involving them. The one exception was men’s Justification scores where, contrary to expectations, it was for men with higher Fear of Crime that sex-role attitudes were more firmly related to beliefs about the justification of wife abuse. 77 Although the planned regression analyses did not reveal significant effects of Fear of Crime on the relationship of sex-role attitudes to wife abuse attitudes, correlation analyses did suggest a clear, although statistically nonsignificant, pattern. With the exception of Justification, wife abuse attitudes were more consistent with sex-role ideology for more vulnerable women and less vulnerable men. This gender difference involving perceived vulnerability might partially explain why some research (e.g., Finn, 1986; Kristiansen & Giulietti, 1990) has found sex-role to be significantly linked with attitudes about violence against women while others (e.g., Miller et al., 1976) found this to be true for women only. The higher agreement in attitudes among less vulnerable men had been anticipated, as minimal perceived vulnerability to crime was not expected to supersede the tendency toward attitude concordance. In contrast, why women’s sex-role attitudes would be more related to their wife abuse attitudes among only those with greater Fear of Crime remains puzzling. Also interesting is the manner in which Justification scores related differently to Fear of Crime and Sex-Role Ideology (see Table 4) than the other wife abuse attitudes. One hypothesis is that Justification implies social acceptance of the behavior, whereas this is not necessarily the case for the other attributions about wife abuse. For example, one may disapprove of the use of 78 force in marriage but still attribute some responsibility to each party and not necessarily believe that intervention should automatically be provided for victims. Thus, justification denotes a judgement about excusing the behavior in a way that the other attitudes do not. And it may be for this reason that Justification scores related differently to measures of Fear and Sex-Role than did the other IBWB scales. The hypothesis that Fear of Crime would be positively associated with more feminist sex-role beliefs (Hypothesis 4) was unsupported. It was anticipated that, as Fear of Crime increased, individuals would be more attuned to their own vulnerability to inequality and oppression and would, therefore, endorse more feminist sex-role ideology. Contrary to expectations, it was the men with less Fear of Crime who reported holding more feminist views, while women’s sex-role ideology appeared unrelated to their Fear of Crime. Future Research The present findings, as well as some prior evidence, suggest that gender impacts defensive attributional processes. Further research is necessary to substantiate such a hypothesis and examine the nature of the relationship between gender and defensive attribution. Perceived vulnerability, a central factor in defensive attribution, 79 may be differentially acceptable for men and women and may, therefore, be underreported by men. That Fear of Crime significantly predicted men’s wife abuse attitudes more consistently than women’s suggests that fear may have been more salient to the men than self-reports indicated. More objective measures of vulnerability (e.g., inquiring about the use of alarms, extra entry locks, protective mace) would be helpful in discerning the role of fear in attributional processes. Perhaps the lower fear levels reported by men may have been indicative that a defensive process had successfully occurred. Having objective vulnerability data could help to substantiate such a hypothesis, as objective data might provide evidence of underreporting as well as demonstrate the heightened vulnerability levels that are an integral part of the Defensive Attribution Theory. A key factor in the present study appears to have been the larger impact of Fear of Crime on men’s attributions about wife abuse. Whether Fear’s impact on attitudes is gender-dependent for other crimes is also a notion warranting additional research. Woman-battering is obviously a crime in which gender of victim and perpetrator is clearly defined. Thus, investigations of this crime are not as amenable to examinations of gender differences in attributions, as relevance is clearly biased from the outset. Similarly, given that women appear to have a heightened sensitivity to risk of crime, it might be useful 80 to employ target events in which gender differences in relevance are not as pronounced, i.e., a automobile accident. With differential relevance controlled in this manner, information on gender differences in defensive attribution might be more readily ascertained. Another means to address the gender confound in relevance would be research examining physical abuse in same sex relationships. Relevance is necessary for defensive attribution processes to occur. The likely limited relevance of wife abuse for unmarried college students probably influenced the present findings. Research comparing attributions in married versus single participants and for those varying in experience of observing abuse between parents would provide useful information on the importance of relevance in attributional processes. Similarly, it would be valuable to determine whether relevance could be manipulated via pretest communications, for example, informing participants that woman abuse was a significant problem on their college campus. In this manner, woman abuse might be made more relevant to samples comprised of college students. APPENDICES Appendix A Fear of Crime Please respond to the following on a seven point scale 1 2 3 4 5 6 7 very strongly disagree neutral agree strongly very strongly disagree agree strongly disagree agree 1. When I am away from home, I worry about the safety of my property. 2. I worry a great deal about my personal safety from crime and criminals. 3. Even in my own home, I’m not safe from people who want to take what I have. 4. There are some parts of the county that I avoid during the day because of fear of crime. 5. There are some parts of the county that I avoid at night because of fear of crime. 6. I feel safe going anywhere in my community or neighborhood in the daytime. 7. I feel safe going anywhere in my community or neighborhood after dark. 8. Crime is more serious than the newspapers and the TV say 0 Victimization Experience 9. During the past 12 months, did anyone damage, destroy or attempt to destroy your home or any property around your home? a. no b. yes How many times? 10. During the past 12 months, did anyone steal or try to steal a car, truck, motorcycle, or farm machinery owned by you or other members of your household? 81 82 11. During the past 12 months, did anyone steal anything from inside your home, such as a stereo, TV, jewelry, gun, or purse, etc.? 12. During the past 12 months, did anyone steal anything that is kept outside your home such as a bicycle, a garden hose, farm tools, or livestock? 13. During the past 12 months, did anyone steal parts attached to a car, truck, or farm machinery owned by any member of your household, such as a battery, hubcaps, or tapedeck? 14. During the past 12 months, did you or any member of your household have anything stolen from them while they were away from home, for instance, at work, school, in a theater, in a restaurant, or while travelling? 15. During the past 12 months, did you or any member of your household have a purse or wallet snatched or pockets picked? 16. During the past 12 months, did you or any member of your household have something stolen from inside a car or truck, such as packages or clothing? 17. During the past 12 months, did anyone break into or somehow illegally get into your house, apartment, garage, or another building on your property? 18. During the past 12 months, did you find a door jimmied, a lock forced, or any other signs of an attempted break-in (do not include business property, second home, or camps)? During the past 12 months, were you or any member of your household a victim of any of the following violent crimes? 19. Did anyone take something of attempt to take something directly from you or any member of your household by using force, such as a stick-up, mugging, or threat? 20. Did anyone beat-up, attack, or hit you or any member of your household? 21. Were you or any member of your household knifed, shot at, or attacked with some other weapon by anyone? 22. Did anyone threaten to beat-up or threaten you or any member of your household with a knife, gun, or some other weapon? 23. Did anyone rape or attempt to rape you or any member of 83 your household? 24. Were any members of your household murdered? Appendix B Inventory of Beliefs About Wife Beating (IBWB) 25. A husband has no right to beat his wife even if she breaks an agreement she has made with him. 26. Even when a wife’s behavior challenges her husband’s manhood, he’s not justified in beating her. 27. A wife doesn’t deserve a beating even if she keeps reminding her husband of his weak points. 28. Even when women lie to their husbands, they do not deserve to get a beating. 29. A sexually unfaithful wife deserves to be beaten. 30. Sometimes it is OK for a man to beat his wife. 31. It would do some wives good to be beaten by their husbands. 32. Occasional violence by a husband toward his wife can help maintain the marriage. 33. There is no excuse for a man beating his wife. 34. A woman who constantly refuses to have sex with her husband is asking to be beaten. 35. Episodes of a man beating his wife are the wife’s fault. 36. Wives could avoid being battered by their husbands if they knew when to stop talking. 37. Battered wives are responsible for their abuse because they intended it to happen. 38. Wives who are battered are responsible for the abuse because they should have foreseen it would happen. 39. Battered wives try to get their partners to beat them as a way to get attention from them. 40. When a wife is beaten, it is caused by her behavior in the weeks before the battering. 41. Most wives secretly desire to be beaten by their 84 85 husbands. 42. Wives try to get beaten by their husbands to get sympathy from them. 43. Women feel pain and no pleasure when beaten up by their husbands. 44. If I heard a woman being attacked by her husband, it would be best to do nothing. 45. If I heard a woman being attacked by her husband, I would call the police. 46. Wife-beating should be given a high priority as a social problem by government agencies. 47. Social agencies should do more to help battered women. 48. Women should be protected by law if their husbands beat them. 49. If a wife is beaten by her husband, she should divorce him immediately. 50. The best way to deal with wife beating is to arrest the husband. 51. How long should a husband who beats his wife spend in prison or jail 1 2 3 4 5 6 7 No time 1 mo. 6 mo. 1 yr. 3 yrs. 5 yrs. 10 yrs. 52. A wife should move out of the house if her husband beats her. 53. Husband who batter should be responsible for the abuse because they intended to do it. 54. Cases of wife heating are the fault of the husband. 55. Husband who batter should be responsible for the abuse because they should have foreseen it would happen. 56. Both parties are equally responsible for the battering. 57. A husband is justified in hitting his wife is he suspects she is having an affair with another man. 58. It is okay for a man to hit his wife/girlfriend if she 86 threw the first punch. 59. Hitting a female partner is acceptable under limited circumstances. 60. Neglect of household responsibilities is NOT sufficient reason for a man to hit his wife/girlfriend. 61. Even if a man has proof that his wife/girlfriend is having an affair with another man, it is not okay for him to use physical force against her. 62. Hitting a female partner is not acceptable under any circumstances. 63. As long as no injuries result, it is okay for a man to occasionally use physical force on his wife/girlfriend. 64. Making threats to hit a wife/girlfriend without actually doing so is harmless. 65. A husband/boyfriend should use only the amount of force absolutely necessary to keep his wife/girlfriend under control; anything more is excessive. Appendix C Sex-Role Ideology Scale (SRIS) 1. The husband should be regarded as the legal representative of the family in all matters of law. 2. A wife’s activities in the community should complement her husband’s position. 3. A woman should have exactly the same freedom of action as a man. 4. The best thing a mother can teach her daughter is what it means to be a woman. 5. A married woman should feel free to have men as friends. 6. Woman’s work and man’s work should not be fundamentally different in nature. 7. Swearing by a man is no more objectionable than swearing by a woman. 8. A woman is not truly fulfilled until she has been a mother. 9. When a man and a woman live together, she should do the housework and he should do the heavier chores. 10. A normal man should be wary of a woman who takes the initiative in courtship even though he may be very interested. 11. It is an outdated custom for a woman to take her husband’s name when she marries. 12. Women should be paid a salary by the state for the work they perform as mothers and homemakers. 13. Women should be much less concerned about make-up, clothing, and body-care. 14. Every child should be taught from an early age to feel a special honor and respect for motherhood. 15. A woman should be appreciative of the glances she receives as she walks down the street. 16. It should be perfectly all right for a mature woman to 87 88 get involved with a younger man. 17. Marriage should not interfere with a woman’s career any more than it should with a man’s. 18. A man’s responsibility to his children is provide them with the necessities of life and discipline. 19. A woman should be careful how she looks, for it influences what people think of her husband. 20. A woman who dislikes her children is abnormal. 21. Homosexual relationships should be as socially acceptable as heterosexual relationships. 22. More day care centers should be available to free mothers from the constant caring for their children. 23. Women should be allowed the same sexual freedom as men. 24. A man’s job is too important for him to get bogged down with household chores. 25. A woman should be no more concerned with her physical appearance on the job than a man. 26. Abortion should be permitted at the woman’s request. 27. The first duty of a woman with young children is to home and family. 28. For the good of the family, a wife should have sexual relations with her husband whether she wants to or not. 29. A woman should be more concerned with helping her husband’s career than having a career herself. 30. Women should not expect men to offer them seats in buses. Appendix D Chronbach’s Alpha Coefficients for Men and Women for the Inventory of Beliefs About Wife Beating, Fear of Crime, Victimization Experience, and Sex-Role Ideology Scales. 2. _EH 1_3 WOMEN Variables IBWB BEATING JUSTIFIED .94 .81 IBWB WIFE GAINS .90 .81 IBWB PUNISH .82 .70 IBWB PERP IS RESPONSIBLE .78 .72 IBWB HELP WIFE .88 .81 FEAR OF CRIME .59 .71 PROPERTY CRIME .65 .67 PERSONAL CRIME .62 .76 VICTIMIZATION TOTAL .75 .78 SEX-ROLE IDEOLOGY .84 .80 89 BI BLIOGRAPHY LIST OF REFERENCES Andrews, B. & Brewin, C. R. (1990). Attributions of blame for marital violence: A study of antecedents and consequences. Journal of Marriage and the Family, 5;, 757-767. Bastian, L. D. (1993). A national crime victimization survey report: Criminal victimization 1992 (NCJ-l44776). Washington, DC: U.S. Government Printing Office. Benson, P. L. & Vincent, S. (1980). Development and validation of the Sexist Attitudes Toward Women Scale (SATWS). Psychology of Women Quarterly, 5, 276-291. Box, 5., Hale, C. & Andrews, G. (1988). Explaining fear of crime. British Journal of Criminology, gg, 340-355. Branscombe, N. R. & Deaux, K. (1991). Feminist attitude accessibility and behavioral intentions. Psychology of Women Quarterly, lg, 411-418. Buhrke, R. A. (1988). Factor dimensions across different measures of sex role ideology. Sex Roles, 18, 309-321. Burger, J. M. (1981). Motivational biases in the attribution of responsibility for an accident: A meta-analysis of the defensive attribution hypothesis. Psychological Bulletin, g9, 496-512. Burt, M. R. (1980). Cultural myths and support for rape. Journal of Personalit and Social Ps cholo , 38, 217-230. Chaikin, A.L. & Darley, J.M. (1973). Victim or perpetrator?: Defensive attribution of responsibility and the need for order and justice. Journal of Personality and Social Psychology, 25, 268-275. Check, J. V. P. & Malamuth, N, M. (1983). Sex role stereotyping and reactions to depictions of stranger versus acquaintance rape. Journal of Personality and Social Psychology, 35, 344-356. Coates, D., Wortman, C. B. Abbey, A. & Holland, E. (1979). 90 91 Reactions to victims. In I. H. Frieze, D. Bar-Tal, & J. S. Carroll (Eds.), New Approaches to Social Problems. San Francisco: Jossey-Bass Publishers. Cohn, E. S. & Sugarman, D. B. (1982). Marital abuse: Abusing the one you love. Victimology: An International Journal, 5, 203-212. Covington, J. & Taylor, R. B. (1991). Fear of crime in urban residential neighborhoods: Implications of between - and within-neighborhood sources for current models. Sociological Quarterly, gg, 231-249. Crowne, D. P. & Marlowe, D. (1964). The Approval Motive; Studies in Evaluative Dependence. New York: John Wiley. Dickstein, L. J. (1988). Spouse abuse and other domestic violence. Psychiatric Clinics of North America, 1;, 611-628. Edwards, C. (1987). Public opinion on domestic violence a review of the New Jersey survey. Response,1Q, 6-10. Ewing, C. P. & Aubrey, M. (1987). Battered woman and public opinion: Some realities about the myths. Journal of Family Violence, ;, 257-264. Eysenck, H. J. & Eysenck, S. B. (1969). Manual of the Eyesenck Personality Questionnaire. London: Hodder & Stoughton. Finn, J. (1986). The relationship between sex role attitudes and attitudes supporting marital violence. Sex Roles, 13, 235-244. Frieze (1979). Perceptions of battered wives. In I. H. Frieze, D. Bar-Tal, & J. 8. Carroll (Eds.), Neg Approaches to Social Problepe. San Francisco: Jossey- Bass Publishers.r Gentemann, K. M. (1984). Wife beating: Attitudes of a non- clinical population. Victimology: An International Journal, 2, 109-119. Gomme, I. (1988). The role of experience in the production of fear of crime: A test of a causal model. Canadian Journal of Criminology, e9, 67-76. Gordon, M. T., Riger, S., LeBailey, R. K., & Heath, L. (1980). Crime, women and the quality of urban life. Signs, 5, 144- 160. 92 Green, A” (1988). Child maltreatment and its victims: A comparison of physical and sexual abuse. Psychiatric Clinics of North America, ii, 591-610. Guralnik, D. B. (Ed.). (1984). Webster’s New World Dictionary (rev. ed.). New York: Simon & Schuster, Inc. Heider; F. (1958). The Psychology of Inteppersonal Relations. New York: Wiley. Hilberman, E. (1980). Overview: The "wife-beater’s wife" reconsidered. American Journal of Psychiatry, 137, 1336-1347. Hilton, N. Z. (1989). When is an assault not an assault? The Canadian public’s attitudes toward wife abuse and stranger assault. Journal of Familv Violence, 3, 323- 337. Janoff-Bulman, R. (1982). Esteem and control bases of blame: Adaptive strategies for victims versus observers. Journal of Personalit , :9, 180-192. Janoff-Bulman, R. & Wortman, C. B. (1977). Attributions of blame and coping in the "real world": Severe accident victims react to their lot. Journal of Personality and Social Psychology, ee, 351-363. Junger, M. (1987). Women’s experiences of sexual harassment: Some implications for their fear of crime. British Journal of Criminology, g1, 358-383. Kalin, R. & Tilby, P. J. (1978). Development and validation of a sex-role ideology scale. Psychological Reports, 5;, 731-738. Kirkpatrick, C. (1936). The construction of a belief-pattern scale for measuring attitudes toward feminism. Journal of Social Psychology, 1, 421-437. Koski, P. R. & Mangold, W. D. (1988). Gender effects in attitudes about family violence. Journal of Family Violence, 3, 225-237. Kosloff, K. (1984). The battered woman: A developmental perspective. Smith College Studies in Social Work, pg, 181-203. Kristiansen, C. M. & Giulietti, R. (1990). Perceptions of wife abuse: Effects of gender, attitudes toward women, and just- world beliefs among college students. Psychology of Women Quarterly, 15, 177-189. 93 Larsen, K; S. & Long, E. (1988). Attitudes toward sex-roles: 'Traditional or egalitarian? Sex Roles, i2, 1-12. Lavoie, F., Jacob, M., Hardy, J., & Martin, G. (1989). Police attitudes in assigning responsibility for wife abuse. Journal of Familv Violence, A, 369-388. Lerner, M. J. (1965). Evaluation of performance as a function of a performer’s reward and attractiveness. Journal of Personality and Social Psychology, ;, 355- 360. Lerner, M.J. (1971). Observer’s evaluation of a victim: Justice, guilt, and veridical perception. Journal of Personality and Social Psychology, eg, 127-135. Lerner, M. J. & Simmons, C. H. (1966). Observer’s reaction to the "innocent victim": Compassion or rejection. Journal of Personality and Social Psychology, A, 203- 210. Miller, F. D., Smith, E. R., Ferree, M. M., & Taylor, S. E. (1976). Predicting perceptions of victimization. Journal of Applied Social Psychology, p, 352-359. Moreland, J. R. & VanTuinen, M. (1980). The Attitude Toward Masculinity Transcendence Scale. Unpublished manuscript, Southern Illinois University at Carbondale. Moss, A. & Ewing, C. P. (1989). Student and voter subjects: Differences in attitudes toward battered women. Journal of Interpereonal Violence, 3, 289-297. Pagelow, M. (1984). Family Violence. New York: Praeger. Perloff, L. S. (1983). Perceptions of vulnerability to victimization. Journal of Social Issues, 32, 41-61. Rand, M. R. (1993). Crime and the nation’s households, 1992 (NCJ-143288). Washington, DC: U.S. Government Printing Office. Renzetti, C. M. (1987). New wave or second stage? Attitudes of college women toward feminism. Sex Roles, 19, 265- 277. Rombough, S. & Ventimiglia, J. C. (1981). Sexism: A tri- dimensional phenomenon. Sex Roles, 1, 747-755. Saunders, D. G., Lynch, A. B., Grayson, M., & Linz, D. (1987). The inventory of beliefs about wife beating: The construction and initial validation of a measure of 94 beliefs and attitudes. Violence and Victime, 2, 39-57. Schiavo, R. S. (1973). Locus of Control and judgements about another’s accident. Psychological Reports, 22, 483-488. Shavery K5 G. (1970). Defensive attribution: Effects of severity and relevance on the responsibility assigned for an accident. Journal of Personality and Social Psychology, mg, 101-113. Shaver, K. G. (1975). An Introduction to Attribution Process. Cambridge, Mass.: Winthrop. Shaver, K. G. & Drown, D. (1986). On causality, responsibility, and self-blame: A theoretical note. Journal of Personality and Social Psychology, pg, 697-702. Shaw, J. I. & Skolnick, P. (1971). Attribution of responsibility for a happy accident. Journal of Personality and Social Psychology, 22, 380-383. Sloan, R. P. & Gruman, J. C. (1983). Beliefs about cancer, heart disease, and their victims. Psychological Reports, Q2, 415-421. Smith, G. & Lab, S. P. (1991). Urban and rural attitudes toward participating in an auxiliary policing crime prevention program. Criminal Justice and Behavior, me, 202-216. Smith, L. N. & Hill, G. D. (1991). Victimization and fear of crime. Criminal Justice and Behavior, 22, 217-239. Smith, M. D. (1987). Changes in the victimization of women: Is there a "new female victim"? Journal Of Research In Crime And Delinquency, 23, 291-301. Spence, J. T. & Helmreich, R. (1972). The Attitudes Toward Women Scale: An objective instrument to measure attitudes toward the rights and roles of women in contemporary society. JSAS Catalog of Selected Documents in Psychology, 2, 66. (MS. No. 153). Spence, J. T. & Helmreich, R. (1973). A short version of the Attitude Toward Women Scale (AWS). Bulletin of the Psvchonomic Sociepy, 2, 219-220. Stark, R. & McEvoy III, J. (1970). Middle class violence. Psychology Today, A, 52-65. Straus, M. A. (1980). Wife-beating: How common and why? In 95 M. A. Straus & G. T. Hotaling (Eds.), The Social Causes of Husband-Wife Violence. Minneapolis: University of Minnesota Press. Sweeny, P., Anderson, K., & Bailey, S. (1986). Attributional style in depression: A meta-analytic review.iJournal of Personality and Social Psychology, pg, 974-991. Tennen, H., Affleck, G., & Gershman, K. (1986). Self-Blame among parents of infants with perinatal complications: The role of self-protective motives. Journal of Personality and Social Psychology, pp, 690-696. Thornton, B. (1984). Defensive attribution of responsibility: Evidence for an arousal-based motivational bias. Journal of Personality and Social Psychology, A9, 721-734. Tomeh, A. K. (1978). Sex-role orientation: An analysis of structural and attitudinal predictors. Journal of Marriage and The Family, 5Q, 341-354. Walster, E. (1966). Assignment of responsibility for an accident. Journal of Personality and Social Psychology, ;, 73-79. Walther, D. J. (1986). Wife abuse prevention: Effects of information on attitudes of high school boys. Journal of Primary Prevention, 1, 84-90. Ward, C. (1988). The attitudes toward rape victims scale. Psychology of Women Quarterly, 22, 127-146. Warr, M. (1987). Fear of victimization and sensitivity to risk. Journal of Quantitative Criminology, 2, 29-46. Yarmey, A. D. (1985). Older and younger adults’ attributions of respon51b111ty toward rape victims and rapists. Canadian Journal of Behavioral Science, 21, 327-338. o LIBRQRIE 65403 312930104 ICHICQN STQTE UNIV ..(t (5.») 9.: 7.1:: c ... ...... 12 . .... . 1 1.1.5... :2 I. r 5 511252,: _. .Z... ..... .: ......