22:... i . li..k......34h Eli..." L. sins“?! éflmufli n- r. .01. . Japan .. '1 u‘{.nfllv I. :1» , .h ; giaum... ;a.§tnm .. a. .. .4 Fan. nun! . L}... . 1%! iii»! l§.I}cl.r.v. 1.! . 3.33 ’ z -. a y ..t...?2..s Nun t .y. .. 5‘... 6’10. Gui).- ..... ....£.-..2:. .21... . i an... 21;-.. » .wh-Ilnx I. ‘3‘ v. ? . :3. . 84h“. .3 .i. 2 {J :33! 3.... .2... 13:7 $5.531! . 3... .59... , . éurrx} (1.1.5 Radian: fix, 325:5} .21... 4.}. S. . “ pi 2‘ 5.36.2.5}: . 5.2 :2! .54. 3:25;.) a 1...... 4t: :1 1...: ;. Lax... . z... .\.rbi..flfiy3’ . II. IV... .4 OH. (I. 3'!sz I ll Ce... \5 .n . a... $911 1.2.}? . 9 3.9 5531‘ . (1| . q . . 23.3.... 1 5. l . t. 9 .. 13$: .. .5139» a . K24. . 1 2:34. I... )II! :{.|...:. 1...}! t? 3901!! :7} 9.... :alriittfizizt : 111......2311 {LG.}IK 52)}. \t ‘ so: .1 k. 5.? 9.51:1; fihrfl. 1.35.651...» $253.. é. . 3;. by"; F 9‘3? .3!l!o.}..l.. .§£..(...s.lnzs.§.. {2...}...- a- .. defies)... t .521... .3. .2, 33.5.2. .. .33.. 5? 1.992133... 9. 5.1.3:- : . . . . 015 i.(.r.A .....J..!. 1! ...1 .11.. . I. v 1:. :1. E25... . . . 5.3... . . .93.: gas 55 . 3...»... .._......§ . p. c - I THESIS I /’,‘ I a"! I'iiiiiiiiiiiiiiiii Michigan State University This is to certify that the thesis entitled Exploring the Link Between Stereotypic Images and Intimate Partner Violence in the African-American Community presented by Tameka L. Gillum has been accepted towards fulfillment of the requirements for Master of Arts, Psychology degree in Major professor Dam January 241 2000 0-7639 MS U i: an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE UAR (372152 JEE§EBH ,V ‘17 I“ A??? 11- E 2W ha 1 4 U 3 SEP 0 12093. "MAY 1 ozoc 11/00 chiRCIDaIeDmpfiS-p.“ EXPLORING THE LINK BETWEEN STEREOTYPIC IMAGES AND INTIMATE . PARTNER VIOLENCE IN THE AFRICAN~AMERICAN COMMUNITY BY Tameka L. Gillum A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Psychology 2000 l/.|‘ .. . rik- :ilivltIKI' \‘latio: '."l I: I'L.-|nb,,i . t!ulti.'|ll.|l|u|l|lfrt\l|li.ttl . . 'I :1 'blr . .i', ‘.[||l|r)|l| ABSTRACT EXPLORING THE LINK BETWEEN STEREOTYPIC IMAGES AND INTIMATE PARTNER VIOLENCE IN THE AFRICAN-AMERICAN COMMUNITY By Tameka L. Gillum This study was an exploratory investigation of the link between stereotypic images of African-American women and intimate partner violence in the African—American community. Researchers have suggested that there may be a link between African-American men’s perceptions of African- American women as jezebels and matriarchs and intimate partner violence committed against them. A community-based sample of 221 African—American men was used to first examine whether African-American men actually endorsed these stereotypic images of African-American women and second, to explore whether a belief in these images related to a belief that it is justified to use violence against an intimate partner. The results of this study indicate that a large percent of African-American men did endorse the stereotypic images of African-American women as matriarchs and jezebels and that this endorsement did positively relate to justification of violence against women. Copyright by Tameka L. Gillum 2000 ACKNOWLEDGEMENTS There are several people that I wish to thank for their assistance in the completion of this project. First and foremost I would like to thank my Lord and Savior for giving me the strength, discipline, wisdom, and patience to complete this project. I would like to thank my advisor, Dr. Cris Sullivan for her wisdom, guidance, and support throughout this process. I thank her for reading numerous drafts of my work in a timely manner, providing constructive feedback towards the completion of this final product and taking the time to meet with me regularly to discuss this process. I would also like to thank my committee members. Dr. Deborah I. Bybee I thank for her statistical expertise, providing very helpful feedback and suggestions on various parts of this research. I wish to thank Dr. Linda Jackson for her expertise in the area of stereotype research. Dr. Pennie Foster—Fishman, I thank for the special assistance and support she extended. I would also like to extend a special thanks to Dr. Oliver J. Williams from the University of Minnesota for his expertise in the area of intimate partner violence in the African—American community iv and taking the time to attend committee meetings both via telephone and in person. I wish to thank my recruiters, Andre and Sean for all their hard work in recruiting participants for this research. I appreciate the time, effort, patience, and commitment that you put into this process as well as your close adherence to the research procedures. Last but certainly not least I would like to thank my husband Andre for his never-ending support and encouragement throughout this endeavor. In addition to the many hours he put into recruitment the love, patience, and understanding that he demonstrated were a comfort and source of strength throughout the this highly-demanding pI‘OCESS . TABLE OF CONTENTS List of Tables Chapter 1 Introduction Stereotypic Images of African—American Women How Stereotypes Influence the Perception and Treatment of African—American Women in General African—American Men’s Acceptance of Stereotypes About African-American Women Current Study Chapter 2: Method Recruitment Measures Chapter 3: Scale Construction Perceptions of African—American Women (PAAW) Scale Justification of Violence Scale Chapter 4: Results Demographics of Participants Men’s Endorsement of Stereotypes Men’s Justification of Partner Violence The Relationship Between Stereotypic Beliefs and Justification of Partner Violence Chapter 5: Discussion Appendix A: Recruitment Training Appendix B: Measure Appendix C: Demographic Pages Appendix D: CFA Model Bibliography vi 10 12 l4 16 16 20 24 24 32 35 35 36 4O 41 49 67 74 78 81 83 Table Table Table Table Table Table Table Table Table Table Table Table Table 10: 11: 12: 13: LIST OF TABLES Recruitment Sites 19 Exploratory Factor Analysis 27 Component Correlations 27 Psychometric Properties of the Matriarch and Jezebel Stereotype Sub-Scales 29 Item—Scale Correlations 31 Reliabilities for Justification of Violence Items 33 Demographics of Research Participants 37 Summary of Multiple Regression Results from Model 1: Relationship between stereotypic beliefs and justification of violence 42 Summary of Multiple Regression Results from Model 2: Relationship between holding the Jezebel stereotype and justification of violence 43 Summary of Multiple Regression Results from Model 3: Relationship between holding the Matriarch stereotype and justification of violence 44 Summary of Multiple Regression Results from Model 4: Effect of holding the Jezebel stereotype and justification of violence after accounting for the Matriarch stereotype 45 Summary of Multiple Regression Results from Model 5: Effect of holding the Matriarch stereotype and justification of violence after accounting for the Jezebel stereotype 46 Summary of Multiple Regression Results from Model 6: Relationship between holding positive perceptions of African—American women and justification of violence 48 vfi Introduction Domestic violence is a pervasive problem in our society. According to recent FBI statistics a woman is battered every nine seconds and four women a day are killed by their intimate partners. It is a problem that cuts across all races, cultures and social classes (Coley & Beckett, 1988; Williams, 1993; Williams, 1994). It is an issue that is thousands of years old but it was only in the 1970's that our country began to address this problem (Coley & Beckett, 1988; Taylor & Hammond, 1987). Many researchers have studied different aspects of this issue resulting in numerous published articles and books. But there is one aspect that has been neglected in this literature. The literature on battered women often overlooks the experiences of women of color (Asbury, 1987; Hampton, 1989; Harrison & Esgueda, 1999). Relatively little empirical community-based research has investigated ethnic differences and similarities in violence against women in U.S. sub—populations (Sorenson, 1996). Researchers have argued that it is important to examine the significance of race and culture in order to understand and respond appropriately to domestic violence and to develop culturally appropriate interventions (Brice- Baker, 1994; Coley & Beckett, 1988; Sorenson, 1996; Uzzell & Peebles—Wilkins 1989; Williams, 1992; Williams, 1993; Williams, 1994a; Williams, 1994b). With this in mind, any research that contributes to our knowledge of racial and cultural differences and similarities is of great value in helping us to better understand and work more successfully toward eliminating the problem of domestic abuse. The current study was designed to (I) examine African American men’s views of African American women, and (2) examine whether acceptance of the use of violence against African American women relates to stereotypic images of them. The existing literature on spouse abuse in the African American community is sparse (Williams, 1992; Williams, 1993; Harrison & Esgueda, 1999). Many researchers have been critical of the mainstream spouse abuse literature’s lack of attention to issues that are unique to African Americans (Asbury, 1987; Coley & Beckett, 1988; Hampton, 1989; Hampton, Gelles & Harrop, 1989; Hampton & Gelles, 1994). Limited publications have identified some of the factors that contribute to domestic violence in African American relationships (Asbury, 1987; Brown, 1985; Cazenave & Straus, 1979; Hampton, 1980; Hampton, 1989; Hine, 1989). A few authors have identified a culturally unique factor that may contribute to violence in interpersonal relationships between African Americans. This factor is the existence of stereotypic images of African American women (Asbury, 1987; Brice-Baker, 1994; Collins, 1991). The two most prevalent stereotypes of African-American women that have been theoretically linked to negative relationships between African—American men and women are the matriarch and jezebel stereotypes. The matriarch is defined as a woman who is overly aggressive, unfeminine, and who emasculates black men (West, 1995, Collins, 1991). The media has often depicted the matriarch as a physically large woman of brown or dark brown complexion whose primary role is to emasculate African American men with frequent verbal assaults, which are conducted in a loud, animated, verbose fashion (Jewell, 1993). The jezebel is defined as a whore, sexually aggressive, sexually promiscuous, and easily sexually aroused (West, 1995, Collins, 1991). This image originated during slavery when white slave owners exercised almost complete control over Black women's sexuality and reproduction (Jewel, 1993). One of the most prevalent images of antebellum America, she was a person governed almost entirely by her libido. She was in every way the counterimage of the mid-nineteenth-century ideal of the Victorian lady (White, 1985). The media has often depicted the jezebel as a mixed race woman with more European features who has functioned primarily in the role of a seductive, hypersexual, exploiter of men’s weaknesses (Jewel, 1993). Stereotypic images have the potential to negatively impact social relationships. Bethea (1995), Dickson (1993) and Willis (1989) have all identified stereotypes as destructive elements in African American interpersonal relationships, contributing to problems in the African American family. It is important that we examine the degree to which African American men may hold these stereotypes because it may have a great impact on how African American men perceive and interact with African American women, therefore affecting the relationships African American men and women have with each other. Belief in these stereotypes by African American men may influence domestic abuse against African American women. Men who believe that their partners are trying to emasculate them or who see women as sexually promiscuous may become angered by this. As a result they may inappropriately act out in violent ways in an attempt to control the situation. Even just the perception that the women are behaving in this manner may lead to violence. It may also cause victim blaming by fostering a belief that these women are at fault, that they provoke their husbands to abuse them (Brice-Baker, 1994). Acceptance of a stereotype not only influences the way in which information is encoded and interpreted about members of a categorized group, but also influences the behavior of both the perceiver and the stereotyped individual. Stereotypes influence information processing and subsequently affect perceptions and interactions with members of stereotyped groups, having important implications for one’s perception of and behavior toward these group members (Hamilton & Sherman, 1994). Stereotypes also influence power dynamics in personal interactions (West, 1995). This issue of power is a very important one when addressing intimate partner violence. When males feel that they are becoming powerless, violence or the threat of violence have been used by some to maintain power in the family (Campbell, 1981, Shepard & Pence, 1988). Clinically based literature has strongly supported the idea that men’s power and control issues underlie the expression and direction of violence towards women (Koss et. a1., 1994). How does this all relate to the African American community? If an African American male perceives African American women to be matriarchs and/or jezebels, if he believes “his woman” is attempting to emasculate him or is sexually promiscuous, he may feel as though he is powerless and that the only way to regain that power is to be physically abusive. In the case of the matriarch, he may feel as though he does not have control over her aggressiveness, her words and/or her attitude. In the case of the jezebel, he may feel as though he does not have sexual control over his partner. Therefore, the perceptions of African American women that these stereotypes may have fostered, along with feelings of powerlessness, may lead some African-American men to respond with violence against their partners. Although scholars have theorized about the power of stereotypes in general, until this study there was no research that attempted to assess the extent to which African American men may hold these stereotypic views of African American women, none that empirically examined the impact of these stereotypes on African American relationships, and none that examined a link between African American men’s belief in stereotypes and abuse of African American women. There were various (limited) literatures, however, that supported one or more of the following three ideas: 1) these stereotypic images exist, 2) these images influence the way in which African American women are perceived and treated by society in general, and/or 3) that African American men may hold stereotypes about African American women and that they may negatively impact African American relationships. These literatures are presented below. Stereotypic Images of African American Women White (1985) traces the historical development of the stereotypic image of the Jezebel back to the time of slavery when White slave owners used Black slave women for their sexual pleasure. Patricia Hill Collins (1991) identifies the matriarch and the jezebel as two of the controlling images that cause African American men to objectify African American women. Collins (1991) contends that these images, created by White Americans during the slave era, have served to control and oppress African American women and reflect the dominant group's interest in maintaining Black women’s subordination. With this in mind it is not difficult to conceive that a belief in these images by African American men may also prompt a desire to control, oppress, and subordinate African American women in relationships, especially since it is believed that some African American men may wish to become “masters” in their relationships by fulfilling traditional, Eurocentric, white defined definitions of masculinity. Collins (1991) also argues that if these African American men are blocked from doing this they may become dangerous to those closest to them, which may imply that attitudes formed from a belief in these stereotypes may increase the risk of violence in relationships between African American men and women. Asbury (1987) used an Afrocentric perspective to examine the experiences of African American women in violent relationships. From a review of the literature the author identified many factors that may be contributing to violence in African American relationships. These factors included 1) flexibility and fluidity with the roles of African American males and females, 2) economic difficulty, which is pervasive in many African American families, 3) early exposure of children to violence in some African American communities, 4) substance abuse, 5) arguments over children and pregnancy, and 6) questions about the wife’s fidelity and sexual problems. Asbury also identified factors in an African American woman’s decision to seek help which included 1) feelings of social isolation, 2) feelings that they may not be understood or welcomed at shelters, and 3) reluctance to seek help because she has internalized common stereotypes about African American women including that of sexual temptress, ugly mammies, bridges that hold the family together, and/or emasculating matriarchs. There was reference to both questions of the African American woman’s fidelity and the stereotypic images that may be internalized, including the jezebel and the matriarch. Brice-Baker (1994) examined domestic violence in African American and African Caribbean families. The author began by presenting factors that interfere with researchers being able to get an accurate estimate of the prevalence of domestic violence and then presented some theories that have been proposed to explain family violence. African American women have been stereotyped as 1) unattractive, 2) the glue that holds the family together, 3) matriarchs, and 4) love objects and sexual temptresses. These images suggest that African American women are somehow at fault for the violence they experience which as the author points out is another form of victim blaming. Through a content analysis of 54 pornographic videos, Cowan & Campbell (1994) found that African American women were portrayed as seductresses, sex objects to be exploited, sexually uncivilized and promiscuous more frequently than white women and were targeted with more acts of aggression. These images are consistent with the jezebel image. The author suggests that such portrayal of African American women has its roots in the stereotypic images that emerged during slavery. Ammons (1995) theorized that opinions of Black women have their genesis in slavery. According to her analysis, beliefs about African American women can be traced to the representations of Black women by the dominant culture, in other words, stereotypes. She speaks of the matriarch and the jezebel as two of the images created to keep Black women down. Ammons (1995) uses everything from examples of African's American women experiences in interacting with the justice system to lyrics from popular music to support the idea that stereotypes impact daily aspects of African American women’s lives. How Stereotypes Influence the Perception and Treatment of African American Women in General While no studies have examined the link between a belief in the jezebel and matriarch images and violence against African American women, a few researchers have addressed the link between these images and how African American women are perceived and treated by others. West (1995) discussed the historical origins of the matriarch and jezebel images, how they impact the psychological functioning of African American women, how they influence society’s treatment of African American women, and how they impact the relationships between psychotherapists and their African American women clients. The matriarch image was linked with chronic anger, psychosomatic conditions, depression, and low self—esteem, masking of vulnerability, relationship problems, and general avoidance or discomfort with displaying strong affect. Belief in this image by psychotherapists may influence their comfort level in their interactions with African American female clients. The jezebel image was connected with sexual exploitation, sexual dysfunction, shame, repression of sexual feelings, promiscuity, and victim blaming. Belief in this image by psychotherapists may influence their perceptions of their client’s sexuality. Helms (1979) argues that Black women have been overlooked and have been treated in a very cursory and denigrating manner by mental health professionals who have come to believe stereotypic images of African American women, particularly that of the matriarch. More recently, Priest (1991) talks about how stereotypes that counselors have about African Americans can negatively affect their interactions with African American clients, which lessens the effectiveness of treatment. Collins (1991) traces the historical development of the stereotypic images of African American women as matriarchs and jezebels and discusses how each image contributes to Black women's oppression and objectification. The author identifies these images as controlling and reflective of the dominant group’s interest in maintaining Black women’s subordination. She also identifies them as powerful influences on African American women's relationships with whites, African American men, each other, and themselves. African American Men’s Acceptance of Stereotypes About African American Women There is a limited literature that directly addresses the question of whether African American men may hold stereotypes of African American women. Staples (1982) suggested that Black families are under greater stress because of a belief held by many Black husbands that their wives will seek sexual satisfaction outside of marriage if they are not satisfied at home. He argued that jealousy in conjunction with community norms that encourage extramarital affairs and regard marriage as a license to physically dominate the woman contribute to violence in African American relationships. The references to the perceived “norm” of extramarital affairs on the part of 12 African American women and increased incidence of Black female aggression support the images of the jezebel and the matriarch. Willis (1989) suggested that when the African American male meets the African American female he sees someone whom he has been told is dominant in the family, “a castrating black woman (the matriarch).” He argues that African American males and females have been programmed from an early age, by society, to be destructive of each other, and as a result mate selection in the African American community is predicated on negative stereotypes which increases the likelihood of problems in the relationship. Plous & Williams (1995) conducted a survey to see whether racial stereotypes that developed during the days of slavery still persist in contemporary American society. The results of this study led the authors to conclude that racial stereotypes from the days of slavery are still present in American society and that what presumably began as White stereotypes of Blacks have now been embraced by the African American community. The authors found that African American respondents were more likely than others to endorse racial stereotypes. While we must be extremely cautious in generalizing these results to the African American population at large, considering that only 10% (about 67 respondents) were African American, the data support the idea that African Americans may, to some extent, internalize White stereotypes of African Americans (Plous & Williams, 1995). Taken together this literature identifies some ways in which stereotypic images have impacted the lives of African American women. If these images are held by African American men they are likely to play a key role in the quality of relationships between African American women and men. An understanding of the extent to which African American men hold these images can help us better understand violence in African American relationships and is important to building stronger, more positive, safer relationships. This would help lead to the betterment of the African American family and community and American society at large. Current Study As a result of the above review, the current study addressed the following research hypotheses: 1) Some African American men do hold these stereotypic views of African American women. 1a) Some African American men’s views of African American women support the jezebel stereotype. 14 1b) Some African American men’s views of African American women support the matriarch stereotype. 2).A belief in these stereotypic images by African American men may lead some to condone violence against African American women. In other words, African American men endorsing either one or both of these stereotypes will be more likely to condone male—to-female intimate partner violence when the woman displays behaviors consistent with the stereotypic images the men hold, such that: 2a) Those men with higher scores on the matriarch sub—scale will be more likely to condone male—to- female intimate partner violence when the woman behaves in ways that are consistent with the matriarch stereotype. 2b) Those men with higher scores on the jezebel sub-scale will be more likely to condone male-to— female intimate partner violence when the woman behaves in ways that are consistent with the jezebel stereotype. 2c) Those men with higher scores on both the matriarch and the jezebel sub-scales will be more likely to condone male-to—female intimate partner violence under all circumstances. 2d) Those men with higher scores on the positive perceptions of African American women will be less likely to condone male—to-female intimate partner violence under any circumstances. Method Recruitment Participants were recruited by one of two African American men who were specifically trained to be recruiters for this study. Recruiters were trained about (1) the purpose of the study, (2) what is and is not to be said to men during recruitment, (3) maintaining anonymity, and (4) lottery procedures (see Appendix A). The decision to use African American men as recruiters was part of an attempt to get more truthful answers and lessen the social desirability that may result from an African American woman (the researcher) asking African American men to fill out surveys about African American women. The recruiters went out individually to various sites for recruitment. The recruiters were told to approach African American men who were alone and who appeared to be within the targeted age range (18 and over). Potential participants were asked to complete a survey about relationships as part of a Michigan State University 16 graduate student’s study. It was explained to them that the survey would take about ten to fifteen minutes to complete, that it was anonymous, and that if they chose to complete the survey they could enter a lottery in which they might win one hundred dollars. The men were recruited from a wide range of locations in five cities in Michigan: Flint, Detroit, Lansing, East Lansing, and Okemos. These locations included Michigan State University's recreation facilities and main library; business locations; public buildings; Wayne State University; and public outdoor areas. It is important to point out a few things about research participants recruited from the two university campuses. Recreation facilities at Michigan State University are used not just by students but also by faculty and staff also. Therefore, not all men from this site were college students. Wayne State University serves a large population of non—traditional college students. Also, those areas targeted by recruiters at Wayne State were areas frequented by university staff also. Therefore, participants recruited from Wayne State University were not solely traditional undergraduate students, but non- traditional students and university employees as well. See Table 1 for a breakdown of recruitment sites. 17 A great deal of effort was devoted to obtaining participants from a wide variety of locations, with a broad range of age, incomes, education status, etc. Researchers were trained to approach any and all men eligible in a particular site. This effort was put forth in order to obtain a community—based sample more generalizable to African-American men. Recruiters went out to various sites at least twice a week and turned in questionnaires on a weekly basis. The researcher met with recruiters on a bi-weekly basis to discuss recruitment. On several occasions the researchers accompanied recruiters to recruitment sites and observed the recruitment process. During a recruitment period of three and a half months, a total of 255 African American men were recruited for participation in this study. Of the 255 surveys turned in to recruiters, 221 were useable. Thirty—four could not be used for the following reasons: 2 men identified as bisexual, 3 identified as gay/homosexual, 12 of the surveys were incomplete (a page or more), 2 of the participants were relatives of the recruiter, 12 men identified as some race other than African American, and 3 were not properly completed. Table 1 Recruitment Sites PUBLIC PLACES Malls Car dealership Professional cleaners facility Doughnut shop Car wash facility Fast food restaurant City streets Merchandise sales site Bus station UNIVERSITY CAMPUS FACILITIES Recreation facilities Libraries Academic buildings Eating facilities PUBLIC FACILITIES Recreation facility State building Public library Adult high school PLACES OF EMPLOYMENT Merchandise sales site Professional service sales site Automotive plant 19 ID 94 56 47 24 I o\° 25 21 11 Lottery procedure: To assure anonymity of the subjects the lottery information was collected in such a way that this information was unable to be connected to the subject's questionnaire. Postcards, which contained the personal information for the lottery, were kept together by each recruiter in an envelope separate from the questionnaires. At the completion of data collection, the postcards from the two recruiters were combined, a drawing was done and a one hundred dollar money order was mailed to the drawn participant. Measures The measure developed for this investigation was a self-constructed two—part questionnaire composed of two scales, the Perceptions of African American Women Scale and the Justification of Violence Scale (see Appendix B). The Perceptions of African American Women Scale was a 27—item scale that was designed to explore whether some African American men hold these stereotypic views of African American women. The scale was designed to include three sub-scales, the Matriarch sub—scale, the Jezebel sub-scale, and a Positive Perceptions sub-scale. The nine items of the Matriarch sub—scale described behaviors that were consistent with the matriarch stereotype (ex. “African 20 American women are too critical of their men”). The nine items of the Jezebel sub—scale described behaviors that were consistent with the jezebel stereotype (ex. “African American women are likely to sleep around”). The nine items of the Positive Perception sub-scale described behaviors that are positive and more desired characteristics of a mate (ex. “African American women deserve to be respected”). Using a six-point likert scale with responses ranging from strongly agree to strongly disagree the participants were asked to respond to each item as they believed them to be characteristic of African American women. The inclusion of the Positive Perception sub-scale on this part of the questionnaire was based on the fact that there are African American men who view African American women in very positive ways and the researcher did not want to force men to only respond to less desirable items. The Justification of Violence Scale was a 24-item scale designed to explore whether men would condone an act of male-to-female intimate partner violence when the female behaved in a way that was characteristic of the jezebel or matriarch images. This scale was designed to include two sub-scales. The twelve items of the Matriarch sub-scale described behaviors that were consistent with the matriarch 21 stereotype (ex. “puts down his manhood”). The eight items of the Jezebel sub—scale described behaviors that were consistent with the jezebel stereotype (ex. “is sexually unfaithful”). Using a six—point likert scale with responses ranging from very justified to very unjustified the participants were asked to indicate how justified they believed it was for a man to hit his partner under certain circumstances. Four filler items were included to attempt to screen out those men who may not have even read the items because the socially desirable thing to do is say that it is never all right to hit a woman and to place a “very unjustified” response under all conditions. The filler items included instances where a man might strike out in self-defense (ex. “tries to cut him with a knife”). Upon going through the questionnaires, the researcher decided not to exclude men from analyses, on this basis, for two reasons. First, 38% (83) of the men endorsed “very unjustified for every single item. Second, of those who did, there was adequate variability on the stereotype sub- scales, indicating that they were reading the questionnaire and most likely taking it seriously. These measures were designed after careful analysis of various psychological, sociological, African American studies, feminist, historical and legal literatures that 22 addressed and/or defined these stereotypes. Based on these readings, careful thought went into the creation of each item of these scales. It was the researcher’s belief that African American men may hold these stereotypes but are either not consciously aware that they do or deny that they hold these views. Many African American men may perceive African American women as matriarchs and jezebels but have not labeled them as such. In other words, they may believe that African American women possess the characteristics of those traits associated with the matriarch or jezebel image but due to lack of exposure to or familiarity with the terms have not linked the stereotypic label with them. Because of this, items did not use the terms jezebel or matriarch, but rather described behaviors that depict these images. Included in the questionnaire following the two scales described above were two pages of demographic questions for the respondents to complete. Demographics included race, age, income, education level, occupation level, relationship history, present relationship status, whether the participant had children, dating preference, and past racial dating history (see Appendix C). Scale Construction Perceptions of African American Women (PAAW) Scale The PAAW scale was originally designed with 27 items. Nine items (items # 2, 6, 9, 12, 15, 18, 21, 24, and, 27) were created to measure the perception of African-American women as matriarchs. Nine items (items # 3, 5, 8, 11, 14, 17, 20, 23, and 26) were created to measure the perception of African-American women as jezebels. Nine items were designed to reflect and measure positive perceptions of African-American women (items #1, 4, 7, 10, 13, 16, 19, 22, 25). These sub—scales were created to l) separately assess the endorsement of each particular stereotype in order to test hypothesis #1 and to 2) assess the relationship of each stereotype to beliefs about justification of violence, testing hypotheses #2a-d. Most items were fairly normally distributed, and the entire 18-item scale (minus the positive items), yielded an alpha of .93 (corrected item- total correlations ranging from .42—.76). The positive perceptions sub-scale yielded an alpha of .87 (corrected item-total correlations ranging from .50—.68). Confirmatory factor analysis (CFA) was used to assess for internal consistency and separation of each of the two stereotype sub-scales. The model tested the two hypothesized sub—scales, each indicated by nine items, with 24 the sub—scales allowed to covary freely (see Appendix D). The criteria used to evaluate the results of the CFA were the following goodness of fit statistics: the chi-square (x2); the comparative fit index (CFI); the root mean-square error of approximation (RMSEA); and the PCLOSE statistic. The CFI compares the fit of the hypothesized model with a null or baseline model in which all paths are fixed at zero. A CFI > .90 is typically seen as an indicator of good fit. The RMSEA estimates a value for model discrepancy in the population, corrected for model complexity. RMSEA values S .05 are typically seen as an indication that the model displays a close fit to the data. The PCLOSE statistic is a “p value” for testing the null hypothesis that the population RMSEA is no greater than .05 (Browne & Cudeck, 1993 & see Amos User’s Guide Version 3.6). The values were x2=437, df(134), .84 (CFI), .10 (RMSEA), and .00 (PCLOSE), which revealed basic problems in fit. Inspection of residuals indicated that lack of fit involved more than low loadings of items on their hypothesized construct. Due to the complexity of the lack of fit between the model and the data, the CFA was not helpful in suggesting modifications that would improve the measure. Because of this, the decision was made to conduct an exploratory 25 factor analysis (EFA) using all items from the intended stereotype sub-scales. The exploratory factor analysis method used an oblique rotation along with Kaiser’s criterion of rotating eigenvalues greater than one. The analysis revealed three factors. High loadings on the first factor included four items that were part of the intended matriarch sub-scale and two that were part of the intended jezebel sub-scale. High loadings on the second factor included five items that were intended for the jezebel sub—scale. The third factor showed high loadings for four items intended for the matriarch sub—scale as well as four items intended for the jezebel sub—scale (see Tables 2&3). The following criteria were used to extract items into two sub-scales from EFA results: factor loadings greater than .65; substantial loading only on one factor; and conceptually fitting with one or the other sub—scale (matriarch or jezebel). From the above criteria, a 4—item matriarch sub-scale (items 15, 21, 24, and 27) and a 4-item jezebel sub—scale (items 5, 11, 17, and 26) were extracted. 26 Table 2 Exploratory Factor Analysis I t em Com Components mua lit 1 2 3 ies PAAW2 expect too much .59 .41 .42 .76 PAAW3 too flirtatious .58 .23 .62 .62 PAAWS likely to cheat** .64 .48 .78 .36 PAAW6 too dominant .57 .62 .20 .63 PAAW8 are teases .57 .39 .58 .68 PAAW9 too aggressive .66 .37 .33 .81 PAAW11 are not faithful** .66 .40 .80 .44 PAAW12 are too demanding .76 .56 .41 .85 PAAW14 use sex to get what they want .51 .67 .48 .47 PAAW15 often insult their men* .63 .79 .37 .33 PAAW17 are likely to sleep around** .69 .55 .80 .43 PAAW18 are too controlling .69 .68 .40 .74 PAAW20 cannot trust to be faithful .29 .29 .54 .28 PAAW21 are too critical* .72 .82 .50 .54 PAAW23 often flirt to make their men .49 .68 .45 .33 jealous PAAW24 often talk down to their men* .77 .87 .42 .43 PAAW26 in their nature to cheat** .52 .38 .71 .25 PAAW27 often attach the manhood of .65 .80 .43 .40 their men* Percent of variance 46 8 7 *indicates items used in the matriarch sub—scale **indicates items used in the jezebel sub-scale Table 3 Component Correlations Component 1 2 3 1 1.00 .437 .459 2 .437 1.00 .420 3 .459 .420 1.00 27 Reliabilities were estimated on the smaller sub- scales, which yielded alphas of .85 for the matriarch sub- scale, .83 for the jezebel sub—scale, and .88 for all eight items combined. Corrected item total correlations ranged from .65-.77, .55-.72, and .54-.72, respectively (see Table 4). A correlation matrix was generated that correlated each of the eight items with the total eight— item scale and each of the newly formed sub-scales. This analysis revealed that the matriarch items correlated more highly with their own sub—scale than with the total scale or the jezebel sub—scale, and the jezebel items correlated more highly with their own sub—scale than with the total scale or the matriarch sub-scale(see Table 5). The correlation between the two stereotype sub-scales was .57. The result of the exploratory factor analysis and the correlation values supported the existence of two distinct sub-scales for use in subsequent analyses. Reliability was also estimated on the nine items designed to measure positive perceptions of African- American women. This analysis supported the existence of a Positive perceptions sub-scale, with an alpha of .87 and corrected item—total correlations ranging from .50-.68. 28 Table 4 Psychometric Properties of the Matriarch and Jezebel Stereotype Sub-scales Matriarch Item Corrected Item— Total Correlation PAAW 15 often insult their men .6496 PAAW 21 are too critical .7040 PAAW 24 often talk down to their men .7691 PAAW 27 often attack manhood of their men .6603 Alpha .8530 Scale mean 3.60* Scale standard deviation l.18* *with responses ranging from 1=strongly disagree to 6=strongly agree Jezebel Item Corrected Item—Total Correlation PAAW 5 likely to cheat .6764 PAAW 11 are not faithful .6991 PAAW 17 are likely to sleep around .7193 PAAW 26 in their nature to cheat .5465 Alpha .8034 Scale mean 2.90* Scale standard deviation 1.04* *with responses ranging from l=strongly disagree to 6=strongly agree 29 Table 4 (cont.) Positive Perception Sub-scale Item Corrected Item—Total Correlation PAAW 1 faithful to their men .5810 PAAW 4 supportive of their men .5537 PAAW 7 can trust not to cheat .4951 PAAW 10 are committed .6822 PAAW 13 can count on to stand by her man .6296 PAAW 16 deserve to be respected .5474 PAAW 19 are beautiful inside and out .6460 PAAW 22 have a loving nature .6381 PAAW 25 are caring individuals .6625 Alpha .8663 Scale mean 4.54* Scale standard deviation .827* *with responses ranging from 1=strongly disagree to 6=strongly agree 30 mcoHumHosnoo empownn004 ummeU OH QMSMMC Hflmflu Cfl QNZGANQ oczoum dooam ob >aoxfla mam ma3<