Analysis paper must include the following:
- Read the chosen articles
- Summarize the articles
Reflect on what the articles mean to you
, what thoughts or feelings were generated. (If you stopped here, you would be submitting a
· What is your reaction to the content of the articles?
Save your time - order a paper!
Get your paper written from scratch within the tight deadline. Our service is a reliable solution to all your troubles. Place an order on any task and we will take care of it. You won’t have to worry about the quality and deadlinesOrder Paper Now
· How do the articles relate to one another?
· What did you learn about violence & abuse in childhood?
Now you are ready to begin writing your analysis paper
. Create a thesis statement or argument that you will defend using the articles to justify your points. In essence, you take a stance on a particular issue, supported by 3-4 points supported by the literature
.Your thesis statement is often the last sentence of your opening introductory paragraph. Often, your thesis statement will be a “should” statement (e.g. social workers should take a feminist perspective when talking about the cause of physical abuse).
You then make your 3-4 points, using the literature to support your argument and linking the paragraphs together to make one seamless argument.
- Conclude with a summary of your argument and your major points as well as a brief discussion of how your argument impacts the clients you serve.
SIDE NOTE: I am providing you the articles. Please put together the ones ending CR3 into one paper 2-3 pages and the Extra into another paper 2-3 pages.
Evaluation will be based on how clearly you respond to the above, in particular:
a) The clarity with which you critique the articles;
b) The depth, scope, and organization of your paper; and,
c) Your conclusions, including a description of the impact of these articles on your thoughts about family violence.
TYPES OF VIOLENCE Childhood Sexual Abuse in the Lives of Black Women: Risk and Resilience in a Longitudinal Study Victoria L. Banyard Linda M. Williams Jane A. Siegel Carolyn M. West Victoria L. Banyard, PhD, is affiliated with the Department of Psychology, University of New Hampshire.Linda M. Williams, PhD, is affiliated with The Stone Center at Wellesley College. Jane A. Siegel, PhD, is affiliated with the Department of Sociology, An- thropology, and Criminal Justice at Rutgers University, Camden. Carolyn M. West, PhD, is affiliated with the Department of Interdisciplinary Arts and Sciences, University of Washing- ton, Tacoma. The research was supported by National Center on Child Abuse and Neglect (90-CA-1406) and Recovery from Sexual Abuse (90-CA-1552). Address correspondence to: Victoria L. Banyard, PhD, Department of Psychology, Uni- versity of New Hampshire, Conant Hall, 10 Library Way, Durham, NH 03824 (E-mail: vlbcisunix.unh.edu). [Haworth co-indexing entry note]: “Childhood Sexual Abuse in the Lives of Black Women: Risk and Resil- ience in a Longitudinal Study.” Banyard, Victoria L. et al. Co-published simultaneously inWomen & Therapy (The Haworth Press, Inc.) Vol. 25, No. 3/4, 2002, pp. 45-58; and:Violence in the Lives of Black Women: Battered, Black, and Blue(ed: Carolyn M. West) The Haworth Press, Inc., 2002, pp. 45-58. Single or multiple copies of this article are available for a fee from The Haworth Document Delivery Service [1-800-HAWORTH, 9:00 a.m. – 5:00 p.m. (EST). E-mail address: [email protected]]. 2002 by The Haworth Press, Inc. All rights reserved.45 SUMMARY.Childhood sexual abuse (CSA) is associated with long-term mental health consequences. This article reviews the results of one longitu – dinal study, whose sample consisted primarily of African American women. The purpose is to give voice to an understudied group of CSA sur – vivors and tohighlight the variability in risk and protective factors. Key find – ings related to mental health consequences, re-traumatization, and resilience are reviewed and set within the broader context of research on African Ameri – can women and child sexual abuse. Implications for future research and clini – cal practice are discussed. [Article copies available for a fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mail address: Website: © 2002 by The Haworth Press, Inc. All rights reserved.] KEYWORDS.Blacks, trauma, mental health, re-victimization Childhood sexual abuse (CSA) has been documented in all segments of society (e.g., Finkelhor, Hotaling, Lewis, & Smith, 1990) with docu- mented long-term consequences for women’s psychological functioning and well-being (e.g., Fergusson & Mullen, 1999). Black feminist survi- vors (Wilson, 1994) have documented many painful personal accounts of childhood sexual abuse. Empirical researchers have also discovered child sexual abuse in this population. For example, Wyatt, Loeb, Solis, and Carmona (1999) reported a 10-year comparison of prevalence rates of child sexual abuse in a community sample of African American and Euro- pean American women. There were no significant changes in reported rates of abuse over 10 years and no racial differences in the characteristics of the abuse experience. However, 29% of the African American women in the sample reported child sexual abuse in the 1994 survey. Although this was lower than the 39% prevalence rate for European American women, it is still a substantial rate of victimization. Research has also doc- umented the wide-ranging negative mental health consequences associ- ated with experiencing child sexual abuse (e.g., Beitchman et al., 1992). These findings suggest that CSA is an important element in understanding Black women’s health (Lawson, Rodgers-Rose, & Rajaram, 1999). Kenny and McEachern (2000) reviewed findings related to ethnicity and child sexual abuse, including patterns of prevalence, disclosure, and consequences. They called for further research that explores the experi- ences of ethnically diverse individuals. Feminist scholars, particularly Black feminist researchers, have asserted the need to reexamine notions of 46 VIOLENCE IN THE LIVES OF BLACK WOMEN difference and to move beyond treating all women as a homogenous group (Collins, 2000; Hare-Mustin & Marecek, 1990; West, 2002). Such cri – tiques have important implications for the study of child sexual abuse, calling for greater attention to how child sexual abuse is experienced in the lives of understudied groups of women. Moreover, in order to understand variability among survivors, researchers need to move beyond simply com – paring survivors to non-survivors (West,Williams, & Siegel, 2000). In response to such feminist critiques, this article will describe the find – ings of a longitudinal study of women, primarily African American, as a focal point for examining patterns of risk and resilience in the lives of CSA survivors. The study’s findings have been discussed in detail else – where (McCahill, Meyer, & Fischman, 1979; Banyard & Williams, 1996; Banyard, Williams, & Siegel, 2000c). The purpose of this article is to summarize and review these findings in the context of other studies of child sexual abuse, particularly studies that focus on African American survivors. THE WOMEN’S STUDY Participants The Women’s Study, which began in the early 1970s, has followed a group of female childhood sexual abuse survivors and a comparison sam- ple for more than 25 years. Participants were drawn from a sample of 206 victims of child sexual abuse who were examined in the emergency room of a large city hospital from 1973 to 1975. They and/or family members were interviewed at the time as part of a larger study on the consequences of sexual assault (McCahill et al., 1979). The sample was composed pri- marily of African American girls (84%) who ranged in age from 10 months to 12 years at the time of the abuse. The reported sexual abuse in- volved sexual contact by force, threat of force, misuse of authority, or by a person who was five or more years older than the child, whether or not force was used. The abuse ranged from genital fondling to sexual intercourse and was perpetrated by a wide range of individuals–fathers, stepfathers, other fam- ily members, friends, acquaintances, and strangers (all males). Soon after each girl was seen in the hospital, consent to participate in the study was obtained from the children and/or their caregivers, and then the child and caregiver were interviewed. Types of Violence 47 An average of 17 years after the abuse, 136 women (66% of the original sample) were located through phone directories, official and government records, and neighborhood canvassing. The survivors were reinterviewed in 1990 or 1991 as part of a follow-up study on the consequences of child sexual abuse. The mean age for the sample at the time of reinterview was 25.5 years. Most (86%) were African American, and most (61%) had never been married. Fifty percent of the women had a high school diploma or GED equivalent. Although 29% were working full- or part-time, the majority of women were unemployed (64%). Their median personal in – come was less than $8,000 in the year before the interview. In 1997, which was 23 years after the abuse, 87 of the original 206 girls were recontacted and reinterviewed. In order to make comparisons, we also contacted 87 women who were seen, but not for sexual abuse, at the same city hospital in the early 1970s. On average, the women were 31.6 years old at this time, and 89% identified themselves as African Ameri- can. Women interviewed at wave 2 were more likely to have been sexu- ally abused by a family member than non-interviewed women and, at wave 3, a greater proportion of interviewed women were African Ameri- can. Overall, however, interviewed and non-interviewed women did not differ on demographic variables or characteristics of their abuse. Procedures Institutional review board procedures were followed for obtaining con- sent. Before each interview, informed consent was obtained, which in- cluded a description of the sensitive nature of the interview material. Participants were advised that they could stop the interview at any time. The caregivers who were interviewed in the early 1970s were not reinter- viewed. All waves of the data were collected through face-to-face interviews. The interviewers were two women, one European American and one Afri- can American, who had received training and supervision to ensure that they were able to establish rapport with the women and conduct inter- views with sensitive and potentially upsetting personal topics. Although the interviewers knew the purpose of the study, they were blind to the de- tails of the women’s histories of victimization. Although a few participants were interviewed in their own homes, al- most all were interviewed in a private office. Each interview began with questions about more neutral aspects of the woman’s life, such as educa- tion and employment status. After sufficient rapport had been established, the interviewers asked questions about other topics such as relationships 48 VIOLENCE IN THE LIVES OF BLACK WOMEN in her family of origin, drug and alcohol use, sexual history, psychological functioning, and detailed questions about sexual victimization. On average, the interviews lasted three hours and were followed by a debriefing period, which gave the women an opportunity to ask questions. There was also a discussion of how she was feeling about the interview. Interviews were suspended if the interviewer determined that the partici – pant was in any distress. Counseling services from a local sexual abuse treatment center were made available to participants if they were inter – ested. REVIEW OF THE RESEARCH The participants’ voices in the Women’s Study revealed the negative consequences associated with child sexual abuse, including a variety of mental health problems, difficulties functioning as a parent, and elevated rates of aggressive behavior. The women also demonstrated how re- sponses to trauma might vary based on the number and types of traumatic experiences. Equally as important, the Women’s Study discovered that survivors, despite their trauma, could be very resilient. Consequences of Child Sexual Abuse Several articles, based on data from The Women’s Study, point to the negative impact of child sexual abuse on the health and well-being of Afri- can American women. Banyard, Williams, and Siegel (2001b) used wave 3 of the study to investigate the adult mental health functioning of CSA survivors and the comparison group. When compared to the non-abused group, survivors of child sexual abuse experienced higher levels of anxi- ety, depression, dissociation, sexual concerns, intrusive symptoms, and an impaired sense of self. The size of the sample in the Women’s Study also permitted the exami- nation of variations among survivors. Banyard, Williams, and Siegel (2001d) discovered that some survivors experienced negative reactions when they disclosed their abuse. For example, the person they told was unsupportive, blamed the survivor, called the survivor a liar, or punished the survivor. These survivors reported higher levels of dissociative symp- toms. Other analyses examined the association between mental health func- tioning and characteristics of the child sexual abuse, such as the use of force, the survivor’s relationship with the perpetrator, and level of Types of Violence 49 invasiveness (Banyard & Williams, 1996; Jasinski, Williams, & Siegel, 2000). Survivors who had experienced force reported higher levels of such problems as depressive symptoms, anxiety, and sleep disturbance. Women who were abused by a family member, when compared to survi – vors who were victims of extrafamilial abuse, were more likely to report an array of mental health symptoms. Higher levels of anxiety and heavy drinking in adulthood also were found among survivors who were older at the time of the abuse and among survivors who had experienced multiple instances of CSA. These findings fit with both the broader literature on the negative effects of child sexual abuse and itslinks to a broad array of men – tal health effects across the life span (see Jumper, 1995; Kendall-Tackett, Williams, & Finkelhor, 1993; Neumann, Houskamp, Pollock, & Briere, 1996; Polusny & Follette, 1995 for reviews) and studies that specifically ex – amine the experiences of AfricanAmerican women (e.g., Rhodes, Ebert, & Meyers, 1993; Russo, Denious, Keita, & Koss, 1997; Thompson, Kaslow, Lane, & Kingree,2000;Wyatt &Newcomb, 1990; West, 2002). Measures used in the Women’s Study also permitted examination of the impact of child sexual abuse on specific behaviors, including the abil- ity to function as a parent and adult aggressive behavior. Banyard, Wil- liams and Siegel (2001a) discovered that adult sexual assault and partner violence, but not CSA, were associated with more negative parenting out- comes. This differs from other studies that found links between child sex- ual abuse and parenting difficulties, such as negative perceptions of one’s self as a parent and greater use of physical punishment (Banyard, 1997; DiLillo, 2001). Future research should explore the impact of CSA on Afri- can American women’s family roles. Other areas of behavior examined in the Women’s Study included criminality and aggression. Several feminist scholars have posited that child sexual abuse is an important factor in the onset of delinquency be- cause it can lead girls to run away, which, in turn, forces them to engage in criminal behavior for survival (Belknap, Holsinger, & Dunn, 1997; Chesney-Lind, 1997). Richie (1996) has shown that victimization plays an important role in what she has referred to as thegender entrapmentof African American women in which women’s criminal behavior is inextri- cably tied with their involvement in violent relationships with their part- ners. In addition, sexual victimization has been associated with aggression in children (Kendall-Tackett et al., 1993); however, before the Women’s Study, there had been few attempts to determine if this relationship per- sisted beyond childhood. Analyses of official arrest records by Siegel and Williams (2001) found that the child sexual abuse victims were at higher risk of arrest than were 50 VIOLENCE IN THE LIVES OF BLACK WOMEN the women in the comparison group as adults, but not as juveniles. How – ever, the victims of abuse were significantly more likely to have been ar – rested for violent offenses in both adolescence and adulthood. African Americans were no more likely to have been arrested than White women, either as juveniles or adults, except for violent offenses; in that case, the African American women faced a substantially higher risk of arrest than the White women. The increased risk of violence faced by victims of sexual abuse was re – flected as well in self-reports of extensive aggressive behavior in the form of physical fighting, with both intimate partners and others (Siegel, 2000). Women with a history of repeated sexual victimization and exposure to other forms of violence, such as witnessing violence, experiencing beat – ings in childhood, and having family members or close friends who were murdered, were at significantly greater risk of perpetrating violence. Those who were victims of violence at the hands of their intimate partners were also more likely to report engaging in physical fighting. Thus, the women’s own violence was coupled with another troubling aspect of their lives: vulnerability to repeat victimization. Re-Victimization Linkages between a history of child sexual abuse and re-victimization in adulthood, defined as the occurrence of at least one incident of sexual abuse during childhood followed by a subsequent incident of adult physi- cal or sexual victimization, have been reported in several studies, includ- ing those with samples of African American women (e.g., Roodman & Clum, 2001; Wyatt, Notgrass, & Gordon, 1995). Almost all of these stud- ies, however, have been based on retrospective reports by adults. In addi- tion to experiencing violence in intimate relationships, many Black women are exposed to a broad array of traumatic experiences, including community violence (Jenkins, 2002). Re-victimized Black survivors are espe- cially vulnerable to mental health problems (Russo et al., 1997). For exam- ple, almost one-quarter of the re-traumatized women surveyed by Hien and Bukszpan (1999) met diagnostic criteria for lifetime posttraumatic stress disorder. The Women’s Study extends our knowledge of re-victimization be- cause it assessed for various forms of trauma in adulthood, including wit- nessing violence, experiencing a serious accident, the traumatic loss of a friend or family member, and various forms of physical and sexual as- saults over the life span. Siegel and Williams (2001) found that the risk of adult re-victmization was not shared equally among all those in the Types of Violence 51 Women’s Study who experienced child sexual abuse. Women whose only experience of youthful sexual victimization occurred when they were un – der age 13 were not at increased risk of adult sexual re-victimization com – pared to those who experienced no child sexual abuse. However, a small group of women were at extremely high risk of adult sexual and physical victimization if they had been victimized as young children (i.e., under age 13) and again in adolescence. In addition, West, Williams, and Siegel (2000) found that survivors who had experienced sexual abuse in both childhood and adulthood were also at increased risk for partner violence in their adult intimate relationships. Furthermore, sexual re-victimzation had an adverse impact on Black women’s reproductive and sexual health. Specifically, when compared to women sexually abused in childhood only, re-victimized women experi – enced more problems with conceiving, repeated vaginal infections, sexu – ally transmitted diseases, and painful intercourse (West et al., 2000). Banyard, Williams, and Siegel (2001b) found that exposure to multiple traumatic events (broadened beyond exposure to sexual re-victimization) across the life span were associated with higher levels of psychological distress. Furthermore, the relationship between child sexual abuse and such outcomes as depressive symptoms, anxiety, and negative feelings about self was mediated by reports of exposure to this variety of other traumatic events. Thus, for some symptoms, the links between child sex- ual abuse and adult mental health symptoms are explained by links be- tween child sexual abuse and re-traumatization. Taken together, with the aforementioned risk of self-reported aggression for those exposed to other forms of violence, these findings underscore the need for careful addi- tional attention to the links between child sexual abuse and a web of other traumatic events. The longitudinal nature of the Women’s Study also permitted examina- tion of risk and protective factors for repeated trauma exposure within the sample of child sexual abuse survivors. West et al. (2000) found that child sexual abuse survivors whose abuse involved physical force were at greater risk for reexperiencing abuse in the form of adultsexual assault. Banyard, Williams, and Siegel (2001c) found increased risk for a wide range of trauma re-exposure in adulthood for women who also experienced extreme poverty, family-of-origin difficulties, depression and dissociation, and sub- stance abuse. Social support was a protective factor. These findings fit with earlier work on risk factors by researchers, such as Koss and Dinero (1989), but extend the findings to a community sample of women, most of whom were African American. These findings further support the need to attend to important differences among women–not only differences between Black 52 VIOLENCE IN THE LIVES OF BLACK WOMEN women and other survivors, but also variability in the experiences of African American women themselves. Resilience In addition to documenting the ways in which child sexual abuse places African American women at risk for re-traumatization and psychological distress, the Women’s Study also sought to examine the complex phe – nomenon of resilience. Despite having experienced childhood sexual abuse, resilient women are functioning relatively well in many aspects of their lives. Hyman and Williams (2001) defined resilience as “physical health, mental health, interpersonal relationships, adherence to commu – nity standards, and economic well being” (p. 203). However, 40 survivors (29% of the sample) had high scores on the re – silience scale, and 25 women (18% of the sample) demonstrated what the researchers referred to asexcellent resilience, which was characterized by competent functioning in nearly all areas assessed. The highly resilient women were less likely to have experienced incest or severe child physi- cal abuse. In addition, they were more likely to have been reared in stable homes, which meant that they experienced fewer moves or foster care placements and less parental drug abuse. Income level of family of origin was not related to resilience. Receiving abuse-specific therapy was not significantly related to resilience, though it should be noted that a rela- tively small portion of the sample of women received this type of profes- sional intervention. Graduation from high school was also predictive of resilience for this sample. For this group of women, social support in the form of receiving sup- port from someone special and significant in their lives was an important protective factor characteristic of the more resilient women. In a similar vein, Siegel (2000) found that women who reported a strong feeling of at- tachment to their mothers were less likely to be in abusive relationships, which, in turn, lowered the risk that they themselves would engage in vio- lence. These results suggest the need to attend not only to strengths within the survivor, but also to building strengths and supports around the survi- vor–including stabilizing the family of the abused child and strengthening informal support networks. These findings fit with the broader literature on understanding compe- tent outcomes following child sexual abuse (e.g., Spaccarelli & Kim, 1995; Valentine & Feinauer, 1993). Harvey (1996), for example, dis- cusses the importance of understanding a variety of ecological factors that may predict recovery from sexual abuse. She highlights the fact that many Types of Violence 53 survivors do not seek professional help but may find other paths to recov – ery. The Women’s Study supports this notion in that only a minority of the study’s participants was in psychotherapy, yet many were able to find paths toward healing through other connections. This study extends dis – cussions of resilience that have more often focused on European Ameri – can women’s experiences, and illustrates the important strengths of African American female survivors. SUGGESTIONS FOR INTERVENTION In summary, a variety of research points to the importance of understand – ing child sexual abuse in the lives of African American women. The voices of participants in the Women’s Study remind us of the negative conse – quences of CSA, including the ways in which it places survivors at greater risk for mental health concerns, such as depression, anxiety, traumatic intru- sions, and impaired views of the self, as well as increased contact with the criminal justice system and concerns in relationships with others. These women also tell us of the variability in survivors’ responses to trauma as not all women experienced the same constellation of effects. Furthermore, these survivors remind us of the strengths of survivors of child sexual abuse, and of Black women in particular. Furthermore, although a relatively small portion of the women used therapy to address their history of abuse (Hyman & Williams, 2001), practitioners should be prepared to assist African Amer- ican survivors of childhood sexualabuse. The following suggestions are made for clinical practice: Establish competence. Therapists should become familiar with a vari- ety of topics, including the psychobiology of trauma; research on memo- ries for trauma, dissociation, and posttraumatic stress disorder; and variables associated with resilience and coping (Enns et al., 1998). Conduct a thorough assessment. Service providers should consider us- ing self-report measures, such as the Trauma History Screen, which is de- signed to enable clinicians to gather information about a broad range of traumatic experiences (Allen, Huntoon, & Evans, 1999). Black women are also at risk for community violence and racial trauma, such as racial profiling by the police (Daniel, 2000; Jenkins, 2002). These forms of trauma should not be overlooked. Assessment and treatment strategies need to consider the complex web of trauma exposure that maycharacter- ize the experiences of some survivors. Treatment plan. It is especially important to be attentive to how the experi- ence of child sexual abuse is influenced by the client’s race, social class, and 54 VIOLENCE IN THE LIVES OF BLACK WOMEN sexual orientation (Enns, 1996). A culturally sensitive treatment plan for Af – rican American survivors may involve addressing such topics as spirituality or stereotypes about Black women’s sexuality (Robinson, 2000; Wilson, 1994). Future work should focus on merging the growing fields of culturally sensitive therapeutic intervention and trauma treatment, work that is being done by Black feminist scholars (Daniel, 2000). Develop a support system. A strong support system has been shown to improve the psychological functioning of victimized Black women and to reduce their likelihood of suicidal behavior (Young, Twomey, & Kaslow, 2000). If necessary or appropriate, the service provider may need to educate relatives, friends, and community members about the risk factors and symp – toms associated with childhood sexual abuse (Fontes, Cruz, & Tabachnick, 2001). Tully (1999) encourages practitioners to recognize the ways in which traditional strengths of African American communities may help the survivor cope in the aftermath of victimization. In conclusion, the Women’s Study provides further evidence of the im- portance of understanding child sexual abuse in the lives of African Amer- ican survivors. The Women’s Study helps document both the long-term negative effects of CSA as well as the fact that despite these life chal- lenges, many survivors are resilient and develop effective coping strate- gies. The voices of participants from the Women’s Study encourage us to continue to develop interventions that appreciate the complex differences among women and to support research that gives voices to the perspec- tives of various survivors of child sexual abuse. REFERENCES Allen, J. G., Huntoon, J., & Evans, R. B. (1999). A self-report measure to screen for trauma history and its application to women in inpatient treatment for trauma-related disorders.Bulletin of the Menninger Clinic,63, 422-429. Banyard, V. L. (1997). The impact of childhood sexual abuse and family functioning on four dimensions of women’s later parenting.Child Abuse & Neglect,21, 1095-1107. Banyard, V. L., & Williams, L. M. (1996). Characteristics of child sexual abuse as corre- lates of women’s adjustment: A prospective study.Journal of Marriage and the Fam- ily,58, 853-865. Banyard, V. L., Williams, L. M., & Siegel, J. A. (2001a).The impact of trauma and de- pression on parenting: An exploration of mediating risk and protective factors.Un- published manuscript. Banyard, V. L., Williams, L. M., & Siegel, J. A. (2001b). The long-term mental health consequences of child sexual abuse: An exploratory study of the impact of multiple traumas in a sample of women.Journal of Traumatic Stress,14, 697-715. Types of Violence 55 Banyard, V. L., Williams, L. M., & Siegel, J. A. (2001c).Retraumatization among adult women sexually abused in childhood: Exploratory analyses in a prospective study of predominantly African American women. Unpublished manuscript. Banyard, V. L., Williams, L. M., & Siegel, J. A. (2001d). Understanding links among childhood trauma, dissociation, and women’s mental health.American Journal of Orthopsychiatry,71, 311-321. Beitchman, J. H., Zucker, K. J., Hood, J. E., DaCosta, G. A., Akman, D., & Cassavia, E. (1992). A review of the long-term effects of child sexual abuse.Child Abuse & Ne – glect,16, 101-118. Belknap, J., Holsinger, K., & Dunn, M. (1997). Understanding incarcerated girls: The re – sults of a focus group study.Prison Journal,77, 381-404. Chesney-Lind, M. (1997).The female offender: Girls, women, and crime. Thousand Oaks, CA: Sage. Collins, P. H. (2000).Black feminist thought: Knowledge, consciousness, and the politics of empowerment. New York: Routledge. Daniel,J. H. (2000). The courage to hear: African American women’s memories of racial trauma. In L. C. Jackson & B. Greene (Eds.),Psychotherapy with African American women: Innovations in psychodynamic perspectives and practice(pp. 126-144). New York: Guilford Press. DiLillo, D. (2001). Interpersonal functioning among women reporting a history of child- hood sexual abuse: Empirical findings and methodological issues.Clinical Psychol- ogy Review,21, 553-576. Enns, C. Z. (1996). The Feminist Therapy Institute code of ethics: Implications for work- ing with survivors of child sexual abuse.Women & Therapy,19, 79-91. Enns, C. Z., Campbell, J., Courtois, C. A., Gottlieb, M. C., Lese, K. P., Gilbert, M. S., & Forrest, L. (1998). Working with adult clients who may have experienced childhood abuse: Recommendations for assessment and practice.Professional Psychology: Re- search and Practice,29, 245-256. Fergusson, D. M., & Mullen, P. E. (1999).Childhood sexual abuse: An evidence based perspective. Thousand Oaks, CA: Sage. Finkelhor, D., Hotaling, G., Lewis, I. A., & Smith, C. (1990). Sexual abuse in a national survey of adult men and women: Prevalence, characteristics, and risk factors.Child Abuse & Neglect,14, 19-28. Fontes, L. A., Cruz, M., & Tabachnick, J. (2001). Views of child sexual abuse in two cul- tural communities: An exploratory study among African Americans and Latinos. Child Maltreatment,6, 103-117. Hare-Mustin, R. T., & Marecek, J. (1990).Making a difference: Psychology and the con- struction of gender. New Haven: Yale University Press. Havey, M. R. (1996). An ecological view of psychological trauma and trauma recovery. Journal of Traumatic Stress,9, 3-23. Hien, D., & Bukszpan, C. (1999). Interpersonal violence in a “normal” low-income con- trol group.Women & Health,29, 1-16. Hyman, B., & Williams, L. (2001). Resilience among women survivors of child sexual abuse.Affilia,16, 198-219. 56 VIOLENCE IN THE LIVES OF BLACK WOMEN Jasinski, J. L., Williams, L. M., & Siegel, J. (2000). Childhood physical and sexual abuse as risk factors for heavy drinking among African American women: A prospective study.Child Abuse & Neglect,24, 1061-1071. Jenkins, E. J. (2002). Black women and community violence: Trauma, grief, and coping. Women & Therapy,25(3 & 4), 29-44. Jumper, S. A. (1995). A meta-analysis of the relationship of child sexual abuse to adult psychological adjustment.Child Abuse & Neglect,19, 715-728. Kendall-Tackett, K. A., Williams, L. M., & Finkelhor, D. (1993). Impact of sexual abuse on children: A review and synthesis.Psychological Bulletin,113, 164-180. Kenny, M. C., & McEachern, A. G. (2000). Racial, ethnic, and cultural factors of child – hood sexual abuse: A selected review of the literature.Clinical Psychology Review, 20, 905-922. Koss, M. P., & Dinero, T. E. (1989). Discriminant analysis of risk factors for sexual vic – timization among a national sample of college women.Journal of Consulting and Clinical Psychology,57, 242-250. Lawson, E. J., Rodgers-Rose, L., & Rajaram, S. (1999). The psychosocial context of Black women’s health.Health Care for Women International,20, 279-289. McCahill, T., Meyer, L. C., & Fischman, A. (1979).The aftermath of rape. Lexington, MA: Lexington Books. Neumann, D. A., Houskamp, B. M., Pollock, V. E., & Briere, J. (1996). The long-term sequelae of childhood sexual abuse in women: A meta-analytic review.Child Mal- treatment,1, 6-16. Polusny, M. A., & Follette, V. M. (1995). Long-term correlates of child sexual abuse: Theory and review of the empirical literature.Applied and Preventive Psychology,4, 143-166. Rhodes, J. E., Ebert, L., & Meyers, A. B. (1993). Sexual victimization in young, pregnant and parenting, African American women: Psychological and social outcomes.Vio- lence and Victims,8, 153-163. Richie, B. E. (1996).Compelled to crime: The gender entrapment of battered Black women. New York: Routledge. Robinson, T. L. (2000). Making the hurt go away: Psychological and spiritual healing for African American women survivors of childhood incest.Journal of Multicultural Counseling and Development,28, 160-176. Roodman, A. A., & Clum, G. A. (2001). Revictimization rates and method variance: A meta-analysis.Clinical Psychology Review,21, 183-204. Russo, N. F., Denious, J. E., Keita, G. P., & Koss, M. P. (1997). Intimate violence and Black women’s health.Women’s Health: Research on Gender, Behavior, and Policy, 3, 315-348. Siegel, J. A. (2000). Aggressive behavior among women sexually abused as children.Vi- olence and Victims,15, 235-255. Siegel, J. A., & Williams, L. M. (2001).The relationship between child sexual abuse and female delinquency and crime: A prospective study. Unpublished manuscript. Spaccarelli, S., & Kim, S. (1995). Resilience criteria and factors associated with resil- ience in sexually abused girls.Child Abuse & Neglect,9, 1171-1182. Types of Violence 57 Thompson, M. P., Kaslow, N. J., Lane, D. B., & Kingree, J. B. (2000). Childhood mal – treatment, PTSD, and suicidal behavior among African American females.Journal of Interpersonal Violence,15, 3-15. Tully, M. A. (1999). Lifting our voices: African American cultural responses to trauma and loss. In K. Nader, N. Dubrow, & B. H. Stamm (Eds.),Honoring differences: Cul – tural issues in the treatment of trauma and loss(pp. 23-48). Philadelphia, PA: Brunner/Ma – zel. Valentine, L., & Feinauer, L. L. (1993). Resilience factors associated with female survivors of childhood sexual abuse.The American Journal of Family Therapy,21, 216-224. West, C. M. (2002). Black battered women: New directions for research and Black femi – nist theory. In L. H. Collins, M. Dunlap, & J. Chrisler (Eds.),Charting a new course: Psychology for a feminist future. (pp. 216-237) Westport, CT: Praeger. West, C. M., Williams, L. M., & Siegel, J. A. (2000). Adult sexual revictimization among Black women sexually abused in childhood: A prospective examination of serious consequences of abuse.Child Maltreatment,5, 49-57. West, C. M. (2002). Battered, Black, and blue: An overview of violence in the lives of Black women.Women & Therapy,25(3 & 4), 5-27. Wilson, M. (1994).Crossing the boundary: Black women survive incest. Seattle, WA: Seal Press. Wyatt, G. E., & Newcomb, M. (1990). Internal and external mediators of women’s sexual abuse in childhood.Journal of Consulting and Clinical Psychology,58, 758-767. Wyatt, G. E., Loeb, T. B., Solis, B., Carmona, J. V., & Romero, G. (1999). The prevalence and circumstances of child sexual abuse: Changes across a decade.Child Abuse & Ne- glect,23, 45-60. Wyatt, G. E., Notgrass, C. M., & Gordon, G. (1995). The effects of African American women’s sexual revictimization: Strategies for prevention. In C. F. Swift (Ed.),Sexual assault and abuse: Sociocultural context of prevention(pp. 111-134). Binghamton, NY: The Haworth Press, Inc. Young, S., Twomey, H., & Kaslow, N. J. (2000). Suicidal behavior in African American women with a history of childhood maltreatment. In T. Joiner & M. D. Rudd (Eds.), Suicide science: Expanding the boundaries(pp. 221-230). Boston: Kluwer Academic Publishers. 58 VIOLENCE IN THE LIVES OF BLACK WOMEN
VIOLENCE AGAINST WOMEN ‘J/ July 2002Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE An Empirical Invbstigatinon of thb Psychological Sbqublab nof Childhood Sbxual Abusb in an Adult Latnina Population JOSIE VEGA HINSON Fullbr Thbological Sbminnary CATHERINE KOVEROLA fnivbrsity of Marylannd, Baltimorb MARY MORAHAN Los Angblbs County andn fnivbrsity of Southnbrn California Violbncb Intbrvbntion Program This study bxaminbd thb bffbcts of childhood sbxual abusb on pbrcbivbd social support among Latino wombn using thb Sblf-In-Rblation thborbtical oribntation and thb Comprb- hbnsivb Modbl of Trauma Impact to bxaminb thb impact of child sbxual abusb. Partici- pants includbd 54 Latina survivors of child sbxual abusb sbbking mbntal hbalth sbrvicbs. Thb rblationship bbtwbbn affbctivb distrbss, intbrpbrsonal sbnsitivity, and pbrcbivbd social support was invbstigatbd. Findings rbvbalbd dbprbssion was a statistically signifi- cant prbdictor of intbrpbrsonal sbnsitivity, sr(3, 49) = .45, p< .005) and of pbrcbivbd social support for both Important Pbrson, r(47) = –.34, p< .05, and Pboplb in Pbrsonal Lifb, r(47) = –.38, p< .01. Clinical implications arb discussbd from a Latino cultural pbrspbctivb. Thb long-tbrm nbgativb impact of child sexual abuse on the survivor has been extensively documented (Briere & Runtz‘b 1988; Bryer‘b Nelson‘b Miller‘b & Krol‘b 198‘f; Finkelhor & Browne‘b 1985; Landecker‘b 1992; Murphy et al.‘b 1988; Roesler & McKenzie‘b 1994; Rowan‘b Foy‘b Rodriguez‘b & Ryan‘b 1994; Wolfe & Wekerle‘b 1994). To date‘b the child sexual abuse research has focused predominantly on the majority Caucasian population with only limited research with minority populations (Arroyo‘b Simpson‘b & Aragon‘b 199‘f). Given the explosion of growth among minority populations in the United States‘b particularly among Latinos‘b there is an increasing need to understand how child sexual abuse affects minority populations. 816 VIOLENCE AGAINST WOMEN‘b‘J Vol. 8 No. ‘f‘b July 200 2816-84 4 © 2002 Sage Publicat‘Jions In this study, the Comprehensive Model of Trauma Impact and Self-In-Relation theoretical orientation are used to provide a con – ceptual framework from which to empirically investigate the impact of child sexual abuse among Latina survivors. The study focusedspecifically on the relationship of affective symptomatology and interpersonal support. COMPREHENSIVE MODEL OF TRAUMA IMPACT The trauma of sexual abuse and its negative impact is not a lin – ear pathway but involves many interactive variables. Childhood sexual abuse does not occur in a vacuum. It happens to a unique individual who lives in a particular context, which includes imme – diate family, extended family, community, culture, and society. To conceptualize the impact of child sexual abuse and to consider the numerous issues involved in assessing child sexual abuse, Koverola and colleagues (Koverola, 1992) developed the Comprehensive Model of Trauma Impact (CMTI) (see Figure 1). This model pro- vides a framework for examining child sexual abuse and how it affects an individual’s areas of functioning, the nature of the trauma, and the particular systemic context in which the abuse occurred. The CMTI identifies a number of areas of functioning that can be affected by child sexual abuse, including cognitive, affective, moral, interpersonal, sexual, and physical. Further- more, the model considers the systemic context in which the indi- vidual exists that is understandably also affected by child sexual abuse, including family, community, and society. The nature of the trauma is another important variable consid- ered by the CMTI. For example, child sexual abuse often co-occurs with other forms of victimization. Survivors may have been con- currently exposed to domestic violence, community violence, psychological abuse, physical abuse, or neglect. This is an impor- tant component of this model because clinicians and researchers have come to realize that many child sexual abuse victims have suffered multiple forms of victimization (Koverola & Friedrich, 2000). Finally, the model recognizes that the effects of the trauma may not occur immediately. Delayed effects, such as anxiety related to sexual intimacy or body image, may not occur until adolescence Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 817 or young adulthood, and parenting issues, due to their own intrafamilial abuse, are not likely to appear until they have their own family. Because individuals are in a constant state of devel- opment, it is important to recognize that events occur on a contin- uum (i.e., pretrauma, trauma, and posttrauma), and the effects of trauma must be examined with this factor in mind. The CMTI is a model and, as such, provides a framework to consider the interplay between the numerous factors believed to be important when conducting research or clinical interventions with child sexual abuse survivors. The model does not provide a theoretical basis to hypothesize the mechanism or process by which these variables are related. For a theoretical perspective to facilitate the understanding of how child sexual abuse has an 818 VIOLENCE AGAINST WOMEN / July 2002 INDIVIDUAL FAMILY COMMUNITY SOCIETY TRAUMA AREAS OF FUNCTIONING PHYSICALCOGNITIVE SEXUALAFFECTIVE INTERPERSONALMORAL PRETRAUMATRAUMAPOSTTRAUMA Figure 1: Comprehensive Model of Trauma Impact SOURCE: Koverola (1992). impact on survivors, we turn to the Self-In-Relation theoretical orientation (Jordan, 1997). SELF-IN-RELATION THEORETICAL ORIENTATION The Self-In-Relation theoretical orientation, developed at the Stone Center for Developmental Service and Studies at Wellesley College, provides a particularly useful theoretical framework for investigating the impact of trauma on areas of an individual’s functioning within a particular context (Surrey, 1991). It is also inherently appropriate for use in research with a minority popula – tion, including Latinas, because it explicitly addresses cross- cultural issues (Jordan, 1997). The Self-In-Relation theoretical orientation conceptualizes iden- tity development as forming in the context of relationships. This orientation emphasizes the centrality of relational development rather than focusing on the development of autonomy and self- sufficiency. Self-differentiation, therefore, takes place in the con- text of relationships without the usual losses associated with per- sonal development. Women continually add to their primary rela- tionships to support their age-appropriate development. People grow because they are in relationships. Positive growth is thus measured by a person’s ability to participate in meaningful rela- tionships that positively affect one another. It is an “ongoing pro- cess of intersubjective relationship” that allows the self to grow in depth and vitality (Surrey, 1991). The Self-In-Relation does not label cultural differences as devi- ant, deficient, or pathological (Surrey, 1991; Turner 1997). It empha- sizes the importance of interconnection, the evolutionary concept of the “relational self,” in women’s development—an important value for Latina culture, especially the relational context of the family (Turner, 1997). Moreover, rather than valuing autonomy and separation, the Self-In-Relation model allows for more tradi- tional values that are dominant in Latina culture. It can serve to validate and legitimize a significant part of the Latina woman’s maturation process, that is, her growth in community and espe- cially within ongoing significant relationships. Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 819 CHILD SEXUAL ABUSE AS AN INTERPERSONAL TRAUMA Child sexual abuse occurs in relationships and causes personal and interpersonal harm. It follows, then, that the affective difficul – ties and the interpersonal nature of child sexual abuse can affect a survivor ’s current and subsequent interpersonal relationships. When child sexual abuse occurs, the sense of self is harmed within an interpersonal context. Therefore, the mechanism to rebuild the self or repair the psychological damage requires being in connec – tion with others (Ducharme, Koverola, & Battle, 1997; Herman, 1992). Healthy adult relationships depend on a secure sense of one’s self in relation to another, the ability to trust another, the capability to regulate one’s emotions, and the judgment to know when one can or cannot rely on another (Cole & Putnam, 1992). Child sexual abuse can impair one’s perception of others. By the time the child sexual abuse survivor reaches adulthood, the cumulative effect of impairments to self and social functioning are likely to influence development and the transitions made into adulthood (Herman, 1992). The documented interpersonal diffi- culties that occur in child sexual abuse survivors include poor social adjustment, difficulty trusting others, social isolation, diffi- culty forming or maintaining relationships, feelings of alienation, isolation, and insecurity, parenting difficulties, physical abuse, and sexual revictimization (Barnett, Miller-Perrin, & Perrin, 1997; Herman, 1992). Studies also indicate that the relationship with the perpetrator affects the long-term adjustment to child sexual abuse (Briere & Runtz, 1988; Browne & Finkelhor, 1986; Huston, Parra, Prihoda, & Foulds, 1995). These researchers found greater impairment in psychological adjustment when the perpetrator was someone in the family, particularly if the perpetrator was the father, a father figure, or someone with whom the individual had a close emo- tional relationship (Barnett et al., 1997; Beitchman et al., 1992; Briere & Runtz, 1988; Browne & Finkelhor, 1986; Huston et al., 1995). In a comparative study that examined the impact of child sexual abuse on survivors who experienced intrafamilial versus extrafamilial sexual abuse, Gregory-Bills and Rhodeback (1995) found that the trauma experienced is likely to be aggravated when the perpetrator was a family member. However, one study 820 VIOLENCE AGAINST WOMEN / July 2002 on depression and child sexual abuse in Latinos reported the woman’s relationship to the perpetrator did not influence her scores on depression when compared to child sexual abuse by a stranger (Roosa, Reinholtz, & Angelini, 1999). Yet, these research – ers cited several methodological problems and stated their results should be interpreted with caution. In sum, these studies not only indicate the prevalence of child sexual abuse within the family, but they also underscore the harmful and long-lasting effects that result from intrafamilial child sexual abuse. FACTORS INFLUENCING THE OUTCOME OF ABUSE An investigation of social support and long-term psychological adjustment after child sexual abuse found that a negative response from others might aggravate the traumatic symptoms (J. Edwards & Alexander, 1992; Herman, 1992). For example, the child sexual abuse victim’s perception of the childhood family environment was found to be a predictor of anxiety and depression in adult women (Yama, Tovey, & Fogas, 1993). Other research indicates that when child sexual abuse occurs, an important mediating variable related to subsequent adult psychological distress is the family environment as perceived by the victim (Alexander, 1992; Nash, Hulsey, Sexton, Harralson, & Lambert, 1993). When child sexual abuse occurs, the family is also affected, whether the abuse occurred within or outside the family. How the family responds can affect the outcome of child sexual abuse. For example, in a study done by Gold (1986), the woman’s perception of her mother ’s response to the child sexual abuse significantly related to her adult functioning. However, it is not only how the family responds but also the response of other sources of support, including teach- ers, other adults, and institutions, that can mitigate or even play a protective role in the outcome of child sexual abuse (Gomes- Schwartz, Horowitz, & Cardarelli, 1990; Herman, 1992). Because the response of others can affect the impact of child sexual abuse, this suggests that social support can serve as a means of helping the survivor cope with and heal from interper- sonal trauma. The influence of the family is also an important fac- tor in the long-term effects of child sexual abuse. Specifically, Mar- tinez (1988) found that Mexican-Americans’ use of the family as a Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 821 supportive network has served as a means to help them tolerate stress and dysfunction better. Finally, the importance of social support in general has been documented to have a positive impact on the treatment and recovery from child sexual abuse (Liese, 1999). CROSS-CULTURAL ISSUES IN CLINICAL RESEARCH Researchers have recognized that there are cultural differences in values and behaviors that affect the outcome of events and do not necessarily point to pathological or deviant problems. For example, because of the differences found in a study comparing child sexual abuse and depression among four ethnic groups, researchers concluded that it is important to examine the effects of child sexual abuse separately for various ethnic groups (Roosa et al., 1999). Sue (1991) pointed out that “in ethnic minority issues, we need to recognize that some standards or criteria assumed to be universal (etics) are actually specific (emics) to mainstream American society” (p. 74). Sue argues that cross-culturally compe- tent researchers need to use an emic-based conceptual framework— namely, a framework designed from the perspective of the partic- ular culture under investigation. This avoids misinterpretations of ethnic minority behaviors. Therefore, unraveling the psycho- logical sequelae of child sexual abuse requires an examination of the values and attitudes that are in effect for the survivor ’s family system, including their particular cultural milieu. LATINA CULTURE DEMOGRAPHICS The termLatinacomprises a very diverse group of people in whom there are distinct differences from others and from other Latinas. Within the Latino population are various groups: the Mexican-origin population constituting 60% (or two thirds); Cen- tral Americans, South Americans, and other Latinos, comprising 22%; Puerto Ricans, 12%; and Cubans, 5%. These various groups 822 VIOLENCE AGAINST WOMEN / July 2002 are concentrated in different parts of the country but are highly urbanized (67%) (Vasquez, 1994). FAMILY AND VALUES For Latinas, family and familial responsibilities are integral aspects of life. “The family” includes concepts of extended family, loyalty/respect for family, and family unity. Cooperation rather than competition and the value of maintaining interpersonal rela – tionships within the family in contrast to the importance of indi – viduality are central components of the Latina family (Comas- Díaz, 1995; Sue & Sue, 1990). Family members, including the extended family, are expected to be a source of emotional support for one another, and familial obligations or responsibility to the family are viewed as a priority above other obligations. There is a high paternal or male dominance with different cultural expecta- tions for men and women within the family. There are varying degrees of acculturation within the family that, particularly among Mexican Americans, usually result in parents who are first- generation U.S. residents being more traditional in their roles, whereas their children adopt a more “modern” approach or become bicultural (Martinez, 1988). Negy and Woods’s (1992) review of the Latino acculturation literature found that some studies indicate that increases in the level of acculturation changes the role of the family for the individual but to varying degrees. Gaines (1997) reported that studies have shown that “familism is more pronounced as a cultural value among Latinas/os than among Anglos,” (p. 43), also suggesting that the family is an important source of social support in Latino culture. Within the Latino family, there are culturally based patterns of childrearing and parental expectations that differ from the dominant Cauca- sian culture in America. These include the expectation of silence among children, the expectation that children should be able to control their emotions, respect for adults within the family, and taboos against discussing sexual issues (Sanders-Phillips, Moisan, Wadlington, Morgan, & English, 1995). This orientation to the family and its values produces the following consequences: (a) the family is valued higher than the individual; (b) depending on the level of acculturation, conflicts in values are likely to occur Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 823 between parents, grandparents, and children; and (c) if marital conflicts occur, divorce is seen as a less acceptable option than working things out for the good of the family (Sue & Sue, 1990). LATINO VALUES MAY AFFECT CHILD SEXUAL ABUSE Latino families, with their high level of interpersonal closeness and involvement, can be a strong source of social support or stress for child sexual abuse survivors. Vasquez (1994) cites the cultural characteristics of support from the family, family obligations, and family as referents (relatives as attitudinal and behavioral models of identity) as the core characteristics of Latino culture. The sup – portive family has been found to provide a buffer, or ameliorative effect, to stress for Mexican Americans (Martinez, 1988). How – ever, it is just as likely that additional stressors, because of the demanding cultural expectations placed on Latinas, may make it more difficult to resolve issues of abuse. For example, Sanders- Phillips et al. (1995) cite a scenario where an abused Latino child may remain silent after the abuse because of cultural constraints or because of placing the family above one’s own needs or wishes. Comas-Díaz (1995) argues that the cultural value of respect (respeto) may actually place children at increased risk for sexual abuse and in the position of maintaining silence. Comas-Díaz points out that the child who is abused by someone deserving of respect will fear breaking the silence due to respect for the abuser and fear of the consequences. In addition, given these cultural values, the Latina’s ability to seek support outside the familial context when she has been sexually abused can be compromised. That is, the cultural value of keeping one’s personal problems within the family can function as a barrier to obtaining social support. In sum, the Latino family’s values may contribute to the impact of child sex- ual abuse either positively or negatively (Comas-Díaz, 1995). MULTIPLE FORMS OF VICTIMIZATION AND CHILD SEXUAL ABUSE There are other issues that are likely to affect the psychological adjustment of Latinas to child sexual abuse. Research indicates that child sexual abuse survivors often suffer multiple forms of victimization due to living in the inner city, in substandard housing, 824 VIOLENCE AGAINST WOMEN / July 2002 and being unaware of resources that can be of help (Koverola, Proulx, Battle, & Hanna, 1996; Nadon, Koverola, & Schluderman, 1998). Examples include exposure to violence and crime, natural disasters, institutional racism, and discrimination (Wyatt, 1990). Stigmatization has been identified as a traumagenic dynamic due to child sexual abuse (Finkelhor & Browne, 1985). Racism can make people feel angry, helpless, or betrayed, and may parallel the dynamics of child sexual abuse, adding to the stigmatization experienced when one is sexually abused (Wyatt, 1990). Minor – ities in general are likely to confront neglect, hostility, and preju – dice, and this heightens the effects of a crisis, such as child sexual abuse (Norris, 1992). Comas-Díaz (1995) points out that among some Latinas, there is evidence of a cultural fatalism in response to sexual abuse:Eso tenia que pasar(“It had to happen”). She argues that this must be examined within the context of colonialism. In response to historical domination and political control, cultural fatalism may have developed as a realistic adaptation to the limi- tations imposed by external realities. The cumulative effects of multiple forms of victimization are often, unfortunately, an all too common part of the Latina’s life experience and must not be overlooked. RATIONALE The long-term negative impact of child sexual abuse on the sur- vivor has already been substantially documented in the literature. However, there is a paucity of research investigating the impact among minorities (Arroyo et al., 1997; Vasquez, 1994). The little research that has been done on Latinas has been done with chil- dren or adolescents, comparing them to members of other ethnici- ties who have experienced childhood sexual abuse (Arellano, Kuhn, & Chavez, 1997; Huston et al., 1995; Kellogg & Hoffman, 1995; Lindholm & Willey, 1986; Mennen, 1994, 1995; Sanders- Phillips et al., 1995). All except one of these studies (Arellano et al., 1997) found racial/ethnic differences between groups. These stud- ies reported elevated symptoms of depression, higher anxiety, differences in type of sexual abuse, and a greater frequency of feel- ing self-blame and ambivalence about their sexual abuse in minor- ity populations. There is a need for further research, with specific consideration given to cultural context. Awareness of the particular Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 825 issues Latina survivors may experience in light of their culture may be helpful in understanding the impact of child sexual abuse. The influence of cultural values, attitudes, and beliefs may have a damaging or, in contrast, an ameliorating effect on the impact of child sexual abuse. Research that does not impose a majority culture ethnocentric bias must, at its core, utilize a theoretical premise that considers cross-cultural issues. Research comparing one cultural group to another, for example, does not address the issues inherent within one specific cultural group. This study, therefore, will examine one specific minority group, Latinas who have experienced child – hood sexual abuse. Furthermore, this research will be guided by a theoretical orientation that places cultural issues at the center of the investigative process. Of particular interest in this study is the premise that the Latino culture is highly relational in nature, that is, strong in its familial and social ties. As such, child sexual abuse is posited to disrupt relational ties within the context of the family, affecting the survivor ’s ability to seek social support. This study proposes that when the child sexual abuse victim suffers from depression, anxiety, or hostility, her perception of others is impaired or damaged, resulting in a heightened negative interpersonal sensitivity. The negative interpersonal sensitivity results in her being unable to accept or benefit from the social sup- port of others. These findings will have important clinical impli- cations. It is only as these clinical symptoms are addressed in ther- apy and only as the survivor is able to be in relationship with others that the survivor can recover adequately from the interper- sonal damage done by sexual abuse in childhood. Using the CMTI’s delineation of critical variables involved in mediating the impact of child sexual abuse and utilizing the self-in-relation theoretical perspective as a guiding premise, this study investigated the affective functioning of a sample of Latinas who have been multiply victimized and are survivors of child sex- ual abuse. The study provides descriptive demographic and psy- chological adjustment information about this significantly under- studied population of minority women. Research has shown the negative psychological sequelae child sexual abuse has had on adults (i.e., depression, anxiety, and hostility). There is a need to understand how this affective damage affects the survivors’ social relationships. Social support is an especially important variable 826 VIOLENCE AGAINST WOMEN / July 2002 among Latinas and has been found to have a positive impact on treatment and recovery from child sexual abuse. Thus, the rela – tionship between affective symptoms and interpersonal sensitiv – ity as well as between these variables and perceived social sup – port was investigated. It was hypothesized that these affective sequelae of child sexual abuse interact with interpersonal sensi – tivity, which will in turn affect perceived social support. HYPOTHESES This study included the following specific hypotheses: 1. The affective variables of depression, anxiety, and hostility will all have a significant and positive correlation with interpersonal sensitivity. 2. The variable of global distress, depression, anxiety, and hostility will have an inverse correlation with social support. 3. Interpersonal sensitivity will have a significant negative correla- tion with social support. METHOD PARTICIPANTS The participants in this study were drawn from a specialty clinic in an inner-city public hospital in Southern California pro- viding services to victims of child abuse, sexual assault, and domestic violence. The participants included 54 clients of Latin descent, aged 18 to 52, who were seeking mental health services and presenting with a history of sexual abuse. Country of origin data were not specifically obtained from participants in the study. However, the Latina population served by the clinic is known to be predominantly Mexican American and Mexican, with a small minority being immigrants from Central American countries. MEASURES All measurement packages were available in English and Span- ish. The published Spanish version of the Symptom Checklist 90-Revised (SCL90-R) was used. All measures, except the SCL90-R were translated into Spanish. A back translation was completed Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 827 for verification. Spanish translation was completed by individuals of Mexican descent residing in Southern California and familiar with the language used within the specific community in which most participants resided. Background Demographic Questionnaire Background demographic information was obtained from an intake questionnaire. The information gathered included age, gender, ethnicity (differentiating only among Caucasian, African American, Latina, Asian, and other), language spoken at home, education, socioeconomic status, history of substance abuse,physi – cal abuse, and child sexual abuse. These data were gathered for descriptive purposes. History of Unwanted Sexual Contact A self-report measure developed by Koverola and colleagues (1996) was used to obtain the history of sexual abuse. Child sexual abuse was defined as any unwanted physical sexual contact that occurred prior to age 18 by someone at least 5 years older than the victim. Peer sexual abuse was defined as any unwanted physical sexual contact that occurred prior to age 18 by someone less than 5 years older. Adult sexual assault is defined as any unwanted physical sexual contact having occurred after the age of 18. For the purposes of this study only child sexual abuse and peer sexual abuse forms were examined because the first incident of sexual abuse would have occurred prior to age 18. The questionnaire documents the following: type of abuse (sexual kissing; fondling of the buttocks, thighs, breasts, and/or genitals; insertion of fin- gers or objects into the vagina or anus; oral sex; attempted or com- pleted vaginal intercourse; and anal intercourse), the age of onset of the abuse, the frequency of abuse, the participant’s relationship to the perpetrator, and threatening behaviors or forms of coercion used. SCL-90-R The Symptom Check List-90-Revised (SCL-90-R) was used to measure the global distress and symptoms of depression, anxiety, hostility, and interpersonal sensitivity (Derogatis & Cleary, 1977). 828 VIOLENCE AGAINST WOMEN / July 2002 The SCL-90-R has been used in previous research to assess the long-term effect of child sexual abuse and to measure current level of distress on child sexual abuse survivors (Greenwald, Leitenberg, Cado, & Tarran, 1990; Murphy et al., 1988; Surrey, Swett, Michaels, & Levin, 1990). The SCL-90-R is designed to assess global distress as well as provide subsections for measur – ing more specific symptomatology, such as depression, somatization, anxiety, and anger (hostility). The participant indicates which symptoms have caused her distress on a scale of 0 to 4. In this study there are five subscales of interest: (a) Global Severity Index (GSI), (b) Depression (DEP), (c) Anxiety (ANX), (d) Hostility (HOS), and (e) Interpersonal Sensitivity (I-S). According to the SCL-90-R, the GSI is an indicator of the current depth or level of the disorder. It combines information about the number of symptoms endorsed with the intensity of perceived distress. It is a single summary measure. The DEP dimension reflects a range of symptoms as manifested in clinical depression. Symptoms of dysphoric mood and affect are common, as are signs of with- drawal of interest, loss of vital energy, and lack of motivation. Other cognitive and somatic symptoms may occur, such as thoughts of suicide and feelings of hopelessness. The ANX dimension reflects general signs of anxiety, such as tension, trembling, rest- lessness, and nervousness. It also includes the symptoms of panic attacks, such as apprehension and feelings of terror and dread. The HOS dimension reflects feelings, thoughts, and actions that are characteristic of the negative affect state of anger. Rage, irrita- bility, aggression, and resentment are common modes of expres- sion that are measured. Finally, the I-S index focuses on feelings of inferiority and inadequacy, especially in comparison with others. Self-doubt, self-deprecation, and marked discomfort or acute self-consciousness during interpersonal encounters, are charac- teristic manifestations of those with a high I-S score. They also express a tendency to have negative expectations regarding inter- personal behavior with others and others’ perceptions of them. Social Relationships Questionnaire Perceived social support was measured by the Social Relation- ships Questionnaire (Abbey, Abramis, & Caplan, 1985). This self- report instrument documents the survivor ’s perception of support Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 829 in relationships. It is divided into three sections with 15 questions in each, with 6 of these questions addressing her social relation – ship with (a) an important individual in her personal life, (b) peo – ple in the participant’s personal life in general, and (c) a spouse or partner. PROCEDURE All clients receiving services at the clinic have an initial intake interview and complete several screening measures. The mea – sures used in this study represent a subset of the data gathered as part of routine intake procedures in the clinic. Data from intakes conducted over a 12-month period were included in this study. Participants were included if they disclosed a history of child sex – ual abuse and indicated that they were Latina. DATA ANALYSIS The data were analyzed using the Statistical Program for Social Sciences (SPSS). RESULTS DEMOGRAPHIC CHARACTERISTICS The sample included 54 Latinas, ranging in age from 18 to 52. The women were predominantly Spanish speaking, 50% being monolingual Spanish, 20.4% bilingual, and 25.9% English speak- ing only. The majority of the women had less than nine grades of schooling, with 9.3% having none to third grade, 16.7% having fourth to sixth grade, 20.4% having seventh to ninth grade, 33.3% having 10th through 12th grade, 11.1% having some college, and 1.9 reporting having a bachelor ’s or above. The sample was of low SES, with 65% having a household income of less than $15,000 per year. The average household included four people. Of the sample, 39% were U.S. natives, 46% reported having resided in the United States more than 5 years, with the remainder (15%) having resided in the United States fewer than 5 years. A total of 39% of the sam- ple never married, 33% were separated, 22% were married or cohabitating and 4% were divorced. 830 VIOLENCE AGAINST WOMEN / July 2002 ABUSE CHARACTERISTICS Age when abuse first occurred ranged from 3 years to 17 years; 15% were abused when they were 5 years old or younger, 39% were abused during ages 6 through 11, and 49% were abused when they were between the ages of 12 and 17. The most common type of perpetrator was someone that the survivor knew, with 69% being familial and consisting of 21.3% parent/stepparent/ guardian, 11.1% a sibling, 36.2% an extended family member, and 51.1% being an unrelated acquaintance. The remainder (25.5%) were abused by a stranger. With respect to the types of manipula – tion used by the perpetrator, 54.5% reported being coerced, 33.3% were threatened, 55.6% were physically forced, and 24.4% were physically hurt. Within the sample, the types of sexual abuse reported included sexual kissing, fondling, insertion of fingers and objects, oral cop- ulation, attempted vaginal intercourse, completed vaginal inter- course, and anal intercourse. Frequencies are reported in Table 1. DEPENDENT MEASURES The mean scores for the distress scales on the SCL-90-R were as follows: I-S, 66.62 (SD= 9.50); DEP, 66.85 (SD= 8.04); ANX, 63.91 (SD= 9.89); HOS, 62.15 (SD= 11.20); and Global Distress, 67.35 (SD= 8.61). The mean score for social support on the Social Relationships Questionnaire for “Important Person” was 3.28 (SD= 1.16), and “People in Personal Life” was 2.96 (SD= 1.06). Reliability analysis of the subscale for social support for “Important Person” was .90 and for “People in Personal Life” .87, using a Cronbach’s alpha. RELATIONSHIP OF AFFECTIVE FUNCTIONING AND INTERPERSONAL SENSITIVITY A Pearson correlation was used to test whether the affective variables of depression, anxiety, and hostility were significantly correlated with interpersonal sensitivity. The relative contribu- tion of these variables was determined using partial correlational analysis. As hypothesized, each of the affective scales on the SCL-90-R was found to correlate significantly with I-S; with DEP, r(51) = .67,p< .0005; with ANX,r(51) = .56,p< .0005; and with Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 831 HOS,r(51) = .53,p< .0005. Post hoc analysis using partial correla- tions and a Bonferroni test with the alpha set at .02 revealed that when controlling for anxiety and hostility, only depression remained significant,sr(3, 49) = .45,p< .001. These findings are reported in Table 2. 832 VIOLENCE AGAINST WOMEN / July 2002 TABLE 1 Sexual Abuse Types and Frequency Type of Sexual Abuse Frequency Sexual kissing Never 19 Once or twice 14 3 to 10 times 6 11 to 20 times 3 More than 20 times 11 Fondling of the buttocks, thighs, breasts, genitals Never 10 Once or twice 19 3 to 10 times 7 11 to 20 times 1 More than 20 times 16 Insertion of fingers or objects into vagina/anus Never 12 Once or twice 18 3 to 10 times 7 11 to 20 times 1 More than 20 times 15 Oral sex Never 13 Once or twice 17 3 to 10 times 7 11 to 20 times 1 More than 20 times 15 Attempted vaginal intercourse Never 31 Once or twice 12 3 to 10 times 4 11 to 20 times 2 More than 20 times 3 Completed vaginal intercourse Never 40 Once or twice 6 3 to 10 times 2 11 to 20 times 2 More than 20 times 2 RELATIONSHIPS OF GLOBAL DISTRESS AND PERCEIVED SOCIAL SUPPORT It was predicted that global distress would be inversely corre- lated with perceived social support. In particular, it was hypothe- sized that depression, anxiety, and hostility would correlate with perceived social support. A zero-order correlation revealed that Global Distress was significantly correlated with perceived social support in the negative direction when the client is considering “People in Personal Life,”sr(47) = –.34,p< .05. Global Distress was not significantly correlated with “Important Person.” Zero-order correlations found depression was significant for both aspects of social support; the finding for “Important Person” wassr(47) = –34,p< .018, and for “People in Personal Life”,sr(47) = –38,p< .007, whereas anxiety and hostility were not significant. The alpha level for each Bonferroni test was adjusted to .02. These findings are reported in Table 3. RELATIONSHIP OF INTERPERSONAL SENSITIVITY AND PERCEIVED SOCIAL SUPPORT A Pearson correlation was computed to determine whether interpersonal sensitivity was negatively correlated with perceived social support. It was found that interpersonal sensitivity was inversely correlated with perceived social support only when the client was considering an “Important Person,”r(48) = –.30,p< .05, and not significant with “People in Personal Life.” These findings are reported in Table 4. In addition, post hoc analysis using a Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 833 TABLE 2 Correlation Matrix: Intercorrelations Between Interpersonal Sensitivity and Affective Variables in Child Sexual Abuse Survivors Variable IS Dep Anx Hos IS — .67*** .56*** .53*** Dep — .72*** .54*** Anx — .64*** Hos— NOTE: IS = interpersonal sensitivity; Dep = depression; Anx = anxiety; Hos = hostility. n= 51. ***p< .0005. protected-F test revealed that both forms ofsocial support were highly correlated with depression; “Important Person” wassr(48) = –.30,p< .05, and “People in Personal Life” wassr(48) = –.33,p< .05. Although “Important Person” and “People inPersonal Life” were correlated with each other,sr(48) = .38,p< .01, when depres- sion was controlled for, neither “Important Person” nor “People in Personal Life” was found to be significantly correlated. DISCUSSION It was predicted that in child sexual abuse survivors, the symp- toms of depression, anxiety, and hostility would be significantly correlated with interpersonal sensitivity. That is, as depression, anxiety, or hostility increased, so would interpersonal sensitivity. As predicted, when depression, anxiety, or hostility was reported as part of the client’s affective symptomatology, interpersonal sensitivity also was a significant symptom. These findings corrob- orate results in the literature that find these affective symptoms to often be the sequelae of child sexual abuse (Briere & Runtz, 1988; Browne & Finkelhor, 1986; Murphy et al., 1988; Surrey et al., 1990). Further examination of the data using post hoc analysis revealed that when each of the affective variables was held constant, depression was the only variable found to be significantly corre- lated with interpersonal sensitivity. This finding provides stron- ger support for previous studies that found that depression is one 834 VIOLENCE AGAINST WOMEN / July 2002 TABLE 3 Correlation Matrix: Intercorrelations Between Global Distress, Affective Variables, and Social Support in Child Sexual Abuse Survivors Variable SSIP SSPPL Global Dep Anx Hos SSIP — .37** –.25 –.34* –.18 –.26 SSPPL — –.34* –.38** –.27 –.19 Global———— Dep — — — Anx — — Hos— NOTE: SSIP = social support by an important person; SSPPL = social support by people in personal life; Dep = depression; Anx = anxiety; Hos = hostility.n= 51. *p< .05. **p< .01. of the most salient long-term negative sequelae of child sexual abuse (Beitchman et al., 1992; Greenwald et al., 1990; Jumper, 1995; Roesler & McKenzie, 1994). Of particular interest to this study is that this result is also consistent with the findings of a study that examined child sexual abuse and depression in four ethnic groups. The results found depression to be a significant symptom among the Mexican American women (n= 472) who reported a history of child sexual abuse (Roosa et al., 1999). There- fore, although interpersonal sensitivity has components of anxi- ety and hostility, it is depression alone that significantly accounts for the interpersonal sensitivity of child sexual abuse survivors. These findings highlight that depression needs to be closely assessed in child sexual abuse survivors, even when they may present with other distress symptomatology. Clinical assessment of child sexual abuse survivors should include a careful differenti- ation of the broad range of distress symptomatology. It was hypothesized that depression, anxiety, and hostility would correlate with perceived social support. This part of the study extends previous research that found global distress and the affective symptoms of depression, anxiety, or hostility to be significant consequences of child sexual abuse (Briere & Runtz, 1988; Browne & Finkelhor, 1986; Greenwald et al., 1990; Murphy et al., 1988; Surrey et al., 1990). Initial analysis revealed that global distress was significantly correlated with perceived social sup- port, but only when participants were considering an important person in their lives (such as a spouse, relative, or close friend) rather than people they viewed as important, such as family or any group. However, when depression was controlled for, there Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 835 TABLE 4 Correlation Matrix: Intercorrelations Between Interpersonal Sensitivity and Social Support in Child Sexual Abuse Survivors Variable IS SSIP SSPPL IS — –.30* –.13 SSIP — .38** SSPPL — NOTE: IS = interpersonal sensitivity; SSIP = social support by an important person; SSPPL = social support by people in personal life.n= 51. *p< .05. **p< .01. was not a significant correlation between global distress and any kind of perceived social support (i.e., neither important person nor important people was significant). Results suggest that depres – sion is the main factor that must be considered in relationship to perceived social support. Social support consists of the positive interpersonal relation – ships a person has with several people. Having healthy relation – ships is an important part of self-formation and personal growth. Healthy relationships, by definition, require give and take from all parties involved. Therefore, the ability to give as well as the ability to receive are necessary parts of the process of being in a healthy relationship. When one is depressed, it understandably becomes more difficult not only to initiate the giving process, but also to receive from others. It has been well documented in the depression literature that when one is depressed, one often feels unworthy, has low self-esteem, and may feel misunderstood or alienated from others. In addition, Latinas have cultural issues that may affect how they manage their depression, such as feel- ings of shame or fear of criticism that may occur for being depressed. These feelings, together with cognitive distortions and the low energy that often accompanies depression, can make sustaining relationships difficult. How the client perceives social support has important clinical implications because it has the potential of negatively or posi- tively affecting treatment and help-seeking behavior. Previous research has shown that social support can influence the impact of child sexual abuse (Liese, 1999). That is, positive social support can reduce the negative symptomatology due to child sexual abuse. The current study extends the findings of the literature and assesses factors that may interfere with a client’s seeking or accepting social support. Because depression was determined to be a significant factor negatively associated with social support, the need to initially assess for depression before proceeding with treatment planning appears to be critical. The issues surrounding depression need to be addressed before the client will be ready to seek or fully benefit from social support. In addition, one source of social support, namely the family, has been found to help child sexual abuse survivors tolerate stress better (Gomes-Schwartz et al., 1990; Herman, 1992). Familialsupport 836 VIOLENCE AGAINST WOMEN / July 2002 has also been identified as an important mediating factor with research investigating Mexican Americans (Martinez, 1988). Therefore, family therapy may also be a beneficial component of treatment recommendations when working with Latinas. Family therapy can facilitate the client’s reconnection with significant family members and provide support. Thus, these findings sug – gest that a part of the treatment goal would be to directly address the need to develop significant relationships despite the client’s negative feelings and perceptions about these relationships. An empathic and supportive relationship in therapy along with devel – oping decision-making skills may help the client ascertain who is trustworthy within the family and who is not, and thus empower and remind them of the choices they still have to make changes in their lives. It was predicted that among child sexual abuse survivors, there is an inverse relationship between interpersonal sensitivity and perceived social support. The results of the study extend previous research that found interpersonal sensitivity to be significantly related to the outcome of child sexual abuse (Greenwald et al., 1990; Murphy et al., 1988; Surrey et al., 1990). Results indicated that there was a statistically significant inverse relationship between interpersonal sensitivity and perceived social support when considering an important person but not for important peo- ple in their lives. However, once again, when depression was held constant, neither variable for perceived social support (important person nor important people) was significantly related to inter- personal sensitivity. This suggests that only depression was a fac- tor affecting perception of social support for these participants. The implication of this finding underscores the importance of assessing for depression and, if present, determining how the individual’s perception of others may have been affected by their depression. An item analysis of the interpersonal sensitivity subscale on the SCL-90-R revealed that the primary constructs on this subscale are the cognitive components or inaccurate perceptions that occur in a relational context, which are often related to depression. For example, some of the questions include the fol- lowing: “feeling critical of others,” “feeling others do not under- stand you or areunsympathetic,” “feeling that people are unfriendly or dislike you,” “feeling uneasy when people are watching you or Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 837 talking about you,” and “feeling very self-conscious with others.” Interpersonally sensitive individuals feel inadequate and inferior when they compare themselves with others. Similarly, depressed individuals feel self-conscious and inadequate around others. Issues of trust and difficulty with intimacy are likely negative con – sequences of child sexual abuse that may have an impact on rela – tionships. Therefore, these findings highlight the importance of directly working with clients on their capacity for healthy social relationships. Focusing on these issues in therapy will help the cli – ent to accept and benefit from the support of others, which is a crucial component in healing from child sexual abuse. In sum, the results of this study revealed that depression is a significant predictor of interpersonal sensitivity. Further, it is depression, not interpersonal sensitivity, which also predicts per – ceived social support. CULTURAL IMPLICATIONS Although these results have implications for non-Latinas, there are a number of considerations specific to the Latino culture that bear mentioning. Fontes (1993) has argued strongly that there are many pressures within a Latino child’s family that will hinder dis- closure of abuse. She notes that the cultural dynamics within the Latino family may often intensify the struggle. A child may remain silent because he or she has been taught from an early age that the needs of the family supersede the needs of the individual— the family always comes first (Falicov, 1982; Gaines, 1997). In a male-dominated culture and one in which “children are to be seen and not heard,” it seems likely that if child sexual abuse occurs and the perpetrator is a male inside the family, it would be exceed- ingly difficult for the child to report it to someone. For example, if a girl told her mother that her father molested her, it would create a crisis situation that might be difficult for the mother to manage. In many cases the father is the sole provider, the head of the fam- ily, and the one who makes the rules (Sue & Sue, 1990). Speaking out may mean losing provision for the family, which can be exceedingly difficult, especially if there are other children to con- sider. These factors need to be considered when treating an adult Latina child sexual abuse survivor. For example, because she may have felt the need to keep her child sexual abuse a secret, the 838 VIOLENCE AGAINST WOMEN / July 2002 survivor may be revealing it to the therapist for the very first time. The cultural value of keeping one’s problems within the family may also make it difficult for the Latina to speak about her child sexual abuse (Falicov, 1982; Sanders-Phillips et al., 1995). In addition, there are taboos within the family regarding talking about sexual issues (Espin, 1986). Furthermore, the Latino child must make sense of the pressure to “respect” the person who may have abused him or her by keeping silent (Comas-Díaz, 1995). The ther – apist needs to find a sensitive way to give the Latina adult survi – vor safety and permission to speak about the unspeakable thing that was done to her. Given Latinas’ strong family ethic, if child sexual abuse occurs within the family and it is discovered, even years later, the shame and betrayal by the family in not keeping her safe are likely issues that need to be addressed. The rippling effect, even if it occurs with a stranger or acquaintance, can cause devastating harm within the family. That is, in a close-knit family, such a harmful and shameful event is likely to cause a crisis within the family. Such a trauma may bring the family in for therapy to address issues concerning the sexually abused child. The therapist needs to have respect for their culturally specific family system or they could lose the opportunity to help these clients (Falicov, 1982). As mentioned, consideration for the family’s and individual’s cul- tural background is the only effective and ethical way to do cross-cultural therapy. The literature has shown that survivors are likely to be multiply victimized because of many variables such as poverty, violence in the neighborhood, living in substandard housing, and being unaware of resources (Koverola et al., 1996). Wyatt (1990) adds that minor- ity children are likely to experience multiple forms of victimiza- tion that include, in addition to the aforementioned factors, insti- tutional racism. The participants in this study similarly were from a low SES, living in substandard housing, in violent neighbor- hoods, and reporting violence in their homes. These additional stressors combined with child sexual abuse create a special set of circumstances that need to be assessed carefully and treated accordingly and responsibly. A therapist must be careful not to revictimize the survivor by failing to evaluate the impact of these additional stressors. Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 839 DIRECTIONS FOR FUTURE RESEARCH AND LIMITATIONS One of the limitations of this study is the generalizability of the results. The literature clearly establishes the heterogeneity evident among ethnic minority populations and, thus, these findings may not be generalizable to other Latino groups nor to other ethnic groups. The population included for this study was Latinas who reported a history of child sexual abuse. It is, however, important to consider the issue of acculturation. Our sample included pre – dominantly Latinas who were U.S. born or who had lived in the United States for more than 5 years. It is interesting to note that the majority were Spanish-speaking or bilingual, continuing to embrace their mother tongue. If language is viewed as a proxy for accultur – ation, this would suggest that the majority of our participants were not highly acculturated. The findings of our study, therefore, may be most applicable to Latinas who, although American born, or who have lived in the United States for a significant period of time, nonetheless identify themselves as Latina and may prefer their native tongue, customs and cultural values. Although anec- dotal, it bears noting that the vast majority of Latinas served in this particular clinic were of Mexican or Mexican American back- ground, so the findings may apply more directly to this subgroup within the larger group of Latinas. A second limitation of this study is the fact that most of the par- ticipants had a history of multiple victimization. The focus of the discussion has been on the sequalae of child sexual abuse, and it is not possible to tease out the differential effects of the different types of victimization experienced. Another limitation is the small sample size. A larger sample may be able to provide stronger rela- tionships between the variables and the possibility of examining issues such as the unique contributions of different types of vic- timization experiences. Future research that would replicate the findings of this study with a larger sample size could prove beneficial. The design of this study as retrospective and self-report has inherent limitations. The limitations of self-reporting have been widely documented and critiqued. The findings have been inter- preted as the participants’ perceptions of their reality. In view of this being a Latina population, pressures or loyalty to the family may have led to a bias in underreporting severity and minimization 840 VIOLENCE AGAINST WOMEN / July 2002 of symptoms. However, in spite of these limitations, the robust – ness of the findings give a relatively high degree of confidence as to the voracity of the findings. The findings suggest a strong likeli – hood that depression is a major factor in child sexual abuse among Latinas and that is related to perceived social support. Further studies are needed that focus on the specific issues sur – rounding the psychological sequelae of child sexual abuse, partic – ularly among those from a minority culture who are multiply vic – timized. There are many complex issues and circumstances involved that affect our understanding of child sexual abuse and the multi – ple interactions and powerful effects that result from different types of trauma that may occur. Given the complexity and multi – plicity of issues for Latinas, future research endeavors need to continue to focus on such underresearched populations in our society. REFERENCES Abbey, A., Abramis, D., & Caplan, R. (1985). Effects of different sources of social support and social conflict on emotional well-being.Basic and Applied Social Psychology,6, 111-129. Alexander, P. C. (1992). Application of attachment theory to the study of sexual abuse.Jour- nal of Consulting and Clinical Psychology,60, 185-195. Arellano, C. M., Kuhn, J. A., & Chavez, E. L. (1997). Psychosocial correlates of sexual assault among Mexican American and White non-Hispanic adolescent females.His- panic Journal of Behavioral Sciences,19, 446-460. Arroyo, J. A., Simpson, T. L., & Aragon, A. S. (1997). Childhood sexual abuse among His- panic and non-Hispanic White college women.Hispanic Journal of Behavioral Sciences, 19, 57-68. Barnett, O. W., Miller-Perrin, C. L., & Perrin, R. D. (Eds.). (1997).Family violence across the lifespan: An introduction. Thousand Oaks, CA: Sage. Beitchman, J. H., Zucker, K. J., Hood, J. E., DaCosta, G. A., Akman, D., & Cassavia, E. (1992). A review of the long-term effects of child sexual abuse.Child Abuse & Neglect,16, 101-118. Briere, J., & Runtz, M. (1988). Symptomatology associated with childhood sexual victim- ization in a nonclinical adult sample.Child Abuse & Neglect,12, 51-59. Browne, A., & Finkelhor, D. (1986). Impact of child sexual abuse: A review of the research. Psychological Bulletin,99, 66-77. Bryer, J. B., Nelson, B. A., Miller, J. B., & Krol, P. A. (1987). Childhood sexual and physical abuse as factors in adult psychiatric illness.American Journal of Psychiatry,144, 1426-1430. Cole, P. M., & Putnam, F. W. (1992). Effect of incest on self and social functioning: A devel- opmental psychopathology perspective.Journal of Consulting and Clinical Psychology, 60, 174-184. Comas-Díaz, L. (1995). Puerto Ricans and sexual child abuse. In L. Fontes (Ed.),Sexual abuse in nine North American cultures(pp. 31-66). Thousand Oaks, CA: Sage. Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 841 Derogatis, L., & Cleary, P. (1977). Confirmation of the dimensional structure of the SCL-90: A study in construct validity.Journal of Clinical Psychology,33, 981-989. Ducharme, J., Koverola, C., & Battle, P. (1997). Intimacy development: The effect of abuse and gender.Journal of Interpersonal Violence,27, 345-352. Edwards, J. J., & Alexander, P. C. (1992). The contribution of family background to the long-term adjustment of women sexually abused as children.Journal of Interpersonal Violence,7, 306-320. Espin, O. M. (1986). Cultural and historical influences on sexuality in Hispanic/Latin women. In J. Cole (Ed.),All American women: Lines that divide, ties that bind. New York: Free Press. Falicov, C. J. (1982). Mexican families. In M. McGoldrick, J. K. Pearce, & J. Giordano (Eds.), Ethnicity and family therapy(pp. 134-163). New York: Guilford. Finkelhor, D., & Browne, A. (1985). The traumatic impact of child sexual abuse: A concep – tualization.American Journal of Orthopsychiatry,55, 530-541. Fontes, L. A. (1993). Disclosures of sexual abuse by Puerto Rican children: Oppression and cultural barriers.Journal of Child Sexual Abuse,2, 21-35. Gaines, S. O.(1997).Culture, ethnicity, and personal relationship processes. New York: Routledge. Gold, E. R. (1986). Long-term effects of sexual victimization in childhood: An attributional approach.Journal of Consulting and Clinical Psychology,54, 471-475. Gomes-Schwartz, B., Horowitz, J., & Cardarelli, A. (1990).Child sexual abuse: The initial effects. Newbury Park, CA: Sage. Greenwald, E., Leitenberg, H., Cado, S., & Tarran, M. J. (1990). Childhood sexual abuse: Long-term effects on psychological and sexual functioning in a nonclinical and nonstudent sample of adult women.Child Abuse & Neglect,14, 503-515. Gregory-Bills, T., & Rhodeback, M. (1995). Comparative psychopathology of women who experienced intra-familial versus extra-familial sexual abuse.Child Abuse & Neglect,19, 177-189. Herman, J. (1992).Trauma and recovery. New York: HarperCollins. Huston, R. L., Parra, J. M., Prihoda, T. J., & Foulds, D. M. (1995). Characteristics of child- hood sexual abuse in a predominantly Mexican-American population.Child Abuse & Neglect,19, 165-176. Jordan, J. V., Kaplan, A. G., Miller, J. B., Stiver, I. P., & Surrey, J. L. (1991).Women’s growth in connection: Writings from the Stone Center. New York: Guilford. Jordan, J. V. (Ed.). (1997).Women’s growth in diversity: More writings from the Stone Center. New York: Guilford. Jumper, S. A. (1995). A meta-analysis of the relationship of child sexual abuse to adult psy- chological adjustment.Child Abuse & Neglect,19, 715-728. Kellogg, K. D., & Hoffman, T. J. (1995). Unwanted and illegal sexual experiences in child- hood and adolescence.Child Abuse & Neglect,19, 1457-1468. Koverola, C. (1992). The psychological effects of child sexual abuse. In A. H. Heger & S. J. Emans (Eds.),Evaluation of the sexually abused child(pp. 15-29). New York: Oxford Uni- versity Press. Koverola, C., & Friedrich, W. N. (2000). Psychological evaluation of the sexually abused child. In A. H. Heger & S. J. Emans (Eds.),Evaluation of the sexually abused child(2nd ed.). New York: Oxford University Press. Koverola, C., Proulx, J., Battle, P., & Hanna, C. (1996). Family functioning as predictors of distress in revictimized sexual abuse survivors.Journal of Interpersonal Violence,11, 263-280. Landecker, H. (1992). The role of childhood sexual trauma in the etiology of borderline per- sonality disorder: Considerations for diagnosis and treatment.Psychotherapy,29, 234-242. Liese, F. C. (1999).The impact of social support on adult survivors of child sexual abuse. Unpub- lished master ’s thesis, California State University, Los Angeles, CA. 842 VIOLENCE AGAINST WOMEN / July 2002 Lindholm, K. J., & Willey, R. (1986). Ethnic differences in child abuse and sexual abuse.His – panic Journal of Behavioral Sciences,8, 111-125. Martinez, C., Jr. (1988). Mexican-Americans. In L. Comas-Díaz & E.E.H. Griffith (Eds.), Clinical guidelines in cross-cultural mental health(pp. 182-203). New York: John Wiley. Mennen, F. E. (1994). Sexual abuse in Latina girls: Their functioning and a comparison with White and African American girls.Hispanic Journal of Behavioral Sciences,16, 475-486. Mennen, F. E. (1995). The relationship of race/ethnicity to symptoms in childhood sexual abuse.Child Abuse & Neglect,19, 115-124. Murphy, S. M., Kilpatrick, D. G., Amick-McMullan, A., Veronen, L. J., Paduhovich, J., Best, C. L., et al. (1988). Current psychological functioning of child sexual assault survivors: A community study.Journal of Interpersonal Violence,3(1), 55-79. Nadon, S., Koverola, C., & Schluderman, E. (1998). Childhood victimization: Antecedents to prostitution.Journal of Interpersonal Violence,28, 213-220. Nash, M. R., Hulsey, T. L., Sexton, M. C., Harralson, T. L., & Lambert, W. (1993). Long-term sequelae of childhood sexual abuse: Perceived family environment, psychopathology, and dissociation.Journal of Consulting and Clinical Psychology,61, 276-283. Negy, C., & Woods, D. J. (1992). The importance of acculturation in understanding research with Hispanic-Americans.Hispanic Journal of Behavioral Sciences,14, 224-247. Norris, F. H. (1992). Epidemiology of trauma: Frequency and impact of different poten – tially traumatic events on different demographic groups.Journal of Consulting and Clini – cal Psychology,60, 409-418. Roesler, T. A., & McKenzie, N. (1994). Effects of childhood trauma on psychological func- tioning in adults sexually abused as children.Journal of Nervous and Mental Disease,182, 145-150. Roosa, M. W., Reinholz, C., & Angelini, P. J. (1999). The relation of child sexual abuse and depression in young women: Comparisons across four ethnic groups.Journal of Abnor- mal Child Psychology,27, 65-76. Rowan, A. B., Foy, D. W., Rodriguez, N., & Ryan, S. (1994). Posttraumatic stress disorder in a clinical sample of adults sexually abused as children.Child Abuse & Neglect,18, 51-61. Sanders-Phillips, K., Moisan, P. A., Wadlington, S., Morgan, S., & English, K. (1995). Ethnic differences in psychological functioning among Black and Latino sexually abused girls. Child Abuse & Neglect,19, 691-706. Sue, D. W., & Sue, D. (1990).Counseling the culturally different: Theory and practice(3rd ed.). New York: John Wiley. Sue, S. (1991). Ethnicity and culture in psychological research and practice. In J. D. Goodchilds (Ed.),Psychological perspectives on human diversity in America(pp. 47-85). Washington, DC: American Psychological Association. Surrey, J. L. (1991). Self-in-relation: A theory of women’s development. In J. V. Jordan (Ed.), Women’s growth in connection: Writings from the Stone Center(pp. 51-66). New York: Guilford. Surrey, J., Swett, C., Michaels, A., & Levin, S. (1990). Reported history of physical and sex- ual abuse and severity of symptomology in women psychiatric outpatients.American Journal of Orthopsychiatry,60(3), 412-417. Turner, C. W. (1997). Clinical applications of the Stone Center theoretical approach to minority women. In J. V. Jordan (Ed.),Women’s growth in diversity: More writings from the Stone Center. New York: Guilford. Vasquez, M.J.T. (1994). Latinas. In L. Comas-Díaz & B. Greene (Eds.),Women of color: Inte- grating ethnic and gender identities in psychotherapy(pp. 114-138). New York: Guilford. Wolfe, D. A., Sas, L., & Wekerle, C. (1994). Factors associated with the development of posttraumatic stress disorder among child victims of sexual abuse.Child Abuse & Neglect,18, 37-50.Hinson et al. / LATINA CHILDHOOD SEXUAL ABUSE 843 Wyatt, G. E. (1990). Sexual abuse of ethnic minority children: Identifying dimensions of victimization.Professional Psychology: Research and Practice,21, 338-343. Yama, M. F., Tovey, S. L., & Fogas, B. S. (1993). Childhood family environment and sexual abuse as predictors of anxiety and depression in adult women.Journal of Orthopsychiatry, 63, 136-141. Josie Vega Hinson recently received her Doctor of Psychology degree from Fuller Theological Seminary’s clinical psychology program. She is currently seeking licensure as a psychologist in New Jersey. Her clinical practice consists of her training experiences, which she tailored toward an emphasis in trauma, particu – larly working with Latino women and children who are sexual abuse survivors as well as their nonoffending parents. She also has a strong interest in psychological testing for Latinos. In December 2001, Dr. Vega Hinson joined the Child Protection Center at Saint Peter’s University Hospital in New Brunswick, New Jersey, where she conducts psychological assessments with children and adolescents who have been physically or sexually abused, neglected, or have witnessed domestic violence. She also works with their parents or caregivers of whom many are Latinos. Catherine Koverola is a researcher and clinician with expertise in the areas of child maltreatment, sexual abuse, domestic violence and cross-cultural service delivery. Her clinical practice also includes the intersection of forensic psychology within the domain of child maltreatment and domestic violence. The emphasis of her research has been on the psychological impact of trauma exposure. Dr. Koverola’s clinical and research experience includes more than 10 years of working with indigenous peoples in isolated communities throughout Canada and Alaska. She also has extensive experience in the inner city with African American families as well as new immigrant and refugee families from Mexico and Central America who have experienced multiple forms of violence and trauma. Dr. Koverola is cur- rently on the faculty at the Center for Families, University of Maryland as Direc- tor of Research. In this capacity, her role is to develop and expand ongoing research efforts that build knowledge concerning the prevention, assessment, and treat- ment of child abuse and neglect. Mary Morahan, M.S.W., L.C.S.W., is director of clinical training at the Violence Intervention Program at the Los Angeles County and University of Southern Cal- ifornia Medical Center in Los Angeles, California. She has more than 20 years experience as a clinical social worker and therapist, with expertise in the areas of domestic violence and child maltreatment, and with an emphasis on services to underserved and marginalized populations. Her clinical work has spanned the range of child welfare, school-based prevention, health care, and outpatient mental health services. As an educator, she has taught social work students at both the undergraduate and graduate levels. She serves on the Executive Board of the Los Angeles County Domestic Violence Council as chair of the Health Issues Committee. 844 VIOLENCE AGAINST WOMEN / July 2002
AffiliaSummer 2001Hyman, Williams Resilience Among Women Survivors of Child Sexual Abuse Batya Hyman and Linda Williams Research has paid little attention to survivors of childhood sexual abuse who do not exhibit long-term negative consequences or who manifest resilience. This study investigated resilient outcomes, that is, competence in the face of adversity, and the factors associated with these outcomes in 136 women who were sexually abused as children. This article presents a study of women who appeared to func- tion competently despite their experience of child sexual abuse (CSA). The article provides a review of the literature on the meaning of resilience–competence and well-being, the factors that may contribute to this outcome, the range of long-term consequences of CSA, and models for analyzing the predictors of competent functioning. REVIEW OF THE LITERATURE Definitions Scholars have long debated the meaning of the termresilience. Some have defined it as a trait a child exhibits, like hardiness Authors’ Note: The research reported in this article was supported by Grant No. CA-90-1406 from the National Center on Child Abuse and Neglect; Linda Williams, principal investigator. AFFILIA, Vol. 16 No. 2, Summer 2001 198-219 © 2001 Sage Publications, Inc. 198 (Anthony, 1987; Murphy & Moriarty, 1976; Rhodewald & Zone, 1989) or invincibility (Neiman, 1988; Werner & Smith, 1982) following exposure to a known risk factor, such as living with a mentally ill mother. Others have acknowledged that resil – ience comprises many factors in interaction (Hauser, Vieyra, Jacobson, & Wertlieb, 1985; James, O’Toole, & Liem, 1993; Neiman, 1988). In their discussion of resilience in maltreated children, Mrazek and Mrazek (1987) proposed 12 characteristics and skills that may foster resilience and warrant research. For example, they suggested that resilient children are intelligent, vigilant, and able to inhibit their affect—skills that enable them to avoid or escape potentially explosive or otherwise danger – ous situations. Furthermore, some of these children can dissoci- ate from intense feelings and not dwell on the abusive events because they are too painful. Resilient children also acquire as much information as possible about the triggers that stimulate abusive behavior in their abusers and may demonstrate altruis- tic behavior; that is, they often try to protect their younger sib- lings from abusive experiences and sometimes help their schoolmates. Mrazek and Mrazek (1987) discussed the need for abused but resilient children to believe they are worthy enough to be loved and noted that children who are loved by younger sib- lings or who are treated sensitively by teachers and social workers may gain this conviction. They also addressed some of the life circumstances that may foster resilience, such as receiv- ing attention from a mentor or another significant person in their lives who encourages them and helps them view the future with hope that it will be better than the present or past abusive circumstances. To date, little research (Romans, Mar- tin, Anderson, O’Shea, & Mullen, 1995) has tested Mrazek and Mrazek’s hypotheses. Garmezy (1985, 1993) demonstrated that important aspects of the person, the environment, the stressful event, and avail- able supports combine to determine positive and adverse out- comes. In addition, Rutter (1990) identified four main protec- tive processes in children who were removed from their Hyman, Williams199 families, placed in institutions, and followed up as young adults: the reduction of the impact of risk, the reduction of neg – ative chain reactions, the establishment and maintenance of self-esteem and self-efficacy, and the opening up of opportuni – ties. In Rutter’s study, the impact appeared to be reduced by altering the meaning of the adversity for the child and eliminat – ing the child’s exposure to the risk factor, such as maternal depression. On the basis of his hypothesis that negative chain reactions heighten the likelihood of adverse long-term conse – quences, Rutter proposed that two types of experiences are especially important in stemming these chain reactions and developing self-esteem and self-efficacy: secure attachment to loving figures and the accomplishment of tasks, such as sports, schoolwork, and the attainment of positions of authority, throughout childhood and adulthood. With regard to the final mechanism, opening up opportunities or turning points in peo- ple’s lives may either positively affect resilience (such as com- pleting high school and entering college or the workplace) or have an adverse affect on it (for example, becoming a teenage mother, which may foreclose some options, including the pros- pect of a successful marriage and the completion of high school). Women’s Well-Being Resilience has been treated as a psychological construct, yet, researchers in several fields (such as family studies, sociology, and economics) have demonstrated that individuals’ well- being consists of multiple and highly correlated features, includ- ing functioning in the workplace, psychological well-being, family relationships, and health, and that these features are highly correlated (Barnett & Marshall, 1991; Hyman, 2000; Kessler & McLeod, 1985; Verbrugge, 1983; Voydanoff & Don- nelly, 1989). Life satisfaction and well-being are dependent, to some degree, on people’s belief that their resources are ade- quate for their needs (Antonovsky, 1987; Benham & Benham, 1981). Access to physical and mental health resources is often determined by income and occupational status (Anson & 200AffiliaSummer 2001 Anson, 1987). Furthermore, the ability to work is often affected by physical and/or mental health limitations (Barnett & Mar – shall, 1991; Chirikos & Nestel, 1985; Waldron, 1980), and women who work outside the home may remain healthy and derive satisfaction from increased social contacts and opportu – nities for success (Verbrugge, 1987). These effects may differ by ethnicity and social class. Ill health and/or economic strains are often identified as factors in the disruption of interpersonal relationships (Barnett & Baruch, 1987; Larson, Wilson, & Beley, 1994; Voydanoff & Donnelly, 1989). Studies have examined the extent to which the relationships among these outcomes are contingent on coping resources and social supports (Baruch, Biener, & Barnett, 1987; Kessler & McLeod, 1985; Long & Kahn, 1993; Pearlin, 1989; Robertson, Elder, Skinner, & Conger, 1991). Investigators of the associa- tions among stress and the health, well-being, and efficiency of female workers have delineated a model for portraying these stress–health interactions (Berkman, 1986; Frankenhaeuser, 1991; Waldron & Jacobs, 1989). This theoretical and empirical model identifies the stressors (or behavioral demands), the individual and social resources available (skills, experience, genetic factors, and social supports), the individual’s cognitive assessment and bodily reactions to the stress, and the outcomes of health, well-being, and efficiency. It is similar to Garmezy’s (1983, 1985) model for understanding the predictors of resilience. Functioning of Adult Survivors of CSA In the past 20 years, much research has documented the ad- verse long-term psychological sequelae of CSA, the mechanisms by which they occur, and methods of intervention (Briere & Runtz, 1988, 1993; Mullen, Martin, Anderson, Romans, & Herbison, 1993; Peters, 1988; Stein, Golding, Siegel, Burnam, & Sorenson, 1988). However, little attention has been paid to the survivors of CSA who do not exhibit negative consequences or who manifest resilience (Draucker, 1995; Harvey, 1996; James et al., 1993; Neiman, 1988; Romans et al., 1995). Studies of the Hyman, Williams201 short-term impact of CSA have found that 21% to 49% of the children studied suffered no symptoms (Kendall-Tackett, Wil – liams, & Finkelhor, 1993; McLeer, Deblinger, Henry, & Orvaschel, 1992). Less information on resilience in adult CSA survivors is available, although Russell (1986) reported that 22% of her representative sample had no long-term effects. The lack of data on the one fifth of survivors who appear to be resil – ient may have implications for both policy makers and practitioners. METHOD Sample The sample for this study was first interviewed from 1973 to 1975 regarding their CSA experiences. All reported victims of sexual abuse in a major northeastern city were brought to the city hospital emergency department for treatment and the col- lection of forensic evidence, at which time details of the sexual assault were recorded. In 1990 and 1991, 136 of the original 206 girls, now ages 18 to 31, were located and reinterviewed (Wil- liams, 1995). The majority of the girls (86%) were African Amer- ican. Because African American children are more likely than White children to come to the attention of child protection authorities (Hampton, 1991), this study provides a unique opportunity to gain insight into the recovery processes of this population. The sexual abuse these women experienced in childhood ranged from sexual intercourse to touching and fondling, all by male perpetrators. Sexual penetration was reported in 60% of the cases, and some type of physical force (pushing, shoving, slapping, beating, or choking) by the perpetrators was reported in 62% of the cases. The perpetrator was a member of the imme- diate or extended family in 34% of the cases, an acquaintance or friend of the family in 27%, a peer in 14%, and a stranger in 25%. Medical evidence of some physical trauma was present in 34% of the cases. For the purposes of the study, CSA was defined as 202AffiliaSummer 2001 sexual contact that was against the child’s wishes, involved force or coercion, and was perpetrated by a male who was 5 or more years older than the victim (Williams, 1995). The Model The theoretical perspective of this study was informed by the theoretical and empirical work of researchers in the field of resilience (Garmezy, 1985; Mrazek & Mrazek, 1987; Rutter, 1985, 1991), as well as of those who have identified the intercon – nections among five essential elements of a woman’s well-being (Barnett & Marshall, 1991; Voydanoff & Donnelly, 1989). The study investigated the factors and mechanisms that predict competent functioning by examining features of the child; her family environment; the abuse she experienced; the supports available to her at various times following the abuse; her opportunities for accomplishing tasks as an adolescent and adult; her experiences, such as teenage pregnancy, that may have foreclosed some choices; and evidence of turning points. Table 1 presents the authors’ conceptual model and the cate- gories of variables that were measured. The independent vari- ables in columns 1 and 2 were drawn from studies of resilience and the authors’ knowledge of factors that influence the adverse outcomes in survivors of CSA. The third column of Table 1 shows the five spheres of adult life and functioning, which have been shown to be affected by CSA, combined to cre- ate a measure of resilience. Resilience Scale Resilience, competent functioning in several interrelated spheres despite adversity, is operationalized as well-being in five essential spheres: physical health, mental health, interper- sonal relationships, adherence to community standards, and economic well-being. To analyze the data, the authors devel- oped a scale to measure “competent outcomes” in these women. Several dilemmas were addressed while developing the scale. First, at what age can it be said that an individual Hyman, Williams203 demonstrates a resilient response to a traumatic experience? Several studies have measured the resilience of children who were exposed to a variety of risk factors for poor outcomes, but few have done so in adults who experienced a traumatic event during their childhood (James et al., 1993). The passage of time between the event and the measurement of resilience leads to 204AffiliaSummer 2001 TABLE 1: Variables Used to Examine Resilience Features of Childhood Years Intervening Years Resilience in Adulthood Characteristics of the Availability and use of Psychological child social supports well-being Health Abuse-specific therapy TSC-40 score Closeness to mother A special relationship Self-esteem Number of moves Opportunity to feel special Foster care or institutionalization Experienced physical abuse Characteristics of the Educational status Health family Graduation from high school Severe illness or Race surgery Income Good in school Use of alcohol and/or Parents abused alcohol drugs or drugs Parental violence Parents’ levels of education Whether mother worked Type of discipline used Characteristics of the sexual Possible adverse outcomes Interpersonal abuse Teenage pregnancy relationships Age of the child Alcohol or drug use Intimate relationships Perpetrator was/was not Arrest during teenage years Parenting a family member Friendships Penetration Participation in the Physical force community Revictimization Economic well-being Sexual Personal earnings Battering Labor force participation Adherence to community standards Arrests in adulthood NOTE: TSC-40 = Trauma Symptom Checklist–40 (Briere & Runtz, 1989). the second dilemma: Should women who demonstrated a less adaptive response at some point during the intervening years but who now exhibit competent functioning be defined as resil – ient? The authors decided to consider women who exhibited competent functioning at the time of the study to be resilient, whatever their previous functioning. The notion of “compe – tent” is the root of the third dilemma: How can one know whether a woman considers her behavior to be competent and whether she developed an adaptive pattern of disguising her problems to appear healthy and functional? Clearly, the authors could not know whether women who appeared to be resilient were disguising their pain, but by designing a measure that took into account the women’s functioning in five spheres, they hoped to distinguish women who appeared outwardly competent in one or two spheres from those who were compe- tent in a range of areas. Most studies of survivors of CSA, including this one, have asked their respondents about the adverse consequences of the childhood victimization, rather than about competent out- comes. Transforming these measures of adverse effects into a scale that is valid in the study of resilience was complicated, and the authors recognize that they may not be able to present a full picture of the factors that contribute to competent function- ing. Nonetheless, five domains of resilient functioning for women were identified: psychological well-being, good health, successful interpersonal relationships, the absence of arrests as an adult, and economic well-being; and a 13-point scale (alpha = .65) was created by assigning 1 point for the resilient response on each of the 13 variables. FINDINGS The frequency distribution of the women’s scores on the resil- ience scale indicated that 71 (52%) of the 136 women demon- strated poor (a score of 2-5 points) or fair (a score of 6-7 points) resilience, and hence for them, the adverse consequences of the abuse predominated. The 40 women (29%) who exhibited good Hyman, Williams205 resilience (a score of 8-9 points) may have overcome some, but not all, of the adverse consequences of the CSA. Finally, 25 women (18%) demonstrated excellent resilience (competent functioning) in the majority of the domains explored; that is, they scored 10 or more points on the 13-point resilience scale. Table 2 presents the differences between the highly resilient (10-13 points) and nonresilient (2-9 points) women on the five domains as derived from chi-square cross tabulations. Five Domains of Resilience Psychological well-being. The women’s psychological well- being scores were measured on two scales—the Trauma Symp- tom Checklist-40, a measure of the psychological sequelae of childhood trauma (TSC-40; Briere & Runtz, 1989), and the self- 206AffiliaSummer 2001 TABLE 2: Aspects of Resilience: Highly Resilient Versus Nonresilient Women (in percentages) (N= 136) Highly Resilient Nonresilient Variable Women (n= 25) Women (n= 111 ) Psychological well-being Low TSC-40 score (below sample median) 84**** 40 High self-esteem score (above sample median) 88**** 40 Health status No severe illness or surgery after age 17 68 54 No current drug or alcohol use 88**** 53 Interpersonal relationships Has male friends/few problems with men 72**** 36 High sexual functioning score (above the sample median) 92**** 35 All her biological children live with her 100*** 78 Has never been reported for child abuse 92 88 Has female friends 100*** 75 Is socially active (belongs to an organization or is socially active several times a year) 80**** 38 Social deviance Has not been arrested as an adult 96** 74 Economic well-being Earns above the sample median 72**** 27 Works full-time 44**** 10 NOTE: TSC-40 = Trauma Symptom Checklist–40 (Briere & Runtz, 1989). **p≤.05. ***p≤.01. ****p≤.001. esteem scale of the Middlesex Hospital questionnaire (Bagley, 1980). A low score on the TSC-40 indicates the absence of psy – chological symptoms, but, as Wortman (1983) noted, this is only a starting point when exploring resilience; therefore, the authors also included a measure of positive self-perception. Of the 25 resilient women, 84% (n= 21) scored low (had fewer problems) on the TSC-40, and 88% (n= 22) scored above the median on the self-esteem scale. In contrast, only 40% (n= 44) of the 111 nonresilient women obtained similar scores on these two measures. Physical health. Health was measured in two ways. One mea – sure was whether the women experienced any severe illness or surgical procedure after age 17. According to this measure, the health of the resilient and nonresilient women did not differ significantly; 68% (n= 17) of the 25 resilient women and 54% (n= 60) of the 111 nonresilient women were in good health. The other measure of physical health was the use of alcohol or drugs, which is also a well-documented adverse conse- quence of CSA (Briere & Runtz, 1990; Najavits, Weiss, & Shaw, 1997; Peters, 1988) and an indicator that a woman may have dif- ficulty caring for herself and her children. Of the 136 women, 41% (n= 56) reported a current problem with drugs or alcohol. However, of the 25 resilient women, 88% (n= 22) did not report a substance abuse problem at the time of the interview, com- pared with 52% (n= 58) of the 111 nonresilient women. Interpersonal relationships. Resilience in the domain of inter- personal relationships was measured in several spheres: rela- tionships with partners, success as a parent, relationships with female friends, and participation in social activities. Because all the women identified themselves as heterosexual, having “very few problems with men” and having “one or more good male friends” were considered resilient responses with regard to relationships with partners. Whereas 72% (n= 18) of the 25 resilient women met this criteria, only 36% (n= 40) of the nonresilient women did so. In relation to sexual functioning, 92% (n= 23) of the 25 resilient women but only 35% (n= 39) of Hyman, Williams207 the 111 nonresilient women scored above the sample median on the measure of sexual functioning derived from an adaptation of Greenwald’s scale (Greenwald, Leitenberg, Cado, & Tarren, 1990). In the area of parenthood, 120 (88%) of the 136 women had children. The 25 highly resilient women had an average of 1.6 children, and the 111 nonresilient women had an average of 2 children—a difference between these two groups that is not significant. With regard to the women’s success as parents, the authors addressed whether a woman replicated her abusive childhood with her children and was reported to the authori – ties for child abuse. Women who were not reported for child abuse obtained one point on the resilience scale and, so as not to penalize them, women who did not have children were also assigned one point. All of the 25 highly resilient women, but only 78% (87) of the 111 nonresilient women were living with their children. Of the 136 women, only 15 said they had been reported for child abuse, and there was no significant difference between the resilient and nonresilient women on this dimension. Only 2 of the 15 were resilient women. In relation to friendships with other women, any woman who reported one or more friendships with other women scored one point. All the resilient women reported female friends, but only 75% (n= 83) of the 111 nonresilient women did so. Because women who are resilient are not socially isolated, either “belonging to an organization” or “participating in social activities several times a year” constituted a resilient response. Of the 25 resilient women, 80% (n= 20) met one of these criteria, versus 38% (n= 42) of the 111 nonresilient women. Arrests. Being arrested as an adult may be viewed as an indi- cator of deviant behavior. Of the 136 women, 22% (n= 30) had been arrested as adults. However, only 1 resilient woman ver- sus 29 of the nonresilient women (26%) had been arrested dur- ing their adult years. 208AffiliaSummer 2001 Economic well-being. In this society, an important measure of one’s worth is the capacity to support oneself and one’s chil – dren. Thus, earning above the sample median and working full- time were considered to be indicators of resilience. Because the authors were interested in a woman’s ability to support herself, they used her personal earnings in this variable, not the earn – ings of all the adults in her household. However, they recognize that they may have understated the resources available to the women and their children. Of the 25 resilient women, 72% (n= 18) earned more than the sample median versus only 27% (n= 30) of the nonresilient women. Because the median income of the sample was $8,500, it may be assumed that women who earned less were having difficulty supporting themselves or were relying on the sup- port of someone else. Only 22 of the total of 136 women— 44% (n= 11) of the 25 resilient women and 10% (n= 11) of the 111 nonresilient women—were working full-time. Because many factors may influence whether a woman works full-time, such as the pres- ence of young children in the home, the condition of the local labor market, and the woman’s level of skills, in future assess- ments, the authors may include women who work part-time but are still able to support themselves in the resilient category. Characteristics Associated With Resilience In Table 3, the highly resilient women are compared with the nonresilient women on the explanatory variables in the model. Characteristics of the CSA. All 136 women in the sample expe- rienced CSA. Although 54% (n= 73) of them were abused by family members, only 32% (n= 8) of the 25 highly resilient women were. Similarly, 58% (n= 79) of the sample, but only 28% (n= 7) of the highly resilient women, experienced severe physical abuse. Neither age at the time of the abuse nor whether the woman experienced genital penetration was asso- ciated with being highly resilient. Hyman, Williams209 Characteristics of the survivor and her family of origin. Resilience was apparent across the racial groups; that is, there was no sig- nificant association between being African American and exhibiting highly resilient behavior. However, three features of a woman’s family of origin were associated with high resil- ience: the absence of severe physical abuse, parents with no substance abuse problems, and a stable family. Thus, 72% (n= 210AffiliaSummer 2001 TABLE 3: A Comparison of the Highly Resilient and Nonresilient Women (in percentages) (N= 136) Resilient Nonresilient Variable Women (n= 25) Women (n= 111 ) Characteristics of the child sexual abuse Abused prior to age 9 64 57 Abused by family member 32*** 59 Abused by someone outside the family 88 82 No physical force 52* 33 No genital penetration 29** 13 Childhood/family characteristics African American race 75 86 Absence of severe physical abuse 72**** 35 Parents did not abuse drugs or alcohol 60*** 28 Child experienced fewer than three changes in caregivers 80**** 44 Stable family: composite score of foregoing three items 80**** 33 Parent completed high school 68 60 Never in foster care or institution 92**** 65 Healthy 96 95 Not poor 84 70 Close to mother 72 73 Mother worked 68 69 Received positive discipline 44 42 Absence of parental violence 80 75 Status in the years since the child sexual abuse Received support and opportunity to feel special 80** 58 Graduated from high school 56** 32 No teenage pregnancy 52 37 No teenage alcohol/drug problems 48 33 Arrested as a teenager 8** 26 Received abuse-related therapy 16 23 Sexually victimized as an adult 12** 35 *p≤.10. **p≤.05. ***p≤.01. ****p≤.001. 18) of the highly resilient women versus 42% (n= 57) of the total sample of 136 did not experience severe physical abuse as chil – dren; 60% (n=15) of the highly resilient women, but only 28% (n= 38) of the total sample, grew up with parents who did not have an alcohol or drug problem; 80% of the highly resilient women, compared with 44% (n= 60) of the total sample, moved fewer than three times; and only 8% (n= 2) of the highly resil – ient women, versus 65% (n= 88) of the total sample, had ever lived in a foster home or institution. Some of the most important findings regarded variables that were not significantly associated with resilience. Neither a girl’s health status before the abuse nor her health during the remainder of her childhood was significantly associated with high resilience. Furthermore, resilience was not associated with whether a girl was poor during childhood, had a close relation- ship with her mother, received positive discipline (discussion and time-out), witnessed parental violence, was kicked out of the family home, or had a mother who worked outside the home. Social support. Many clinicians have long theorized that vari- ous forms of support buffer the developing child from the potential adverse effects of CSA. Although this theory may be true, this study found that two forms of support are not signifi- cantly associated with high resilience—abuse-specific therapy and having an opportunity to feel special—whereas one form is associated with being highly resilient— receiving support from a special person, at some point during her life. These findings appear to contradict Garmezy’s (1983, 1985) work on support as a protective factor in children’s lives, but they may have cap- tured the fact that the protective effect of support during child- hood diminishes over the years. In addition, one irony in the recovery from trauma is that the traumatic experience can dis- rupt the support networks at the precise moment they are most needed. This disruption may be especially likely when the abuse is perpetrated by a family member, when institutional services interfere with the child’s support network, or when the child’s hardy personality prevents her from eliciting support. Hyman, Williams211 The authors also evaluated the association between the women’s current use of social supports and current levels of resilience. When asked whether they had considered seeking help with their personal problems during the past year, 75 (55%) of the 136 women said no, and 21 women (15%) said they had considered seeking help, but had not actually pursued it. Of the 39 women (29%) who sought assistance, 12 turned to friends; only 3 of the 12 women were highly resilient (p= .10). Neither seeking help from a member of the clergy nor seeking help from a family member was significantly associated with being highly resilient. It is interesting to note, however, that when the women were asked, “Is there anyone who lives out – side your household whose opinion you would consider seri – ously in making an important decision, ” 77% of the sample said yes. Of this group, 21% were highly resilient, compared with only 10% of those who would not turn to someone outside their household. Education. The resilient women were significantly more likely than the nonresilient women to have graduated from high school (56% vs. 32%). Furthermore, the highly resilient women were more likely to report that they did well in school and got along well with the teachers (48% vs. 26%). One must wonder to what extent these positive relationships bolstered the likelihood that the woman completed high school. In addi- tion, completing high school is an important predictor of an individual’s annual earnings and labor force status (Pindyck & Rubinfeld, 1991), two of the variables that were used to mea- sure resilience. Teenage pregnancy. In light of the apparent importance of completing high school for these survivors, the authors exam- ined whether pregnancy during high school might explain the differential graduation rates. It was found that 63% (n= 70) of the nonresilient women and 48% (n= 12) of the resilient women were pregnant as teenagers. The trend in these statistics, although not significant, may help explain who finished high school. 212AffiliaSummer 2001 Revictimization during adulthood. Two forms of victimization during adulthood were explored: rape and battering. Of the 25 highly resilient women, 12% (n= 3) experienced sexual victim- ization as adults compared with 35% (n= 39) of the nonresilient women. Furthermore, 40% (n= 10) of the resilient women and 60% (n= 67) of the nonresilient women reported an episode of battering by a boyfriend at some point in their lives. Although the difference in the two groups’ experiences in this regard was not significant (p= .07), clearly, those who were less likely to experience battering tended to demonstrate resilience. Factors That Predict Resilience Table 4 presents the predictive strength of factors from the sur- vivors’ childhoods and adolescence for resilience in adulthood. Building on the results of bivariate analyses, the authors identi- fied six variables that explain the resilience of these women: growing up in a stable family, not experiencing incest (perpe- trator not a family member), not experiencing physical force as a part of sexual abuse, not being arrested as a juvenile, graduat- ing from high school, and not being revictimized as an adult. To create a parsimonious model, the authors reduced three family- related variables (parents did not abuse alcohol or other sub- stances, the child did not experience severe physical abuse, and Hyman, Williams213 TABLE 4: Predictors of Resilience in Women Who Were Sexually Abused in Childhood: Regression Summary Table (N= 119) VariableB SE Bβ During childhood Stable family 1.19 .41 .24*** No incest –.94 .41 –.19** No physical force .91 .44 .18** During adolescence/adulthood Ever arrested as a juvenile –1.14 .46 –.19** Graduated from high school 1.35 .40 .26**** No sexual victimization after age 18 .87 .44 .16** R 2(Adjusted) = .34 F= 11.07,p< .00001 **p≤.05. ***p≤.01. ****p≤.001. the child did not experience more than two changes in care – givers) into a new variable, stable family, on the premise that the women who lived in stable families as children were more likely to be resilient adults. All the women in the sample were survivors of CSA, but those who experienced incest or CSA accompanied by physical force were less likely to be resilient, as were those who were arrested as teenagers. The strongest predictor of resilience was whether a woman graduated from high school, given that 56% of the highly resilient women, but only 32% of the nonresilient women did so. In the model, the final predictor of resilience was whether the woman was revictimized as an adult. Because this variable measured an event that may have been contempo – raneous with the resilience measure, it was not surprising to find that women who had been recently raped were less likely to be resilient. DISCUSSION The highly resilient women benefited from growing up in more stable homes, characterized by fewer moves and less parental drug abuse, were less likely to have experienced incest or severe physical abuse, and were more likely to have graduated from high school. The significance of the number of changes in caregivers may be an indicator that there is a threshold beyond which children who have also experienced additional stressors (such as an alcoholic parent and CSA) begin to have significant difficulties. The strength of the stable family variable in all the multivariate analyses indicates that family context plays a role as important as CSA in predicting a resilient outcome. Social support received since the CSA did not appear to explain the women’s resilience, but it is difficult to disentangle the association between support and resilience. For example, resilient girls may be more likely to reach out for support or may exhibit characteristics that draw support to them, leading to resilient adult outcomes. 214AffiliaSummer 2001 Paradise, Rose, Sleeper, and Nathanson (1994) suggested that the child’s existing psychosocial circumstances determine, at least in part, the nature of the functional outcome. As is char – acteristic of research on the consequences of CSA, the authors cannot provide this baseline data, although they were able to control for social class, family circumstances, and race at the time of the abuse. Some researchers seem to be interested in studies of resil – ience as a way of searching for a trait or factor that can be bot – tled and doled out to all survivors of CSA. Although the unique set of intrapsychic, social, and timing factors that may allow one person to successfully navigate the wake of trauma cannot be bottled, the types of skills that may be taught or enhanced to increase the likelihood of a successful outcome can be identi- fied. Thus, for example, social workers should direct their ener- gies toward helping students complete high school. In addi- tion, they must be careful not to use the findings of research on competent functioning to reinforce the idea that nonresilient children and adults are failures. One way to avoid this use of the findings is to focus on environmental factors that can be influenced, such as social support, rather than on individual factors that are less amenable to intervention. Researchers need to sharpen their understanding of resil- ience as a complex construct and to design studies that capture the richness of the processes during the years following the abuse. As they develop a more sophisticated understanding of the processes involved in resilience, they will be able to create new measures of resilience that are not predicated on the notion that resilience is the absence of pathology. Therefore, future research should investigate the factors associated with resilient outcomes in women who experience more than one childhood stressor. For example, is there an interaction effect between the CSA experiences and other childhood traumas that decreases the likelihood of a resilient outcome? Or, instead, do girls who survive more than one childhood trauma develop skills to cope with these experiences that foster resilience in adulthood? Among the many other questions for future research are these: Should the examination of competence be limited to Hyman, Williams215 particular domains, such as psychological functioning? Are there cultural differences in defining competent functioning, and, if so, how can these differences best be captured? Do resil – ient women follow a different developmental trajectory from that of nonresilient women after CSA? What will happen to the women in this study who were identified as resilient when they face future life stresses if, as the authors have argued, resilience is a function of a particular point in a woman’s life? The findings of this study lay a foundation for future exami – nations of competent functioning in adult survivors of CSA. In feminist research, it is important to recognize and represent women as agents in their own lives. Therefore, feminist researchers need to find ways to capture those parts of women’s experiences that are attributable to proactive behav- ior, rather than solely to a passive response to trauma. By iden- tifying particular intervening variables that suggest the more complex processes, this study has begun that work. REFERENCES Anson, O., & Anson, J. (1987). Women’s health and labour force status: An enquiry using a multi-point measure of labour force participation.Social Science and Medicine,25(1), 57-63. Anthony, E. J. (1987). Risk, vulnerability, and resilience: An overview. In E. J. Anthony & B. J. Cohler (Eds.),The invulnerable child(pp. 3-48). New York: Guilford. Antonovsky, A. (1987).Unraveling the mystery of health: How people manage stress and stay well. San Francisco: Jossey-Bass. Bagley, C. (1980). The factorial reliability of the Middlesex Hospital Question- naire in normal subjects.British Journal of Medical Psychology,53, 53-58. Barnett, R. C., & Baruch, G. K. (1987). Social roles, gender, and psychological distress. In R. C. Barnett, L. Biener, & G. K. Baruch (Eds.),Gender and stress (pp. 176-190). New York: Free Press. Barnett, R. C., & Marshall, N. L. (1991). The relationship between women’s work and family roles and their subjective well-being and psychological distress. In M. Frankenhaeuser, U. Lundberg, & M. Chesney (Eds.), Women, work, and health: Stress and opportunities(pp. 111-136). New York: Plenum. Baruch, G. K., Biener, L., & Barnett, R. C. (1987). Women and gender in research on work and family stress.American Psychologist,42, 130-136. 216AffiliaSummer 2001 Benham, L., & Benham, A. (1981). Employment, earnings and psychiatric diagnosis. In V. Fuchs (Ed.),Economic aspects of health(pp. 203-220). Chi – cago: University of Chicago Press. Berkman, L. F. (1986). Social networks, support, and health: Taking the next step forward.American Journal of Epidemiology,123, 559-562. Briere, J., & Runtz, M. (1988). Symptomatology associated with childhood sexual victimization in a nonclinical adult sample.Child Abuse and Neglect, 12, 51-59. Briere, J., & Runtz, M. (1989). The Trauma Symptoms Checklist (TSC-33): Early data of a new scale.Journal of Interpersonal Violence,4, 151-163. Briere, J., & Runtz, M. (1990). Differential adult symptomatology associated with three types of child abuse histories.Child Abuse and Neglect,14, 357-364. Briere, J., & Runtz, M. (1993). Childhood sexual abuse: Long-term sequelae and implications for psychological assessment.Journal of Interpersonal Vio – lence,8, 312-330. Chirikos, T. N., & Nestel, G. (1985). Further evidence on the economic effects of poor health.Review of Economics and Statistics,67, 61-69. Draucker, C. B. (1995). A coping model for adult survivors of childhood sex- ual abuse.Journal of Interpersonal Violence,10, 159-175. Frankenhaeuser, M. (1991). The psychophysiology of sex differences as related to occupational status. In M. Frankenhaeuser, U. Lundberg, & M. Chesney (Eds.),Women, work, and health: Stress and opportunities(pp. 39- 61). New York: Plenum. Garmezy, N. (1983). Stressors of childhood. In N. Garmezy & M. Rutter (Eds.), Stress, coping, and development in children(pp. 43-84). New York: McGraw-Hill. Garmezy, N. (1985). Stress-resistant children: The search for protective fac- tors. In J. Stevenson (Ed.),Recent research in developmental psychopathology. Oxford, UK: Pergamon. Garmezy, N. (1993). Vulnerability and resilience. In D. C. Funder, R. D. Parke, C. A. Tomlinson-Keasey, & K. Widaman (Eds.),Studying lives through time: Personality and development(pp. 377-398). Washington, DC: American Psy- chological Association. Greenwald, E., Leitenberg, H., Cado., S., & Tarren, M. J. (1990). Child sexual abuse: The long-term effects on psychological and sexual functioning in a non-clinical and non-student sample of adult women.Child Abuse and Neglect,14, 503-514. Hampton, R. (1991). Race, ethnicity, and child maltreatment: An analysis of cases recognized and reported by hospitals. In R. Staples (Ed.),The Black family: Essays and studies(pp. 178-191). Belmont, CA: Wadsworth. Harvey, M. R. (1996). An ecological view of psychological trauma and trauma recovery.Journal of Traumatic Stress,9, 3-23. Hauser, S. T., Vieyra, M.A.B., Jacobson, A. M., & Wertlieb, D. (1985). Vulnera- bility and resilience in adolescence: Views from the family.Journal of Early Adolescence,5, 81-100. Hyman, B. (2000). The economic consequences of child sexual abuse for adult lesbian women.Journal of Marriage and the Family,62, 199-211. Hyman, Williams217 James, J. B., O’Toole, J., & Liem, J. (1993).Assessing psychological resilience in survivors of childhood sexual abuse(Final report). Palo Alto, CA: Spring Foundation for Research on Women in Contemporary Society. Kendall-Tackett, K. A., Williams, L. M., & Finkelhor, D. (1993). Impact of sex – ual abuse on children: A review and synthesis of recent empirical studies. Psychological Bulletin,11 3, 164-180. Kessler, R. C., & McLeod, J. D. (1985). Social support and mental health in community samples. In S. Cohen & S. I. Syme (Eds.),Social support and health(pp. 219-240). New York: Academic Press. Larson, J. H., Wilson, S. M., & Beley, R. (1994). The impact of job insecurity on marital and family relationships.Family Relations,43, 138-143. Long, B. C., & Kahn, S. E. (1993). A theoretical integration of women, work, and coping. In B. C. Long & S. E. Kahn (Eds.),Women, work, and coping(pp. 296-312). Montreal: McGill-Queen’s University Press. McLeer, S. V., Deblinger, E. B., Henry, D., & Orvaschel, H. (1992). Sexually abused children at high risk for post-traumatic stress disorder.Journal of the American Academy of Child and Adolescent Psychiatry,31, 875-879. Mrazek, P. J., & Mrazek, D. A. (1987). Resilience in child maltreatment victims: A conceptual exploration.Child Abuse and Neglect,11, 357-367. Mullen, P., Martin, J., Anderson, C., Romans, S. E., & Herbison, G. P. (1993). Childhood sexual abuse and mental health in adult life.British Journal of Psychiatry,163, 721-732. Murphy, L. B., & Moriarty, A. E. (1976). The development of a vulnerable but resilient child. InVulnerability, coping, and growth(pp. 295-333). New Haven, CT: Yale University Press. Najavits, L. M., Weiss, R. D., & Shaw, S. R. (1997). The link between substance abuse and posttraumatic stress disorder in women: A research review. American Journal of Addictions,6, 273-283. Neiman, L. (1988). A critical review of the resiliency literature and its rele- vance to homeless children.Children’s Environments Quarterly,5, 17-25. Paradise, J. E., Rose, L., Sleeper, L. A., & Nathanson, M. (1994). Behavior, fam- ily function, school performance, and predictors of persistent disturbance in sexually abused children.Pediatrics,93, 452-459. Pearlin, L. (1989). The sociological study of stress.Journal of Health and Social Behavior,30, 241-256. Peters, S. D. (1988). Child sexual abuse and later psychological problems. In G. E. Wyatt, & G. J. Powell (Eds.),Lasting effects of child sexual abuse (pp. 101-117). Newbury Park, CA: Sage. Pindyck, R. S., & Rubinfeld, D. L. (1991).Econometric models and economic fore- casts. New York: McGraw-Hill. Rhodewald, F., & Zone, J. B. (1989). Appraisal of life change, depression, and illness in hardy and nonhardy women.Journal of Personality and Social Psy- chology,56, 81-88. Robertson, E. B., Elder, G. H., Skinner, M. L., & Conger, R. D. (1991). The costs and benefits of social support in families.Journal of Marriage and the Family, 53, 403-416. Romans, S. E., Martin, J. L., Anderson, J. C., O’Shea, M. L., & Mullen, P. E. (1995). Factors that mediate between child sexual abuse and adult psycho- logical outcome.Psychological Medicine,25, 127-142. 218AffiliaSummer 2001 Russell, D.E.H. (1986).The secret trauma: Incest in the lives of girls and women. New York: Basic Books. Rutter, M. (1985). Resilience in the face of adversity: Protective factors and resistance to psychiatric disorder.British Journal of Psychiatry,147, 5498-5611. Rutter, M. (1990). Psychosocial resilience and protective mechanisms. In J. Rolf, A. S. Marten, D. Cicchetti, K. H. Nuechterlein, & S. Weintraub (Eds.),Risk and protective factors in the development of psychopathology (pp. 181-214). New York: Cambridge University Press. Rutter, M. (1991). Protective factors: Independent or interactive?Journal of the American Academy of Child and Adolescent Psychiatry,30, 151-152. Stein, J. A., Golding, J. M., Siegel, J. M., Burnam, M. A., & Sorenson, S. B. (1988). Long-term psychological sequelae of child sexual abuse: The Los Angeles Epidemiologic Catchment Area Study. In G. E. Wyatt & G. J. Powell (Eds.),Lasting effects of child sexual abuse(pp. 135-154). Newbury Park, CA: Sage. Verbrugge, L. M. (1983). Multiple roles and physical health of women and men.Journal of Health and Social Behavior,24, 16. Verbrugge, L. M. (1987). Role responsibilities, role burdens, and physical health. In F. J. Crosby (Ed.),Spouse, parent, worker: On gender and multiple roles(pp. 154-166). New Haven, CT: Yale University Press. Voydanoff, P., & Donnelly, B. W. (1989). Economic distress and mental health: The role of family coping resources and behaviors.Lifestyles,10, 139-162. Waldron, I. (1980). Employment and women’s health: An analysis of causal relationships.International Journal of Health Services,10, 435. Waldron, I., & Jacobs, J. A. (1989). Effects of multiple roles on women’s health: Evidence from a national longitudinal study.Women and Health,15(1), 3-19. Werner, E. E., & Smith, R. S. (1982).Vulnerable but invincible: A study of resilient children. New York: McGraw-Hill. Williams, L. M. (1995). Recovered memories of abuse in women with docu- mented child sexual victimization histories.Journal of Traumatic Stress,8, 649-673. Wortman, C. B. (1983). Coping with victimization: Conclusions and implica- tions for future research.Journal of Social Issues,39, 195-221. Batya Hyman, Ph.D., is an assistant professor in the Department of Social Work, College of Human Services, Arizona State University West, 4701 Thunderbird Road, P.O. Box 37100, Phoenix, AZ 85069-7100; e-mail: [email protected] Linda Williams, Ph.D., is the codirector of the National Violence Against Women Prevention Research Center, Centers for Research on Women, Wellesley College, 106 Central Street, Wellesley, MA 02481; e-mail: [email protected] wellesley.edu. Hyman, Williams219
http://cad.sagepub.com/ Crime & Delinquency http://cad.sagepub.com/content/55/3/363 The online version of this article can be found at: DOI: 10.1177/0011128707306017 2007 2009 55: 363 originally published online 14 December Crime & Delinquency Kerryn E. Bell Gender and Gangs : A Quantitative Comparison Published by: http://www.sagepublications.com can be found at: Crime & Delinquency Additional services and information for http://cad.sagepub.com/cgi/alerts Email Alerts: http://cad.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: http://cad.sagepub.com/content/55/3/363.refs.html Citations: at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Gender and Gangs A Quantitative Comparison Kerryn E. Bell The Ohio State University Research and theory about female gang involvement remain scarce. Drawing on the National Longitudinal Study of Adolescent Health, this study addresseswhether males and females differ in risk factors associated with gang membership (e.g., community characteristics, parent-child relation ships, associations with deviant friends). Integrating theory and research from social disorganization, social control, and feminist perspectives on crime/ delinquency, few differences are found between boys and girls in terms of risk factors associated with gang membership and outcomes associated with gang involvement. Instead, the results indicate that parental social control, attachment, and involvement; school safety; peer fighting; age; and race similarly influence boys’ and girls’ gang involvement. Keywords: gang; gender; delinquency A growing interest in the prevalence and nature of youth gangs has been documented in the criminological literature (Cummings & Monti, 1993; Curry, Ball, & Fox, 1994; Esbensen, Deschenes, & Winfree, 1999; Huff, 1996; Spergel, 1990; Thornberry, Krohn, Lizotte, Smith, & Tobin, 2003). In part resulting from this, there has been increasing interest in female gang participation (Moore, 1991). Unfortunately, there remains little consensus concerning the extent and nature of female gang activity (see J. Miller, 2001, for a discussion of this point). Partially, this lack of consensus reflects the nongeneralizability of results from previous studies (e.g., Joe & Chesney-Lind, 1995; Laidler & Hunt, 2001; J. Miller, 1998, 2001; Wang, 2000). Moreover, research on the topic using the Denver Youth Survey (Esbensen & Huizinga, 1993; Esbensen, Huizinga, & Weiher, Crime & Delinquency Volume 55 Number 3 July 2009 363-387 © 2009 SAGE Publications 10.1177/0011128707306017 http://cad.sagepub.com hosted at http://online.sagepub.com 363 Author’s Note: Please direct all correspondence to Kerryn E. Bell, Department of Sociology, The Ohio State University, 300 Bricker Hall, 190 N. Oval Mall, Columbus, OH 43210; phone: (614) 886-0012; fax: (614) 292-6687; e-mail: [email protected] Article at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from 1993), the Rochester Youth Development Survey (Browning, Thornberry, & Porter, 1999; Thornberry & Burch, 1997), and the Seattle Social Development Study (Hill, Lui, & Hawkins, 2001), among others, has not employed nationally representative samples (Bjerregaard, 2002; Deschenes & Esbensen, 1999; Esbensen et al., 1999; Palmer & Tilley, 1995) or focused extensively on gender differences in factors associated with gang involvement.To address these shortcomings, the current study explores gender differ- ences in gang involvement using data from the National Longitudinal Study of Adolescent Health (Add Health), a nationally representative sample of school-age children (Grades 7 to 12), and quantitative methods to assess whether characteristics associated with gang involvement differ by gender. Addressing and building on themes from previous work, this article con- siders whether males and females differ in risk factors associated with gang membership. This is especially important because Thornberry et al. (2003) suggested that not only does gang involvement affect the lives of individu- als while in a gang but that it has negative long-term consequences as well. Background Gender Similarities in Gang Membership Recent research on gang involvement explores reasons that youth join gangs and highlights three common themes: neighborhood disadvantage, having gang-involved family or friends, and problems within the family such as neglect, lack of supervision, and drug or alcohol addiction. In a qualitative study of female gang members in Hawaii, Joe and Chesney-Lind (1995) argued that it is common for females growing up in poor neighbor- hoods to turn to gangs and that all of the females in their research des cribe a visible gang presence in their neighborhood. In addition, 90% of the female gang members report a family member, usually a sibling, in a gang. Documenting the extensive nature of victimization in these females’ lives, the study also reveals that three quarters of the sample report experiencing physical abuse and 62% report sexual abuse in their family. J. Miller’s (2001) qualitative study of female gang members in Columbus, Ohio, and St. Louis, Missouri, documents similar themes leading young women to join gangs. Most notably, females’ neighborhood contexts pro- vide exposure to gangs. In addition, many young women document the strong influence of gang-involved family members, and some females 364 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from describe the existence of serious family problems such as violence and drug abuse, leading them to avoid home in an attempt to meet their social and emotional needs elsewhere. Both of these studies, however, are nongener- alizable due to their reliance on small non–nationally representative samples.Along with these studies of female gang members, research similarly indicates that male gang members disproportionately come from lower- class communities characterized by social instability, poverty, and unem- ployment (Cartwright & Howard, 1966; Hagedorn, 1988; Jankowski, 1991; W. B. Miller, 1958; Moore, 1985; Oehme, 1997; Spergel, 1990). Furthermore, Covey, Menard, and Franzese (1992) argued that the probability of males joining a gang increases with greater neighborhood disadvantage and that males are greatly influenced to join a gang if parents, siblings, and fr iends are or have been members. Gender Differences in Gang Membership Despite these similarities in reasons why males and females join gangs, however, Campbell’s (1991) findings suggest that female gang members come from more disadvantaged backgrounds than male gang members. Specifically, she stated that females in gangs are more likely than gang males to come from broken, unemployed homes and report family violence including parental drug abuse. Moore and Hagedorn (1996) furthered this argument by making the distinction that male gang members usually come from conventional working-class families, whereas female gang members are more likely to come from “underclass” families that are abusive as well. These latter findings are consistent with the idea that females may need more of a “push” to initiate their involvement in serious offending than is the case for males (Steffensmeier & Allan, 1996). Thus, it is argued that females join gangs because they have few options in very disadvantaged neighborhoods and they are seeking a “familial” peer group and emotionally fulfilling relationships that they do not find in other areas of their lives (Campbell, 1991; Chesney-Lind, Shelden, & Joe, 1996; Fleisher, 1998; Harris, 1988). Fleisher (1998) found some variation by gender in this regard as males are often drawn to the “action” associated with gang involvement. However, Esbensen et al. (1999) found that with the exception of males being more likely to join a gang to acquire money, there are no gender differences in reasons to join a gang. In their study, both males and females reported joining gangs, above all else, for protection. Bell / Gender and Gangs 365 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Theoretical Context When considering gender and crime, there is still an invisibility of women and girls (Belknap, 2001). Naffine (1996) noted it is with some alarm that it is still possible to study criminals without referring to their gender. Thus, do similar or different factors, or a combination of both, influence men and women to offend? Traditional theories lean toward sim- ilar factors, whereas feminist criminologists argue for different influences over men and women. Such feminist theories increasingly take into account intersections of various kinds of disadvantage, which can be noted in the gang context (Giordano, Cernkovick, & Randolph, 2002). In considering gender and gang involvement, both traditional and feminist theories are incorporated here. Traditional Theories Social Disorganization Theory There is considerable growth in interest in urban crime, including under- standing the proliferation of gangs that has resulted in part from the r apid deterioration of living conditions in many urban areas. The implication that this deterioration has for adolescent supervision and thus involvement in gang activity is often interpreted through the framework of social disorga- nization theory. Social disorganization theory assumes that delinquency emerges in neighborhoods where neighborhood relations and social institu- tions have broken down and can no longer maintain effective social controls (Sampson & Groves, 1989; Shaw & McKay, 1942). Although Sampson and Groves’s (1989) work was significant in revi- talizing social disorganization theory, Lowenkamp, Cullen, and Pratt (2003) in particular validated this work by replicating this study. Using the 1994 British Crime Survey, the authors provided further proof that support for social disorganization theory is not time specific but an enduring social reality. Bursik and Grasmick (1993) contributed to the above basic model of social disorganization theory as well by suggesting that disorder (Skogan, 1990) may have a mediating effect on the relationship between social disorganization and levels of social control. Specifically, residential instability, heterogeneity, and poverty affect interpersonal relationships within the community that open opportunities for crime. Thus, Bursick and Grasmick opened the door for a consideration of social control theor y at the individual level to supplement a discussion of social disorganization theory. 366 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Using the above, then, it is assumed that in disadvantaged neighbor- hoods, less supervision of both boys and girls results in greater opportunity for deviant peer exposure, which in turn increases the likelihood of adoles- cents’ association with gangs. Furthermore, the breakdown of parent-child and neighborhood relations provides greater opportunity for males and females to associate with delinquent peers and to become immersed in gangs. Social Control Theory Social control theory assumes that all individuals would naturally com- mit crime if left to their own devices, yet most people do not commit crime as a result of certain controlling forces (i.e., individuals form social bonds that constrain them from acting on their antisocial nature). Hirschi ( 1969) argued that both males and females who are most tightly bonded to groups such as the family, school, and peers are less likely to commit delinquent acts through the attachment, commitment, involvement, and belief such social bonds entail. Therefore, adolescents with strong attachments to fam- ily and/or those youth who are more involved and supervised by parents, or feel a sense of belonging, are less likely to belong to gangs. Walker-Barnes and Mason (2001) also suggested that gang membership is tied to deficient family relationships and poor social control such as lack of supervision. They drew on various studies showing that parenting behavior is broadly related to general delinquent behavior (e.g., Blaske, Borduin, Henggeler, & Mann, 1989; Cernkovich & Giordano, 1987; DiLalla, Mitchell, Arthur, & Pagliocca, 1988; Keenan, Loeber, Zhang, Stouthamer-Loeber, & Van Kammen, 1995; Laub & Sampson, 1988; Mason, Cauce, Gonzales, & Hiraga, 1994; Shields & Clark, 1995). Feminist Perspective Statistically, criminologists see a great difference between amount and types of crime that men and women commit, commonly referred to as the gender gap in crime. As more feminist scholars have insisted that gender be considered in criminological research (Campbell, 1990; Joe & Chesney- Lind, 1995; Messerschmidt, 1995; J. Miller, 2002; Moore, 1991), researchers know more about girls’ experiences in gangs today. To that end, this study argues that traditional theories alone cannot explain gender differences in crime. Rather, a feminist perspective of offending emphasizes consideration of female victimization and how the social organization of gender shapes female involvement in crime. Chesney-Lind and Faith (2000) documented that victimization is at the heart of much female offending and that this pat- tern of gender entrapment best explains women’s inv olvement in crime. 1 Bell / Gender and Gangs 367 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from In terms of gang involvement specifically, feminist scholars have played a significant role in bringing to light the overlapping nature of female offending and victimization. J. Miller (2001) stated that there is evidence that many female gang members have histories of childhood victimization or find themselves in abusive adult relationships and that this sexual vic- timization continues within the gang. Furthermore, feminist scholars argue that considering the interconnected nature of women’s victimization and offending, women use violence differently than do men, that is, as a pro- tective measure in response to their vulnerability. Campbell (1993) argued that for gang males, then, violence provides power whereas for gang females, violence is a result of fear because violence is a better alter native than being victimized again. Campbell’s argument goes along with the idea of gender differences in crime involvement, specifically indicating that there are differences in reasons why men and women join gangs and that such differences need to be taken into account. In considering gender and gang involvement, then, both traditional and feminist theories discussed above are necessary to explain such involvement and will be shown through this research. It becomes further significant to include feminist theories as J. Miller (2002) pointed out research on girls in gangs suggests the need for ge nder- specific as well as gender-neutral interventions to address male and female gang members. Furthermore, Chmelynski (2006) documented how, in rec- ognizing and combating an increase of female gang members in schools, districts have implemented gender-specific programming to target girls, recognizing a necessity for such programming. This would support feminist theory assertions that reasons boys and girls join gangs, although they con- tain some similarities, are different as well. Hypotheses Hypothesis 1 The above literature and theories lead to the initial hypothesis that sim- ilar factors are associated with male and female gang involvement, yet these factors influence males and females to different degrees. Specifically, it is expected that the effect of neighborhood disadvantage, disorder, and crime; family disadvantage and structure; and parenting and peer behaviors will have a stronger effect on females’ likelihood of being in a gang in compar- ison to males’ likelihood. This is consistent with a feminist perspective and the idea that females may need an extra “push” to initiate involvement in delinquency or entrance into a gang (Steffensmeier & Allan, 1996). 368 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Hypothesis 2 Because there is some evidence that adolescents join gangs for a sense of protection, the second hypothesis is that adolescents who are more co n- cerned for their safety are more likely to be involved in gangs. For example, males and females living in neighborhoods with greater disorder and crime rates would be expected to report an increased need for safety. However, because previous literature suggests that both males and females in gangs live in neighborhoods with greater disorder and crime rates, gender differ- ences in the impact of safety on gang involvement are not expected. Tied to this is the use of exposure to peer deviance as a test of an intervening vari- able looking at the influence of social disorganization theory. It is expected that greater exposure to peer deviance is associated with gang membership similarly for both males and females. Specifically, it is hypothesized that both males and females are more likely to join a gang when they have greater association with deviant peers. Hypothesis 3 The third hypothesis assumes, as social control theory would argue, that parent-child relationship quality is associated with gang membership. Specifically, the ideas that adolescents are more likely to join gangs if they have much conflict with parents or have parents reporting a substance abuse problem are assessed. Once again, these variables are expected to similarly affect both males and females; however, they may have a stronger effect on females rather than males because females appear to need a stronger cata – lyst toward gang involvement and subsequent delinquency. Data and Method This study employs data from Add Health, a nationally representative sample of adolescents in Grades 7 to 12 with data collected in three waves between 1994 and 2002. Add Health is well suited for the current study because it includes measures of neighborhood characteristics, measures o f parenting practices and family disadvantage, and self-report data on ado- lescent involvement in gangs. On the other hand, because Add Health is a school-based data set, it can be assumed that serious gang youth would not be included. Even so, 5.4% of the boys and 2.5% of the girls in the Add Health data set claim gang involvement, which is very comparable to other research looking at female gang involvement (see the 2000 National Youth Gang Survey at http://www.ncjrs.org/pdffiles1/ojjdp/fs200204.pdf). Bell / Gender and Gangs 369 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Add Health consists of one in-school survey followed by three waves of in-home interviews as well as a parental questionnaire administered during the first in-home interview. A random sample of those interviewed in the initial in-school survey was chosen to participate in the longer in-home interviews. During the in-home interview, interviewers recorded the home address of each adolescent, allowing researchers to link neighborhood res- idence to census and other contextual data available at the state, county, tract, and block-group levels (Billy, Wenzlow, & Grady, 1998). The final sample size is limited to respondents with nonmissing data on the final vari- ables used in the analysis as discussed in Table 1. Thus, drawing from the parental and in-school questionnaires, Wave 1 and Wave 2 in-home inter- views, and contextual data from the 1990 U.S. Census, 7,212 respondents remain in the final sample. 2 Measures Dependent Variable The dependent variable for the research question evaluating gender dif- ferences in factors associated with gang involvement is a dichotomous indi- cator of gang involvement measured at Wave 2. Gang involvement assesses males’ and females’ response to the question, “Have you been initiated into a named gang?” Those who answer “yes” are coded 1, and those who answer “no” are coded 0. The strength of this measure is that it focuses on the robustness of self-nomination into a gang. On the other hand, the accu- racy of responses to the above question on a self-report survey cannot be verified. Furthermore, as with all cross-sectional data, because the question on gang involvement is only asked at Wave 2, establishing time order with risk factors and gang membership is a problem. Appendix A provides a detailed description of all variables included in the study. When studying gangs, it is important to define what is meant by the term “gang.” Esbensen, Winfree, and Taylor (2001) noted that there is little consensus among researchers as to what constitutes a gang or who is a gang member and that a popular strategy in defining gangs has been self-identification. Kornhauser (1978) argued that a major problem for gang research has been distinguishing between group and gang delinquency, thus suggesting that definitions of gangs are imprecise. To address this issue, previous research makes the distinction between gangs and other delin- quent groups by looking at the greater criminal propensity of gangs, foc us- ing specifically on the escalation in scope and form of gang activity in recent years (Klein, 1995; Klein & Maxson, 1989; W. B. Miller, 1982). 370 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from 371 Table 1 t Tests Describing Differences Between Females and Males (Within Gang Status) and Between Gang Members and Non–Gang Members (Within Gender) Female Male Gang Nongang GangNongang MSD MSD MSD M SD Community characteristic Neighborhood disadvantage index 0.97* a .85 –0.14 .27 1.11 .53 –0.43 .24 Neighborhood disorder 1.65* a,b .12 1.48* c .03 1.64* d .08 1.46 .02 Neighborhood crime e 5.89* a,b .43 5.25* c .28 5.69* d .36 5.24 .27 Family Family disadvantage 0.06* b .23 0.02* c .15 0.10* d .31 0.05 .22 Single (two biological parent reference) 0.03 .17 0.02* c .15 0.07* d .25 0.05 .22 Blend (two biological parent reference) 0.04 .19 0.02 .15 0.07 .26 0.05 .22 Other (two biological parent reference) 0.04 .21 0.02 .15 0.10 .30 0.05 .22 Parenting behaviors Alcohol abuse index 0.26 .25 0.06* c .05 –0.06* d .16 0.15 .06 Availability index 0.17* b .60 0.64* c .09 0.23* d .29 0.68 .10 Social control index 5.40* a,b .20 5.38* c .06 5.52* d .12 5.38 .06 Attachment index –0.85* b .80 0.98* c .15 –0.35* d .49 1.62 .15 Involvement index 5.99* a,b .46 6.09* c .11 5.23* d .33 5.56 .11 Safety School 2.93* a,b .27 3.63* c .05 3.19* d .17 3.74 .05 Neighborhood 0.02 .15 0.05 .21 0.05 .22 0.08 .26 Peer deviance Peer fighting 0.96* a,b .07 0.75* c .02 1.12* d .06 0.83 .02 Control variables Age 14.21* a,b .21 14.79* c .13 14.75* d .19 14.94 .14 Black (White reference) 0.02 .15 0.03* c .16 0.07* d .26 0.05 .22 Hispanic (White reference) 0.06* a,b .23 0.02 .14 0.08 .28 0.05 .22 Asian (White reference) 0.03.16 0.03 .16 0.06 .24 0.05 .23 Native American (White reference) 0.08* b .28 0.02 .15 0.10* d .30 0.05 .23 Other race (White reference) 0.00 .00 0.03 .16 0.08 .28 0.05 .23 Immigrant 0.00* a,b .06 0.03 .16 0.07 .25 0.05 .23 N 963,719 1863,211 a. Significant difference between female and male gang members. b. Significant difference between female gang members and non–gang members. c. Significant difference between female and male non–gang members. d. Significant difference between male gang members and non–gang members. e. Means and standard deviations are divided by 1,000. *p< .05. at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Furthermore, Thornberry et al. (2003) suggested that the more complex structure of the gang facilitates more serious forms of criminal activity, thus distinguishing the gang from other delinquent groups. The approach taken in this study is similar to previous researchers and uses self-nomination. Battin, Hill, Abbott, Catalano, and Hawkins (1998) used self-nomination and gang involvement in their study and noted that self-nomination and gang membership independently predict delinquency beyond the effect of having delinquent friends and prior delinquency. Esbensen et al. (2001) also used increasingly restrictive definitions of a gang and found that from a research perspective, self-nomination is the most robust measure of gang membership capable of distinguishing gang youth from non–gang youth. Using the precedent of previous research, this study considers self-nomination into a gang as a strong dependent measure of gang involvement.Independent Variables Neighborhood characteristics . Independent variables in this study con- sist of neighborhood and family characteristics as well as parent-child rela- tionship quality and exposure to peer violence. All of these variables, measuring risk factors associated with gang involvement, are measured at Wave 1. Three neighborhood characteristics are assessed to test social dis- organization theory: neighborhood disadvantage, neighborhood crime, and neighborhood disorder. Block-level data from the 1990 census are used to measure neighborhood disadvantage associated with adolescents’ resi- dence. 3Neighborhood disadvantage is measured as an index incorporating the proportion of female-headed households, the unemployment rate, the proportion of people below poverty level, and the proportion of households receiving public assistance. Neighborhood disorder is derived from the parental questionnaire and measures parents’ report of how big a problem drug dealers and users are in the neighborhood (1 =no problem to 3 =big problem). Neighborhood crime rates are measured as the total crimes per 100,000 in the reporting county in 1990, based on information provided by the FBI Uniform Crime Reports. Using drug dealing and subsequent disor- der as well as higher crime rates as correlates of a more disadvantaged neighborhood, this study is able to assess neighborhood exposure to disor- der and crime. Family disadvantage. As is the case with much literature measuring social disorganization theory, it is not just the influence of the neighborhood struc- tural factors but also the family unit itself. For e xample, Lowenkamp et al. 372 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from (2003), in replicating Sampson and Groves’s (1989) test of social disorga- nization theory, used a measure of family disruption that considers family structure. The research here uses a measure of family disadvantage that considers respondents’ families’ receipt of public assistance, a measure of family structure, and parental alcohol abuse. Receipt of public assistance is based on a survey item that asks parents whether they receive public assis- tance such as welfare (0 =no,1 = yes). Family structure is assessed with a dummy variable that compares respondents residing with two-parent bio- logical families to single, blended (i.e., step- and adoptive families), and other family types. The last measure of family disadvantage, parental alco- hol abuse, is based on a measure that asks the respondents’ parents how often they drink alcohol (coded 1 =never to 6 = nearly every day) and how often in the past month they had five or more drinks on one occasion (coded 1 = never to 6 = 5 +times). These two items are standardized and summed such that higher values indicate a greater likelihood of parental abuse of alcohol. As previous research on gender and gang involvement suggests the influence of family substance abuse (Campbell, 1991; Joe & Chesney-Lind, 1995; J. Miller, 2001), parental alcohol abuse is also an important concept to test here. Parent-child relationship quality. Social control theory can be assessed by looking at four aspects of parent-child relationship quality. First, a parental availability index is created based on a question that asks how often the resident mother and/or father is home when the adolescent leaves and returns from school (ranging from 1 =never to 6 = she/he takes me/brings me home from school ) and whether the resident mother and/or father is home when the adolescent goes to bed (ranging from 1 =never to 5 = always). These items are standardized and summed to create an index of parental availability. A parent-child social control index is created by incorporating the following items: how often the parent allows the adoles- cent to make his or her own decisions about the time he or she has to be home on weekends, the people he or she hangs around with, what he or she wears, how much television and what he or she watches, what time he or she has to go to bed on weeknights, and what he or she eats. Each item ranges from 0 = noto 1 = yes,and the items are summed to create the parent-child social control index. A parent-child attachment index measures how close the adolescent feels to his or her resident mother and/or father, how much he or she thinks the parents care for him or her, whether he or she feels the parents are warm and loving toward him or her, and how satisfied he or she is with the Bell / Gender and Gangs 373 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from parental relationship. The range for this index is 1 =not close at all/do not feel parent cares at all to 5 =extremely close/feel parent cares very much. These items are summed and standardized to create the attachment index. Finally, a parent-child involvement index asks whether, in the past 4 weeks, the adolescent reports that his or her resident mother/father par- ticipated in the following activities with him or her: gone shopping, played a sport, gone to a religious service or a movie, talked about a personal prob- lem or schoolwork, or worked on a school project. Each item on the parent- child involvement index ranges from 0 =no to 1 = yes, and the items are summed to create the involvement index. All of these measures are described in Appendix A. Safety. Safety, or a sense of protection, has also been described as a rea- son to join a gang (Esbensen et al., 1999). Thus, respondents are asked how safe they feel in school and in their neighborhood. Whether the adolescent feels safe in school is coded from 1 ( strongly disagree) to 5 (strongly agree), with a higher value indicating a feeling of greater safety. Neighborhood safety is also measured with a simple dichotomous response for whether the indi – vidual feels safe in his or her neighborhood (0 =no, 1 = yes). Finally, the role of peer influence, an indicator of peer violence, is included from the preconstructed network data. Each friend in the respondent’s friendship network reports on how often he or she has been involved in a physical fight. Each friend’s response is summed and divided by the number of friends in the network, resulting in a measure of friends’ average involve- ment in violence. 4These items measuring safety are also being used to test social disorganization theory and are used as measures of the individual’s reaction to his or her environment. Control Variables Controls for age (measured in years), race (measured with six dummy variables for Whites, Blacks, Hispanics, Asians, Native Americans, and other races, with Whites as the reference category), a measure of school dropout between Waves 1 and 2 (0 =nondropout, 1 = dropout), and a mea- sure of immigrant status (0 = non–foreign born,1 = foreign born) are also included. Appendix A provides definitions of all control variables included in the analyses. Appendix B provides a correlation matrix describing bivari- ate associations among all variables by gender (female correlations are above the diagonal and male correlations are presented below the diagonal). Of note, Appendix B indicates little evidence of high collinearity among independent variables. 374 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Analytic Strategy To test whether the neighborhood and family characteristics as well as parenting and peer behaviors described above are associated with gang membership, survey-corrected logistic regression is employed due to the dichotomous nature of the dependent variable. 5To test gender differences in the strength of the associations, interactions between gender and the independent variables of interest are introduced. Results Bivariate Analysis Table 1 presents initial bivariate tests describing differences between females and males within gang status and between gang members and non–gang members within gender. Gender Differences for Gang Members Focusing first on gender differences for gang members, Table 1 indicates that female and male gang members differ in terms of the influence of neighborhood disadvantage, disorder, and crime (female gang members experience less neighborhood disadvantage but greater neighborhood dis- order and crime). Female gang members also differ from male gang members in terms of parental social control and involvement (girls experi- ence less social control and more parental involvement), school safety (girls experience less safety), and peer fighting (female gang members experience less peer fighting). Finally, female gang members differ from male gang members in terms of their age (female gang members are on average younger), their race (female members are less likely to be Hispanic), and immigrant status (male gang members are more likely to be immigrants than are female gang members). Female Gang Members Versus Non–Gang Members Focusing on risk factors associated with gang membership and compar- ing female gang members to female non–gang members, female gang members report greater neighborhood disorder and crime, family disadvan- tage, and peer fighting yet less parental availability, attachment, and involvement and school safety than female non–gang members. Female gang members also experience more peer fighting than their nongang coun- terparts. Furthermore, female gang members are more likely to be Hispanic Bell / Gender and Gangs 375 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from or Native American, more likely to be younger, and less likely to be immi- grants compared to female non–gang members.Male Gang Members Versus Non–Gang Members In comparing male gang members and non–gang members, initial descriptive findings suggest that similar to females, males in gangs report greater neighborhood disorder and crime than males not in gangs. In addi – tion, male gang members show greater family disadvantage and are more likely to come from single-parent households than is the case for nongang males. Male gang members also show significantly less parental availabil- ity, attachment, and involvement as well as less school safety and greater exposure to peer fighting than their nongang counterparts. In addition, these bivariate results indicate that male gang members are more likely to be Black or Native American and younger than is the case for male non–gang members. Overall, descriptive results support the hypotheses that similar influences such as neighborhood factors as well as parenting and peer behaviors affect both boys and girls, but to different degrees. Multivariate Analysis Table 2 presents the results of multivariate analyses. Table 2 presents the results of logistic regression addressing whether community and family characteristics, parenting behaviors, safety, and peer deviance are associ- ated with male and female gang involvement. Although interactions were conducted with all independent variables, only the base model and those with significant interactions are presented here. Results indicate very few gender differences in the effect of risk factors on gang involvement. In terms of gender differences, females who are Hispanic do have greater odds of gang membership in comparison to Hispanic males. Furthermore, immi- grant status is more protective for females than is the case for males. However, the effect of community disadvantage, family disadvantage, par- enting behaviors, safety, and peer deviance on gang involvement is not moderated by gender. Model 1 serves as the baseline model and indicates that females are sig- nificantly less likely to be involved in a gang than males. Specifically, the odds that females will be gang members are 0.6 times lower than those for males. Furthermore, three parenting behaviors are significantly associated with gang membership. Counterintuitively, greater parental social control and involvement increase the odds of gang involvement. Perhaps this indi- cates parents’ increased efforts to mediate the effects of gang membership. 6 376 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from 377 Table 2 Logistic Regression of Community and Family Characteristics, Parenting Behaviors, Safety, and Peer Deviance on Gang Initiation (Standard Errors in Parentheses) a Model 1 Model 2 Model 3 BOdds bOdds bOdds Female –1.21 (.20)*** 0.39 –1.43 (.31)*** 0.35 –1.14 (.21)*** 0.42 Community Neighborhood disadvantage index 0.06 (.03) 1.04 0.06 (.03) 1.03 0.06 (.03) 1.04 Neighborhood disorder 0.14 (.15) 1.28 0.14 (.15) 1.28 0.14 (.15) 1.29 Neighborhood crime b 0.00 (.03) 1.00 0.00 (.03) 1.00 0.00 (.03) 1.00 Family Family disadvantage 0.30 (.33) 1.41 0.31 (.33) 1.42 0.29 (.33) 1.38 Single (two biological parent reference) 0.02 (.29) 0.71 0.04 (.28) 0.72 0.04 (.29) 0.71 Blend (two biological parent reference) 0.22 (.27) 1.29 0.23 (.27) 1.28 0.22 (.27) 1.29 Other (two biological parent reference) 0.10 (.39) 1.36 0.07 (.41) 1.37 0.11 (.40) 1.37 Parenting behaviors Alcohol abuse index –0.05 (.05) 0.98 –0.05 (.05) 0.98 –0.05 (.05) 0.98 Availability index 0.05 (.04) 1.01 0.05 (.04) 1.01 0.05 (.04) 1.01 Social control index 0.13 (.06)* 1.07 0.13 (.05)* 1.07 0.13 (.06)* 1.07 Attachment index –0.09 (.02)*** 0.92 –0.09 (.02)*** 0.93 –0.09 (.02)*** 0.92 Involvement index 0.08 (.03)* 1.04 0.08 (.03)* 1.04 0.08 (.03)* 1.04 Safety School –0.28 (.06)*** 0.81 –0.27 (.06)*** 0.81 –0.28 (.06)*** 0.81 Neighborhood 0.19 (.25) 1.02 0.18 (.25) 1.01 0.19 (.25) 1.02 Peer deviance Peer fighting 0.50 (.10)*** 1.73 0.49 (.10)*** 1.73 0.49 (.10)*** 1.73 (continued) at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from 378 Table 2 (continued) Model 1Model 2 Model 3 BOdds bOdds bOdds Control variables Age –0.16 (.05)*** 0.87 –0.16 (.05)*** 0.87 –0.16 (.05)*** 0.87 Hispanic (White reference) 0.98 (.23)*** 2.67 0.53 (.30) 2.18 0.98 (.23)*** 2.65 Black (White reference) 0.16 (.30) 1.26 0.25 (.34) 1.39 0.15 (.30) 1.25 Asian (White reference) 1.16 (.34)*** 2.24 1.19 (.44)** 2.20 1.12 (.34)*** 2.23 Native American (White reference) 1.07 (.42)** 2.65 0.95 (.54) 2.13 1.06 (.42)** 2.63 Other race (White reference) 0.46 (.64) 1.22 0.59 (.66) 1.60 0.48 (.65) 1.28 Immigrant –0.33 (.36) 0.69 –0.35 (.37) 0.70 0.04 (.37) 0.97 Female ×Hispanic 1.25 (.47)** 1.65 Female ×Black –0.31 (.54) 0.76 Female ×Asian –0.07 (.80) 1.03 Female ×Native American 0.40 (.79) 1.64 Female ×Immigrant –4.51 (1.12)*** 0.12 Intercept –1.65 (1.00)–1.68 (.98) –3.05 (1.08)** R 2 .10*** .11*** .11*** a. Only those models with significant interactions are included here. b. Means and standard deviations are multiplied by 1,000. *p < .05. **p < .01. ***p < .001. at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from More consistent with prior research, less parental attachment is associa ted with a greater likelihood of gang membership, and feeling less safe in school is also associated with increased likelihood of gang involvement. Greater exposure to peer fighting increases the probability of gang involve- ment, whereas older age reduces the likelihood of gang involvement. Last, being Hispanic, Asian, or Native American is indicative of greater gang involvement in comparison to the White reference category. Conclusion Overall, the importance of this research is the use of a large, nationally representative sample to assess associations of gang involvement for males and females that help elucidate whether the nature of gang involvement is gendered. Findings provide little evidence that risk factors associated with gang involvement differ for males and females. This study does find, however, significant gender differences in terms of entrance into a gang for Hispanics and if the respondent is an immigrant . Specifically, Hispanic females are significantly more likely to be involved in gangs than are Hispanic males. However, females who are first-generation immigrants to the United States are significantly less likely to be involved in gangs than males of the same status. Perhaps this lesser involvement of first-generation females in gangs indicates greater parental control of s uch girls on introduction to a new culture. Future research should continue to evaluate whether gender differences in gang experience are conditioned by race or ethnic identity. In addressing the research question (Are there gender differences in the risk factors associated with gang membership?), this study finds largely similarities between males and females in the factors associated with gang involvement. Neighborhood disadvantage, parent-child relationship factors, school safety, and exposure to violent peers similarly affected male and female adolescents’ likelihood of being in a gang. These findings corrobo- rate previous studies and traditional theories of crime such as social control theories, which suggest that the reasons for male and female involvement in crime—and in this case, reasons for joining a gang—are similar. However, in contrast to qualitative findings, which highlight the rate of neighborhood disadvantage for gang membership, findings here indicate that neighborhood disadvantage is unassociated with gang membership once relevant control variables are introduced. Perhaps this reflects a diffu- sion of gangs across the United States so that they are no longer limited to Bell / Gender and Gangs 379 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from disadvantaged urban areas in just a few cities (Dukes & Stein, 2003; Klein, 1995; Thornberry et al., 2003).Similar to prior findings, this study does indicate that gang involvement is associated with parenting behaviors and exposure to peer deviance among male and female adolescents. The importance of parental attach- ment for both males and females is consistent with previous literature, which suggests that adolescents, particularly females, join gangs to find “familial” and emotionally fulfilling relationships that they do not find in other areas of their lives (Campbell, 1991; Chesney-Lind et al., 1996; Fleisher, 1998; Harris, 1988). Exposure to peer violence is also associated with a greater likelihood of gang involvement and may be tied to less parental attachment. Thus, although the hypotheses were not proven, gender differences were found when it comes to the involvement of boys and girls of different races and ethnicities and immigrant statuses. It was argued at the beginning of this article that both traditional and feminist theories combine to explain gender and gang involvement. Social control theory was supported as sim- ilarities were found between gender when considering reasons why males and females join a gang. Surprisingly, a lack of difference by neighborhood or gender does not provide much support for social disorganization or fem- inist theories. Perhaps this article suggests it is best to get at gende r differ- ences through a combination of quantitative and qualitative research as such differences appear subtle and difficult to assess through just quantita- tive research. Future research on gender and gangs should thus include mul- tiple methods. Future research should also take a longitudinal approach to the evalua- tion of differences between male and female gang members such that causal statements about gang involvement can be made. Such research could also, then, look at gang influence on delinquency and victimization. In addition, reliance on self-report data alone can be improved by incorpo- rating various data sources into research. Specifically, using self-report data such as Add Health as well as official police reports or in-depth interviews on gang activity and with gang members in an area could provide greater reliability regarding gang involvement. Finally, the gang type and its impact on gender are also important. Future research should consider different types of gangs (i.e., gangs with females as auxiliary members through a ll- girl gangs). These suggestions for future research could augment research on gender differences in gang involvement. 380 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Appendix A Operationalization of Variables Variable NameDescription Dependent variable Gang initiation (Wave 2) 0 = no,1 = yes Independent variable Gender 0 = male, 1 = female Community characteristic Neighborhood disadvantage Using 1990 census, in block proportion households(µ =.93) that are female householder, unemployment rate, proportion people with income below poverty level in 1989, proportion of households receiving public assistance Neighborhood disorder How big a problem drug dealers and users are in the neighborhood (1 =no problem, 3 = big problem) Neighborhood crime Total crime rate per 100,000 in the reporting county Family Family disadvantage Is your family receiving public assistance? (0 =no, 1 = yes) Family structure Dummy variable for two biological parents (reference category), single parent, blended (i.e., step-/adoptive), and other Parenting behavior Alcohol abuse (µ = .64) How often have you drunk in the past month (1 =never, 6 = nearly every day), how often 5+ drinks in the past month (1 =never, 6 = 5 +times) Availability (µ = .70) Resident mother is home when respondent leaves for school (1 =never,6 = she takes me to school ), resident mother is home when respondent returns from school (1 =never, 6 = she brings me home from school ), resident mother is home when respondent goes to bed (1 =never, 5 = always), resident father is home when respondent leaves for school (1 =never, 6 = he takes me to school), resident father is home when respondent returns from school (1 = never,6 = he brings me home from school), resident father is home when respondent goes to bed (1 =never, 5 = always) Social control (µ = .66) Whether parents let respondents make own decisions about time they are home on weekends, people they hang around with, what they wear, how much television they watch, programs they watch, time they are in bed on weeknights, what they eat (0 =no, 1 = yes) Bell / Gender and Gangs 381 (continued) at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Attachment (µ =.86) How close respondents feel to resident mother, how much they feel resident mother cares, most of the time resident mother is warm and loving, respondents are satisfied with resident mother relationship, how close they feel to resident father, how much they feel resident father cares, most of the time resident father is warm and loving, respondents are satisfied with resident father relationship, how much they feel parents care (1 =not close at all/do not feel parents care, 5 = extremely close/feel parents care very much) Involvement (µ = .74) In the past 4 weeks, went shopping with resident mother, played a sport with resident mother, went to a religious service with resident mother, talked about dating with resident mother, went to a movie with resident mother, talked about a problem with resident mother, talked about schoolwork with resident mother, worked on a school project with resident mother, talked about other school items with resident mother, went shopping with resident father, played a sport with resident father, went to a religious service with resident father, talked about dating with resident father, went to a movie with resident father, talked about a problem with resident father, talked about schoolwork with resident father, worked on a school project with resident father, talked about other school items with resident father (0 = no,1 = yes) Peer deviance Fighting Safety School You feel safe in your school (1 =strongly disagree, 5 = strongly agree) Neighborhood Do you usually feel safe in your neighborhood? (0 =no, 1 = yes) Control variable Age Measured in years Race Dummy variable for White (reference category), Black, Hispanic, Asian, Native American, and all other races Immigrant Whether respondent is first-generation immigrant to the United States (0 = non–foreign born, 1 = foreign born) Not in school Whether respondent is in school or not (0 =nondropout, 1 = dropout) 382 Crime & Delinquency Appendix A (continued) Variable Name Description at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Appendix B Correlation Matrix (Females Above Diagonal and Males Below Diagonal) 123456789101112 1. Gang initiation — .05 .04 .04 .06 .01 .01 .02 .02 –.02 .00 –.05 2. Neighborhood .12 — .37 .30 .29 .24 –.02 .10 –.02 –.07 –.02 –.18 disadvantage index 3. Neighborhood disorder .07 .32 — .11 .17 .18 –.03 .02 .03 –.09 –.04 –.16 4. Neighborhood crime .04 .25 .06 — .06 .11 –.01 .03 .01 –.02 –.01 –.06 5. Family disadvantage .06 .30 .19 .05 — .26 –.03 .05 .06 –.08 –.06 –.17 6. Single (two biological .06 .21 .14 .12 .20 — –.24 –.10 .06 –.58 –.01 –.57 parents reference) 7. Blend (two biological .04 .03 .04 –.01 .01 –.22 — –.08 .10 .00 .03 –.04 parents reference) 8. Other (two biological .00 .13 .03 .07 .02 –.09 –.08 — –.06 –.02 .02 –.06 parents reference) 9. Alcohol abuse index –.03 –.05 .03 .01 .01 .08 .06 .00 — –.09 .06 –.07 10. Availability index –.04 –.08 –.10 –.02 –.08 –.56 –.01 .02 –.13 — –.08 .65 11. Social control index .02 –.04 .01 –.03 –.03 .06 –.01 .03 .06 –.14 — –.04 12. Attachment index –.10 –.16 –.16 –.06 –.18 –.61 –.05 –.05 –.12 .70 –.10 — 13 14 15 16 17 18 19 20 21 22 23 13. Involvement index — .19 .09 –.14 –.07 –.04 –.16 –.04 –.03 –.01 –.01 14. School safety .18 — .21 –.12 –.11 –.06 –.09 .01 –.06 .01 .01 15. Neighborhood safety .05 .13 — –.11 .05 –.06 –.12 –.01 –.02 .01 .01 16. Peer fighting –.07 –.05 –.06 — –.22 .03 .04 –.01 .07 –.03 –.03 17. Age –.09 –.14 .05 –.24 — –.01 .05 .02 –.02 –.02 .05 18. Hispanic (White reference) –.02 –.04 –.09 .05 .02 — –.14 –.06 –.05 –.03 .18 19. Black (White reference) –.11 –.10 –.05 .06 .04 –.13 — –.08 –.07 –.04 –.05 20. Asian (White reference) .01 –.04 .01 –.06 –.01 –.06 –.08 — –.03 –.02 .36 21. Native American –.03 –.02 –.04 .03 –.03 –.05 –.06 –.03 — –.01 –.03 (White reference) 22. Other race .03 –.04 –.02 .01 –.06 –.04 –.05 –.02 –.02 — .12 (White reference) 23. Immigrant –.03 –.04 –.06 –.06 .07 .19 –.02 .36 –.03 .02 — Bell / Gender and Gangs 383 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Notes 1. Campbell (1990) also argues that the general tendency is to minimize and distort moti- vations of female gang members as a result of gender bias on the part of male gang researchers, who describe females’ experiences from the male gang members’ point of view. 2. Cases were lost when restricting respondents to all three waves of data, namely, the in- school questionnaire and subsequent Wave 1 and Wave 2 in-home interviews. Furthermore, the final sample size was also restricted through incorporation of the parental questionnaire, net- work data, and finally inclusion of the National Longitudinal Study of Adolescent Health sam- pling weights necessary to adjust for the stratified cluster sampling design. 3. Based on high factor loadings in a principal components analysis, items in the neigh- borhood disadvantage index were summed and standardized at the block level. 4. For more information on the use of network data to assess peer deviance, see Haynie (2001). 5. All regression analyses incorporate survey weights that adjust the standard errors to reflect the clustered nature of the National Longitudinal Study of Adolescent Health sampling design (see Chantala & Tabor, 1999). 6. Unfortunately, it is impossible to determine if this interpretation is correct with t he National Longitudinal Study of Adolescent Health data because gang involvement is measured only during the second wave of in-home interviews. To assess this requires measures of par- enting practices and gang involvement at multiple points in time. References Battin, S. R., Hill, K. G., Abbott, R. D., Catalano, R. F., & Hawkins, J. D. (1998). The contribution of gang membership to delinquency beyond delinquent friends. Criminology, 36(1), 93-111. Belknap, J. (2001). The invisible woman: Gender, crime, and justice ( 2nd ed.). Belmont, CA: Wadsworth/Thomson Learning. Billy, J. A., Wenzlow, A., & Grady, W. (1998). The National Longitudinal Study of Adolescent Health: Part I, codebooks for the Waves I and II contextual database. Retrieved August 2002 from http://www.cpc.unc.edu/projects/addhealth/strategies/design.html Bjerregaard, B. (2002). Self-definitions of gang membership and involvement in delinquent activities. Youth & Society, 34 (1), 31-54. Blaske, D. M., Borduin, C. M., Henggeler, S. W., & Mann, B. J. (1989). Individual, family, and peer characteristics of adolescent gender offenders and assaultive offenders. Developmental Psychology, 25,846-855. Browning, K., Thornberry, T., & Porter, P. (1999). Highlights of findings from the Rochester Youth Development Study (Fact Sheet). Washington, DC: U.S. Department of Justice. Bursik, R. J., & Grasmick, H. G. (1993). Neighborhoods and crime: The dimensions of effec- tive community control. San Francisco: Jossey-Bass. Campbell, A. (1990). Female participation in gangs. In C. R. Huff (Ed.), Gangs in America. Newbury Park, CA: Sage. Campbell, A. (1991). Girls in the gang(2nd ed.). Cambridge, MA: Basil Blackwell. Campbell, A. (1993). Men, women and aggression. New York: Basic Books. Cartwright, D., & Howard, K. (1966). Multivariate analysis of gang delinquency: Ecological influences. Multivariate Behavioral Research, 1, 321-337. 384 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Cernkovich, S. A., & Giordano, P. C. (1987). Family relationships and delinquency. Criminology, 25, 295-319. Chantala, K., & Tabor, J. (1999). Strategies to perform a design-based analysis using the Add Health data. Retrieved August 2002 from http://www.cpc.unc.edu/projects/addhealth/ strategies.html Chesney-Lind, M., & Faith, K. (2000). What about feminism? In R. Paternoster & R. Bachman (Eds.), Explaining criminals and crime (pp. 287-302). Los Angeles: Roxbury. Chesney-Lind, M., Shelden, R., & Joe, K. A. (1996). Girls, delinquency, and gang member- ship. In C. R. Huff (Ed.), Gangs in America(pp. 185-204). Newbury Park, CA: Sage. Chmelynski, C. (2006). When “mean girls” turn to female violence. Education Digest, 72(4), 37-39. Covey, H. C., Menard, S., & Franzese, R. J. (1992). Juvenile gangs.Chicago: Thomas. Cummings, S., & Monti, D. J. (1993). Gangs: The origins and impact of contemporary youth gangs in the United States. Albany: State University of New York Press. Curry, G. D., Ball, R. A., & Fox, R. J. (1994). Gang crime and law enforcement recordkeep- ing. Washington, DC: National Institute of Justice. Deschenes, E. P., & Esbensen, F. (1999). Violence and gangs: Gender differences in percep- tions and behavior. Journal of Quantitative Criminology, 15 (1), 63-96. DiLalla, L. F., Mitchell, C. M., Arthur, M. W., & Pagliocca, P. M. (1988). Aggression and delinquency: Family and environmental factors. Journal of Youth and Adolescence, 17, 233-246. Dukes, R. C., & Stein, J. A. (2003). Gender and gang membership. Youth & Society, 34(4), 415-441. Esbensen, F., Deschenes, E., & Winfree, L. T. (1999). Differences between gang girls and gang boys: Results from a multisite survey. Youth & Society, 31 (1), 27-53. Esbensen, F., & Huizinga, D. (1993). Gangs, drugs, and delinquency in a survey of urban youth. Criminology, 31(4), 565-590. Esbensen, F., Huizinga, D., & Weiher, A. (1993). Gang and non-gang youth: Differences in explanatory variables. Journal of Contemporary Criminal Justice, 9, 94-116. Esbensen, F., Winfree, L. T., & Taylor, T. J. (2001). Youth gang and definitional issues: When is a gang a gang, and why does it matter? Crime & Delinquency, 47(1), 105-130. Fleisher, M. (1998). Dead end kids: Gang girls and the boys they know. Madison: University of Wisconsin Press. Giordano, P., Cernkovick, S., & Randolph, J. L. (2002). Gender, crime, and desistance: Toward a theory of cognitive transformation. American Journal of Sociology, 107 (4), 990-1064. Hagedorn, J. M. (1988). People and folks: Gangs, crime and the underclass in a rustbelt city. Chicago: Lakeview Press. Harris, M. G. (1988). Cholas: Latino girls and gangs. New York: AMS Press. Haynie, D. L. (2001). Delinquent peers revisited: Does network structure matter? American Journal of Sociology, 106 (4), 1013-1057. Hill, K., Lui, C., & Hawkins, J. D. (2001, December). Early precursors of gang membership: A study of Seattle youth. Juvenile Justice Bulletin,pp. 1-5. Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press. Huff, C. R. (1996). Gangs in America(2nd ed.). Thousand Oaks, CA: Sage. Jankowski, M. S. (1991). Islands in the streets: Gangs and American urban society. Berkeley: University of California Press. Joe, K. A., & Chesney-Lind, M. (1995). “Just every mother’s angel”: An analysis of gender and ethnic variations in youth gang membership. Gender & Society, 9(4), 408-431. Bell / Gender and Gangs 385 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Keenan, K., Loeber, R., Zhang, Q., Stouthamer-Loeber, M., & Van Kammen, W. B. (1995).The influence of deviant peers on the development of boys’ disruptive and delinquency behavior: A temporal analysis. Development and Psychopathology, 7, 715-726. Klein, M. W. (1995). The American street gang: Its nature, prevalence and control. New York: Oxford University Press. Klein, M. W., & Maxson, C. L. (1989). Street gang violence. In N. Weiner & M. Wolfgang (Eds.), Violent crime, violent criminals (chap. 8). Newbury Park, CA: Sage. Kornhauser, R. R. (1978). Social sources of delinquency. Chicago: University of Chicago Press. Laidler, K. J., & Hunt, G. (2001). Accomplishing femininity among the girls in the gang. British Journal of Criminology, 41, 656-678. Laub, J. H., & Sampson, R. J. (1988). Unraveling families and delinquency: A reanalysis of the Glueck’s data. Criminology, 26, 355-380. Lowenkamp, C., Cullen, F., & Pratt, T. C. (2003). Replicating Sampson and Grove’s test of social disorganization theory: Revisiting a criminological classic. Journal of Research in Crime & Delinquency, 40(4), 351-373. Mason, C. A., Cauce, A. M., Gonzales, N., & Hiraga, Y. (1994). Adolescent problem behav- ior: The effect of peers and the moderating role of father absence and the mother-child relationship. American Journal of Community Psychology, 22, 723-743. Messerschmidt, J. (1995). From patriarchy to gender: Feminist theory, criminology and the challenge of diversity. In H. Rafter & F. Heidensohn (Eds.), International feminist perspectives in criminology: Engendering a discipline (pp. 167-188). Philadelphia: Open University Press. Miller, J. (1998). Gender and victimization risk among young women in gangs. Journal of Research in Crime and Delinquency, 35 (4), 429-453. Miller, J. (2001). One of the guys: Girls, gangs, and gender. New York: Oxford University Press. Miller, J. (2002). Young women in street gangs: Risk factors, delinquency, and victimization risk. In W. Reed & S. Decker (Eds.), Responding to gangs: Evaluation and research (pp. 67-105). Washington, DC: National Institute of Justice. Miller, W. B. (1958). Lower class culture as a generating milieu of gang delinquency. Journal of Social Issues, 14 (3), 5-19. Miller, W. B. (1982). Crime by youth gangs and groups in the United States. Washington, DC: National Institute of Juvenile Justice and Delinquency Prevention. Moore, J. (1985). Isolation and stigmatization in the development of the underclass: The case of Chicano gangs in East Los Angeles. Social Problems, 33, 1-10. Moore, J. (1991). Going down to the barrio: Homeboys and homegirls in change. Philadelphia: Temple University Press. Moore, J., & Hagedorn, J. M. (1996). What happens to girls in the gang? In C. R. Huff (Ed.), Gangs in America (pp. 205-220). Thousand Oaks, CA: Sage. Naffine, N. (1996). Feminism and criminology. Philadelphia: Temple University Press. Oehme, C. G. (1997). Gangs, groups, and crime: Perceptions and responses of community organizations. Durham, NC: Carolina Academic Press. Palmer, C. T., & Tilley, C. F. (1995). Genderual access to females as a motivation for joining gangs: An evolutionary approach. Journal of Gender Research, 32(3), 213-217. Sampson, R. J., & Groves, W. B. (1989). Community structure and crime: Testing social- disorganization theory. American Journal of Sociology, 94 (4), 774-802. Shaw, C., & McKay, H. (1942). Juvenile delinquency and urban areas. Chicago: University of Chicago Press. 386 Crime & Delinquency at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from Shields, G., & Clark, R. D. (1995). Family correlates of delinquency: Cohesion and adapt-ability. Journal of Sociology and Social Welfare, 22, 93-106. Skogan, W. (1990). Disorder and decline: Crime and the spiral of decay in American neigh- borhoods. Berkeley: University of California Press. Spergel, I. A. (1990). Youth gangs: Continuity and change. In N. Morris & M. Tonry (Eds.), Crime and justice: An annual review of research (chap. 4). Chicago: University of Chicago Press. Steffensmeier, D., & Allan, E. (1996). Gender and crime: Toward a gendered theory of female offending. Annual Review of Sociology, 22 (1), 459-488. Thornberry, T., & Burch, J. (1997). Gang members and delinquent behavior (Bulletin). Washington, DC: U.S. Department of Justice. Thornberry, T., Krohn, M., Lizotte, A., Smith, C., & Tobin, K. (2003). Gangs and delinquency in developmental perspective. Cambridge, UK: Cambridge University Press. Walker-Barnes, C., & Mason, C. (2001). Ethnic differences in the effect of parenting on gang involvement and gang delinquency: A longitudinal, hierarchical linear modeling perspec- tive. Child Development, 72 (6), 1814-1831. Wang, J. Z. (2000). Female gang affiliation: Knowledge and perceptions of at-risk girls. International Journal of Offender Therapy and Comparative Criminology, 44 (5), 618-632. Kerryn E. Bell is a doctoral candidate in sociology at The Ohio State University. Her research interests include intersections of gender, race, and crime. She is currently looking at changes in patterns of female offending over time. Bell / Gender and Gangs 387 at RUTGERS UNIV on January 3, 2011 cad.sagepub.com Downloaded from
Adolescent Male and Female Gang Members’ Experiences with Violent Victimization, Dating Violence, and Sexual Assault Angela R. Gover &Wesley G. Jennings & Richard Tewksbury Received: 25 April 2008 / Accepted: 28 August 2008 / Published online: 21 January 2009 # Southern Criminal Justice Association 2009 AbstractA substantial amount of scholarship has been devoted to examining the relationship between gang membership and criminal offending. This research has produced a wealth of qualitative and quantitative studies indicating that gang membership increases the likelihood of criminal offending for both males and females. Less research, however, has examined the relationship between gang membership and violent victimization. The present study adds to the literature by examining the relationship between gender, gang membership, and three types of victimization. Specifically, this paper focuses on whether self-reported gang membership is uniquely related to victimization experiences for females compared to males. Results from a statewide survey of public high school students in South Carolina indicate that gang membership is significantly related to the risk of victimization for both males and females. The implications of these findings for research and policy are discussed. KeywordsGangs. Gender. Victimization. Dating violence Am J Crim Just (2009) 34:103–115 DOI 10.1007/s12103-008-9053-z NO9053; No of Pages A. R. Gover School of Public Affairs, University of Colorado Denver, 1380 Lawrence Street, Suite 525, Denver, CO 80217-3364, USA e-mail: [email protected] W. G. Jennings (*) Department of Justice Administration, University of Louisville, Brigman Hall, Room 215, Louisville, KY 40292, USA e-mail: [email protected] R. Tewksbury Department of Justice Administration, University of Louisville, Brigman Hall, Louisville, KY 40292, USA e-mail: [email protected] Introduction For nearly 100 years social scientists have conducted research on the relationship between gang activity and criminal offending (Esbensen and Huizinga1993;Esbensen and Winfree1998; Thrasher1927). This research has produced a number of qualitative and quantitative studies about the influence of gang membership on criminal activities (see Hughes2005;Thornberryetal.2003). Gangs, defined as“social groups that are organized around delinquency”(Miller1998, p. 429) are commonly associated with the sale of drugs and commission of violent acts (Hagedorn1998; Peterson et al.2004). In addition, membership in a gang increases criminal activity even among those who are already engaged in crime (Thornberry1995). Recent observational studies as well as survey research indicate that there is a strong association between membership in a gang and a variety of criminal activities (Katz et al.2005; Peterson et al.2004; Taylor et al.2007). The accumulation of research on gangs indicates that gang membership increases a youth’s involvement in crime and violence, which in turn increases their likelihood of becoming a victim of a violent crime (Hunt and Joe-Laidler2001; Miller 1998,2001; Miller and Decker2001; Peterson et al.2004; Taylor et al.2007). Criminological research has clearly established the relationship between participation in gangs and participation in delinquency, including violent behavior (Curry et al.1996, 2002; Dukes et al.1997; Egley and Ritz2006;Thornberryetal.2003). Research indicates that delinquent lifestyles lead to risk taking behaviors that are associated with victimization (Lauritsen et al.1991;MustaineandTewksbury2000). Consistent with this theoretical perspective, gang members are at an increased risk for violent victimization because they tend to be the main targets of violence from rival gangs and because violence occurs within gangs (Curry and Decker1998;Miller1998,2001; Miller and Decker2001; Taylor et al.2007). As gangs are increasingly organized there is a corresponding increase in the risk of victimization for members (Decker et al. 2008). Gang membership increases one’s risk of victimization because involvement in deviant behavior and risk taking activities (e.g., fighting, alcohol and drug consumption, etc.) places individuals in environments with offenders seeking criminal opportunities and victims (Mustaine and Tewksbury2000). Although a long history of gang research is available, it has not been until the last decade that female gang members have been seriously addressed (Esbensen et al. 1999; Fleisher1998; Miller1998,2001 ; Miller and Brunson2000; Miller and Decker2001; Moore and Hagedorn2001). One fact established in this literature is that females, although still marginalized and largely relegated to secondary and supportive roles in mixed sex gangs (see Fleisher1998; Turley2003), are moving towards equality in gang roles and activities (Anderson et al.2003). As such, not only are female gang members increasingly involved in committing delinquent (including violent) acts, but so too may they be increasingly vulnerable to violent victimization – because of their presence in potential offender rich environments. Gender and Gangs Frederic Thrasher’s(1927) classic studyThe Gangmentioned that female gangs and gang members were rare and that when present gang girls were near universally 104 Am J Crim Just (2009) 34:103–115 auxiliary members of male gangs. This view and understanding of females’relationship with gangs persisted for a half-decade. Not until the 1970s, accompanying women’s strides toward social, political and economic equality were female criminal offenders, including gang members, addressed by scholars, law enforcement officials and others as significant and serious public safety concerns (see Hughes2005). By the 1970s though, researchers began to recognize that females played important roles in youth gangs. Brown’s(1977) study of Black female gangs in Philadelphia, for example, found that“the female is an intrinsic part of the gang’s group identity and is involved in various gang functions, rather than just ancillary activities”(p. 226). This evolution of women’s roles has become a staple of contemporary gang research, with increasing recognition that females play significant, yet still not equal, roles in gangs (Fleisher 1998; Joe and Chesney-Lind1995; Miller1998,2001; Thornberry et al.2003). Although exact numbers may not be possible to identify, it is widely believed that the true number of female gang members is higher than official estimates indicate (Esbensen et al.2001). Across varying research sites and sampling methods studies have estimated that anywhere from less than 10% to up to 50% of gang members are females (Esbensen and Deschenes1998; Esbensen and Winfree1998). As their numbers have grown and their roles have evolved, females have experienced both increasing independence in some gang structures (Harris1988; Taylor1993)and an enhanced visibility in gang (and all criminological) research (Chesney-Lind1997; Curry1998; Esbensen et al.2001). Although gang involvement increases the risk that females will be involved in serious crimes, there remain both qualitative and quantitative differences between female and male gang members’forms, frequency and severity of delinquency and violence. Bjerregaard and Smith (1993), for example, found that serious delinquency was lower among gang girls compared to gang boys. Evaluation data from the Gang Resistance Education and Training (GREAT) program (Deschenes and Esbensen1999) also indicates that female gang members report lower levels of violent crime than their male counterparts. Interestingly, although female gang members’offending behaviors have been assessed and contrasted with that of males’, few attempts at examining the relationship between women’sgang membership and victimization have been completed (Miller and Decker2001). Gender, Gang Membership, and Victimization Gang membership, for both males and females, is associated with increased risks of criminal, especially violent, victimization (for review, see Taylor2008).Whilemalegang members typically have higher ratesof victimization (Taylor et al.2007), female’s rates and experiences are not insignificant, and are often the result of different traits, roles, relationships and behaviors than men’s victimizations. For female gang members the risks of victimization arise from both retaliatory behavior associated with being in a gang and sexual and physical victimizationfrom fellow male gang members (Fleisher 1998; Hunt and Joe-Laidler2001 ; Miller1998,2001). In fact, ethnographic research has suggested that young women may be used as“sex objects”by their own gang members (Campbell1990; Fleisher1998; Miller1998; however, also see Rosenbaum1996). Both ethnographic and survey research suggests that females’risk of victimization is clearly shaped by gender roles. For example, Miller (1998) notes that young women Am J Crim Just (2009) 34:103–115 105 adopting“masculine attributes”may increase their risk of victimization because they become more heavily involved in risky and delinquent behavior, and their victimization experience (e.g., sexual victimization) may increase the likelihood that they are victimized in the future because they become viewed as“weaker”than their male gang counterparts (pp. 433–434). Hunt and Joe-Laider (2001) also suggest that female gang members are at greater risk of victimization through the experience of gang initiation (also see Knox2004), conflicts with males in rival gangs, as well as conflicts with female gang members in both their own and rival gangs. Where females’victimization risks most clearly diverge from those of their male counterparts, however, is in the patriarchal and subjugating behavior of men in their own gangs. Girls in male dominated gangs are more likely to be sexually assaulted as a method of gang initiation (Fleisher1998;Knox2004;Miller1998). Additionally, female gang members may be required to sleep with multiple male gang members as a method of initiation. This pattern, however, does not appear to exist in gangs not dominated by males (Joe and Chesney-Lind1995; Miller2001). Research also suggests that gender may moderate the risk of serious violence for female gang members (Miller1998,2001). Qualitative interviews with female gang members and quantitative research on gang homicides in St. Louis, for example, suggests that females are at reduced risk of serious violent victimization (e.g., homicide) because they are not expected to engage in the most serious forms of violence that increase one’s risk for violent victimization (Miller and Decker2001). As Miller and Decker (2001)note,“Fewer expectations were placed on them [girls] with regard to involvement in criminal activities such as gun use, drug sales, and other serious crimes; this situation apparently limited their exposure to risk of gang-related victimization”(p. 131). In essence, by limiting their immersion in criminogenic settings and activities, female gang members’exposure to potential victimizers is reduced, thereby reducing instances of victimization. Thus, while the research reviewed above tends to focus on female gang members and victimization, it is certainly important to recognize and reiterate that victimization among male gang members is a prevalent and oftentimes frequent event (for review, see Taylor2008). Current Study At present, the relationship between women’s gang membership and victimization risks has been established through ethnographic work (e.g. Miller2001). However, there remains a gap in our knowledge about women gang members’victimization experiences, how these experiences contrast with their male counterparts, and what (if any) factors are related to greater and lesser victimization experiences for women in gangs. Also lacking at present is empirical assessments of the relationship between gang membership and victimization for females in general population samples. The present study adds to the literature by examining the relationship between gang membership and the risk of victimization for females and males among a sample of public high school students in the state of South Carolina. Analysis not only documents and empirically assesses victimizations, but also highlights gender differences in victimization patterns between male and female gang members across an entire state. 106 Am J Crim Just (2009) 34:103–115 Data and Method The data used in this study come from the 1999 South Carolina Youth Risk Behavior Survey (YRBS). The YRBS is an ongoing state and national survey conducted by state contracts for the Centers for Disease Control (CDC), Division on Adolescent and School Health. The original purpose of the YRBS was to estimate the prevalence of health-related risk behaviors among public high school students over time. The YRBS uses a three-stage, stratified cluster sampling design that, when weighted, represents the age, race, and gender composition of all South Carolina public high schools. In the first stage of the design, all South Carolina high schools were stratified into small (58–876 students), medium (877–1,284 students), and large (1,285–2,577 students) schools. In the second stage of the sampling design individual schools were sampled according to each of three strata, which resulted in a final sample of 58 (22 small, 21 medium, and 15 large) public high schools. Schools in each stratum were then rank-ordered according to their minority representation and selected systematically using a random start, with each stratum proportional to size of the enrollment. The third stage of the sampling design involved sampling the requisite number of classes in each school until approximately 100 students were enrolled. The unit of analysis for these data is the individual respondent (N = 4,597). This survey had an 83% student response rate. Dependent Variables Violent victimization experiences were assessed using three separate dichotomous measures. High school students were asked whether they had been injured in a physical fight in the past 12 months in which the injury required treatment from a doctor or nurse (violent victimization). Students were also asked whether they had ever been hit, slapped, or physically hurt on purpose by their girlfriend or boyfriend during the previous 12 months (dating violence). Finally, students were asked to report whether they had ever been forced to have sexual intercourse when they did not want to (sexual assault) during their lifetime. Thus, the violent victimization and the dating violence measures were annual prevalence estimates and the sexual assault measure was a lifetime prevalence measure because of the way the questions were asked in the SCYRBS. Explanatory Variables The main explanatory variables of interest are gang membership and gender. Gang membership was assessed by asking students to indicate if they had been a member of a gang during the previous 12 months. Gender was measured as a dichotomous variable (female = 1). Independent variables include family structure, age, and race as control variables since research indicates that these demographic factors have been shown to be associated with the risk of victimization (Lauritsen and White2001). Youth from two-parent households, for example, are less likely to engage in activities that place them at risk for victimization (Coker et al.2000;Gover2004). Therefore, family structure was Am J Crim Just (2009) 34:103–115 107 measured as a dichotomous dummy variable denoting households in which the respondent lived with both of their natural or adopted parents in comparison with single parent or otherwise configured homes (1 = two-parent family). Race was measured as a dichotomous dummy variable (1 = black, 0 = all other races). Age was measured as a continuous variable ranging from a minimum of 12 to a maximum of 18 years old. Description of Participants Ta bl e1presents the descriptive statistics for the dependent and independent variables. The mean age of the sample was 16 years (SD= 1.24). Of the participants, 42.3% were Black, 49% were White, and 7.7% were of other ethnic groups (e.g., Hispanics and Asians). Approximately 53% of the participants were female, and roughly 51% of the youth lived in nuclear families with two-parent households. Eleven percent (N = 528) of the sample indicated that they were active gang members during the preceding 12 months. As shown in Table1, victimization of these students was a relatively rare event. In terms of victimization experiences, 11% of the sample reported that they had been sexually assaulted during their lifetime. Approximately 12% of the sample reported being physically beaten by the person they dated during the prior year. Only 3.9% of the sample reported being physically injured during a fight in the past 12 months that required treatment by a doctor or nurse. These descriptive statistics indicate that both gang membership and victimization occur at relatively low rates within the general sample of high school adolescents. Findings The focus of this study is on gender differences in victimization patterns between male and female gang members. Table2provides a descriptive comparison of gender differences in victimization patterns and demographic characteristics for the entire sample as well as within the sample of gang members. For the entire sample, the data indicate that the risk of sexual assault is significantly higher for females compared to males. Specifically, 11.6% of females report being forced to have sex against their will during their lifetime compared to only 6% of males. In addition, females report a significantly higher prevalence of dating violence during the previous 12 months compared to males (11.3% versus 8.7%). In contrast, the Table 1Sample characteristics of youth Variables Mean Value SD Sexual Assault (%) 11.19 31.53 Dating Violence (%) 12.42 32.98 Violent Victimization (%) 3.92 19.40 Gang Membership (%) 11.50 31.90 African-American (%) 42.27 49.40 Average Age (years) 15.99 1.24 Nuclear Family (%) 51.64 49.97 Female (%) 53.36 49.89 108 Am J Crim Just (2009) 34:103–115 prevalence of violent victimization resulting in injuries from a physical fight during the previous 12 months is significantly higher for males (4%) compared to females (1.7%). With regard to the demographic characteristics of race, family structure, and age there appears to be few absolute differences between males and females, although males are significantly more likely than females to live in nuclear families with both biological parents (54.8% versus 51.1%). The results indicate greater parity with regard to victimization patterns between male and female gang members, with the exception of gang boys being more likely than gang girls to be injured in fights. Surprisingly, the results indicate that the prevalence of sexual assault is equally high for both male and female gang members. Specifically, 24.5% of male and 28.2% of female gang members report having been sexually assaulted during their lifetime. Additionally, there are not statistically significant gender differences in the prevalence of dating violence for gang members. More than one-quarter of male (28.1%) and female (29.9%) gang members report having been victims of dating violence in the previous year. This simple comparison of gender differences, therefore, suggests that for both males and females those involved in gangs have a greater risk of experiencing victimization. These comparisons, however, are limited because they cannot assess the independent relationship between gang membership, gender, and victimization patterns. To examine this issue a series of multivariate analyses are presented. Multivariate Analysis Our analysis focuses on the relationship between gang membership, gender, and victimization experiences among South Carolina public high school students. Because the dependent variables of interest are dichotomous, logistic regression models are used (Long1997). Because these data were collected using a complex survey design involving multistage cluster sampling, the use of standard logistic regression would be inappropriate (Lee et al.1986). To handle these issues a variance function estimation was used that was developed for use with multistage sampling designs that also involve weights (Eltinge and Sribney1996). Specifically, the complex survey estimation routine written for Stata version 6.0 was used, which estimates a logistic regression model using a pseudo-maximum likelihood routine that allows for probability sampling weights, stratification, and clustering (Stata Corporation1999). Results from the series of logistic regression models are displayed in Table3. First, models were estimated on the full sample to examine the independent relationship between gang membership, Table 2Characteristics of youth by gender and gang membership Total Sample (N = 4,591) Gang Youth Sample (N = 528) Male (46.6%) Female (53.4%) Male (63.8%) Female (36.2%) Sexual Assault (%) 6.0 11.6* 24.5 28.2 Dating Violence (%) 8.7 11.3* 28.1 29.9 Violent Victimization (%) 4.0 1.7* 14.0 9.0* African-American (%) 40.6 42.9 45.5 43.6 Nuclear Family (%) 54.8 51.1* 43.2 43.3 Average Age (years) 16.0 15.9 16.0 16.0 * t-test within sample comparison of means p < .05 Am J Crim Just (2009) 34:103–115 109 Table 3Logistic regression models of victimization by gender a Sexual Assault Dating Violence Violent Victimization Full (N = 4,482)Male (N = 2,088)Female (N = 2,401)Full (N = 4,488)Male (N = 2,088)Female (N = 2,406)Full (N = 4,460)Male (N = 2,065)Female (N = 2,402) Gang Membership 3.66* 4.58* 2.88* 3.53* 3.74* 3.28* 3.84* 3.43* 5.34* African-American 1.15 1.47* 0.97 1.17 1.21 1.14 1.02 0.86 1.37 Nuclear Family 1.11* 1.19* 1.07 1.07 1.16* 1.01 0.96 0.92 1.06 Age 0.75* 0.86 0.68* 0.85 0.91 0.81 0.79 0.81 0.81 Female 1.84* – – 1.36* – – 0.48* – – Nagelkerke R 2 0.07 0.11 0.03 0.05 0.07 0.03 0.07 0.05 0.05 a – Odds ratios reported * p < .05 110 Am J Crim Just (2009) 34:103–115 gender, and victimization experiences. Second, separate models for males and females were estimated to assess the interactive relationship between gender and gang membership on victimization. The results from all of the full models indicate that gang membership is significantly related to a higher risk of violent victimization. Self-reported membership in gangs increases the odds of sexual assault by 266%. Gang membership increased the odds of violent dating victimization by 253%. In terms of serious injuries from fighting, gang membership increased the odds by 284%. It is clear from these findings that gang membership is highly significant for the risk of violent victimization. The results also indicate that gender is associated with victimization risks. Females are at an increased risk of sexual assault and dating violence. For example, being a female increased the odds of sexual assault by 84%. Males, on the other hand, were at an increased risk of serious injuries from fights. Being a female decreased the odds of injuries from fights by 52%. The direction of the effect for family structure suggests that nuclear families provide some protection from victimization risk, but these effects are only statistically significant for sexual assault. Together these findings are consistent with previous research and indicate that victimization patterns are associated with gang membership among adolescents. The separate logistic regression models (by gender) indicate that gang membership is significantly related to all three forms of violent victimization for both males and females. For example, gang membership is associated with a 358% increase in the odds of sexual assault for males. Gang membership is also associated with a 274% increase in the odds of dating violence and a 243% increase in the odds of serious injuries for males. The models for females indicate similar results. Gang membership is associated with a 188% increase in the odds of experiencing sexual assault for females (lifetime estimate). Self- reported membership in a gang is also associated with a 228% increase in the odds of dating violence for females. For females who are members of gangs the odds of serious injuries from fights that require medical attention increases by 434%. These results suggest that for both sexes victimization experiences are significantly associated with gang membership. Whereas prior research has argued that female gang members have heightened risks for sexual victimization, this study suggests that such risk may not be unique to females. A comparison of coefficients test (see Paternoster et al.1998) for gang membership on all three measures of victimization between male and female models indicates nonsignificant differences. Consequently, gang membership does not have a gender specific effect on these measures of violent victimization. Therefore, it appears that there are global differences in the gender specific risks of victimization (e.g., girls are more likely than boys to experience sexual assault and dating violence), and being a member of a gang is associated with a greater risk of victimization for all adolescents. Discussion This study examined gender differences in gang membership and the risk of victimization. Most research on victimization of females in gangs has been restricted to single site qualitative studies. While these studies are informative about the process and nature of victimization for female gang members, they have not been able to make inferences about the overall risk of victimization for youth gang Am J Crim Just (2009) 34:103–115 111 members and the relative risk for females involved in gangs. Eleven percent (N = 528) of youths surveyed from the state of South Carolina self-reported membership in gangs, and approximately 36% of gang members were female. These results are consistent with other research and suggest that female gang membership is not rare (Esbensen and Deschenes1998; Esbensen and Winfree1998). In addition, the results from this research suggest that risks of victimization are not significantly different for male and female gang members. The unexpected finding from this study, however, was that female gang members did not have a greater risk for dating violence and sexual assault compared to similarly situated male gang members. There are, however, a number of limitations to the existing data that suggest caution should be exercised in interpreting the results. First, the data collected and analyzed focuses only on public high-school students in the state of South Carolina. While the sample is representative of the high school population in this state, it does not capture the risks associated with gang membership among the most at-risk youth – those who have dropped out of school. Second, the data used in this survey are cross-sectional and therefore do not allow one to make any assertions regarding causality or time-ordering of victimization and gang membership. As a result, the findings from this study can only discuss the general risk enhancement of violent victimization associated with gang membership and cannot explain whether gang membership specifically causes victimization. It is likely that some victimization experiences occurred before youth joined a gang and some victimization occurred subsequent to gang membership. Longitudinal data are needed to explain how much gang membership is the direct cause of future victimization. Third, when interpreting the findings from this research it is particularly important to note the differences in the measurement for the victimization experiences in that violent victimization and dating violence were annual prevalence estimates and sexual assault victimization was a lifetime prevalence estimate. However, it is reasonable to suspect that past sexual assault victimization experiences in the home may be a greater antecedent to gang membership for girls compared to boys (Chesney- Lind et al.1996). Fifth, the South Carolina Youth Risk Behavior Survey does not contain variables that allow one to test known theoretical constructs of victimization (e.g., routine activities, lifestyles, and association with delinquent peers) (Akers1998; Cohen and Felson1979; Hindelang et al.1978). As a result, we are not able to discern the independent contribution of gang membership relative to other known theories of victimization. Thus, the findings from this study should be interpreted with attention to the fact that there are likely a number of other theoretically relevant variables that if available may have affected the results presented herein (i.e., omitted variable bias). Finally, these data do not detail the situations that led to dating violence and sexual assault. It is possible that, for example, males use more liberal interpretations of dating violence and sexual assault compared to females. In addition, these data do not explain the extent to which males and females are embedded in the gang structure. It is possible that males on average have stronger gang attachments than females and as a result are more likely to become victims of violence. Taken together, adolescence is a key stage in the life course of individuals as they prepare for transitions into adult life. Part of a positive transition into healthy adulthood appears to be the avoidance of negative health related experiences, such as sexual assault and dating violence. Since gang membership is related to an adolescent’s risk for experiencing victimization, gang involvement is a key social ingredient in the life course. 112 Am J Crim Just (2009) 34:103–115 Implications From a primary prevention perspective, schools or communities should be encouraged to focus on positive youth development and the creation of alternatives to gang affiliation before hardcore gang activity takes hold (Evans et al.1999;Harperand Robinson1999). This may be particularly important given the current lack of success plaguing most gang intervention projects (Cohen et al.1995). Findings from this study suggest that programs focused on preventing or decreasing rates of physical fighting (resulting in injury), dating violence, and sexual assault should not ignore the relationship between experiencing these events and gang membership, if prevention programs are to be ecologically sound. The gang can be a micro-system that exerts the most powerful influence over adolescent behavior. If gang membership is ignored, prevention/intervention efforts may fail to bring about sustainable changes. Preventing children and adolescents from joining gangs promises to be a cost- effective long-term strategy. For example, one way to achieve and sustain gang and violence prevention/intervention is to target individuals and their environments (Hawkins et al.1992; Pentz et al.1989). The overall emphasis of gang and violence prevention programs should be on promoting positive health and social behaviors that are incompatible with a wide range of adolescent risk behaviors (Jessor et al. 2003). Gang membership and violence prevention is imperative vis-à-vis the intentional and unintentional injuries caused by physical fighting, dating violence and sexual assault. These behaviors and their consequences take their toll in direct medical costs, indirect economic costs to communities, and in the emotional and social costs of premature adolescent mortality and disability (Valois et al.1995; Valois and McKeown1998). Additional prevention/intervention efforts that have been shown to be effective with gangs include“lever-pulling strategies”where law enforcement tries to proactively disrupt street-level drug markets, aggressively serve warrants, assist with mounting federal prosecutions, and changing the community supervision of gang members who are out on parole and probation (see Weisburd and Braga2006). Ultimately, while much is known about the gang membership-offending link, research has only recently begun to examine the victimization experiences among gang members in general, and among male and female gang members in particular. We hope that the findings from this study provide a stepping stone for future quantitative research in this area. References Akers, R. (1998).Social learning and social structure: A general theory of crime and deviance. Boston, MA: Northeastern University Press. Anderson, J. F., Brooks, W., Langsam, A., & Dyson, L. (2003). The“new”female gang member: Anomaly or evolution?Journal of Gang Research,10,47–65. Bjerregard, B., & Smith, C. (1993). Gender differences in gang participation, delinquency, and substance use.Journal of Quantitative Criminology,4, 329–355. Brown, W. K. (1977). Black female gangs in Philadelphia.International Journal of Offender Therapy and Comparative Criminology,21, 221–228. Campbell, A. (1990). Female participation in gangs. In C. R. Huff (Ed.),Gangs in America. Thousand Oaks, CA: Sage. Chesney-Lind, M. (1997).The female offender: Girls, women, and crime. Thousand Oaks, CA: Sage. Am J Crim Just (2009) 34:103–115 113 Chesney-Lind, M., Shelden, R. G., & Joe, K. (1996). Girls, delinquency, and gang membership. In C. R. Huff (Ed.),Gangs in America(pp. 185–204, 2nd ed.). Thousand Oaks, CA: Sage. Cohen, L., & Felson, M. (1979). Social change and crime rate trends: A routine activities approach. American Sociological Review,44, 588–608. Cohen, M., Williams, K., Bekelman, A., & Crosse, S. (1995). Gang involvement. In M. W. Klien, C. L. Maxon, & J. Miller (Eds.),The modern gang reader(pp. 245–254). Los Angeles: Roxbury. Coker, A. L., McKeown, R. E., Sanderson, M., Davis, K. E., Valois, R. F., & Huebner, E. S. (2000). Severe dating violence and quality of life among South Carolina high school students.American Journal of Preventive Medicine,19, 220–227. Curry, G. D. (1998). Female gang involvement.Journal of Research in Crime and Delinquency,35,100–118. Curry, G. D., & Decker, S. (1998).Confronting gangs: Crime and community. Los Angeles: Roxbury Press. Curry, G. D., Ball, R. A., & Decker, S. H. (1996).Estimating the national scope of gang crime from law enforcement statistics. Washington, DC: National Institute of Justice. Curry, G. D., Decker, S. H., & Egley, A. (2002). Gang involvement and delinquency in a middle-school population.Justice Quarterly,19, 275. Decker, S., Katz, C. M., & Webb, V. J. (2008). Understanding the black box of gang organization: Implications for involvement in violent crime, drugs sales, and violent victimization.Crime & Delinquency,54, 153–172. Deschenes, E. P., & Esbensen, F. A. (1999). Violence and gangs: Gender differences in perceptions and behavior.Journal of Quantitative Criminology,15,63–96. Dukes, R. L., Martinez, R. O., & Stein, J. A. (1997). Precursors and consequences of membership in youth gangs.Youth and Society,29, 139–165. Egley, A., & Ritz, C. E. (2006).Highlights from the 2004 National Youth Gang Survey. Washington, DC: Office of Juvenile Justice & Delinquency Prevention. Esbensen, F. A., & Deschenes, E. P. (1998). Multisite examination of youth gang membership: Does gender matter?Criminology,36, 799–827. Esbensen, F. A., & Huizinga, D. (1993). Gangs, drugs, and delinquency in a survey of urban youth. Criminology,31, 565. Esbensen, F. A., & Winfree Jr., L. T. (1998). Race and gender differences between gang and nongang youths: Results from a multisite survey.Justice Quarterly,15, 505–526. Esbensen, F. A., Deschenes, E. P., & Winfree Jr., L. T. (1999). Difference between gang girls and gang boys: Results from a multisite survey.Youth and Society ,31,27–53. Esbensen, F. A., Osgood, D. W., Taylor, T. J., Peterson, D., & Freng, A. (2001). How Great is G.R.E.A.T.? Results from a longitudinal quasi-experimental design.Criminology and Public Policy,1,87–118. Eltinge, J., & Sribney, W. (1996). Linear, logistic and probit regressions for survey data.The Stata Technical Bulletin Reprints,6, 239–245. Evans, W. P., Fitzgerald, C., Weigel, D., & Chvilicek, S. (1999). Are rural gang members similar to their urban peers? Implications for rural communities.Youth & Society,30, 267–282. Fleisher, M. S. (1998).Dead end kids: gang girls and the boys they know. Madison: University of Wisconsin Press. Gover, A. R. (2004). Risky lifestyles and dating violence: A theoretical test of violent victimization. Journal of Criminal Justice,32, 171–180. Hagedorn, J. M. (1998). Gang violence in the postindustrial era. In M. Tonry, & M. Moore (Eds.),Youth violence: Crime and Justice: A Review of Research, Vol. 24(pp. 365–419). Chicago: University of Chicago Press. Harper, G. W., & Robinson, W. L. (1999). Pathways to risk among inner-city African American adolescent females: The influence of gang membership.American Journal of Community Psychology,27, 383–404. Harris, M. C. (1988).Cholas: Latino girls and gangs. New York: AMS. Hawkins, J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin,11 2,64–105. Hindelang, M. J., Gottfredson, M. R., & Garofalo, J. (1978).Victims of personal crime: An empirical foundation for a theory of personal victimization. Cambridge, MA: Ballinger Publishing Co. Hughes, L. A. (2005). The representation of females in criminological research: A content analysis of American and British journal articles.Women & Criminal Justice,16,1–28. Hunt, G., & Joe-Laidler, K. (2001). Situations of violence in the lives of girl gang members.Health Care for Women International,22, 363–384. Jessor, R., Turbin, M. S., Costa, F. M., Dong, Qi, Zhang, H., & Wang, C. (2003). Adolescent problem behavior in China and the United States: A cross-national study of psychosocial protective factors. Journal of Research on Adolescence,13, 329–360. 114 Am J Crim Just (2009) 34:103–115 Joe, K. A., & Chesney-Lind, M. (1995).“Just every mother’s angel:”An analysis of gender and ethnic variations in youth gang membership.Gender and Society,9, 408–431. Katz, C. M., Webb, V. J., & Decker, S. H. (2005). Using the Arrestee Drug Abuse Monitoring (ADAM) program to further understand the relationship between drug use and gang membership.Justice Quarterly,22,58–88. Knox, G. (2004). Females and gangs: Sexual violence, prostitution and exploitation.Journal of Gang Research,11,1–15. Lauritsen, J. L., & White, N. (2001). Putting violence in its place: The influence of race, ethnicity, gender, and place on the risk for violence.Criminology and Public Policy,1,37–60. Lauritsen, J. L., Sampson, R. J., & Laub, J. H. (1991). The link between offending and victimization among adolescents.Criminology,29, 265–292. Lee, E. S., Forthofer, R. N., & Lorimor, R. J. (1986). Analysis of complex survey data: Problems and strategies.Sociological Methods and Research,15,69–100. Long, J. S. (1997).Regression models for categorical and limited dependent variables. Thousand Oaks, CA: Sage. Miller, J. (1998). Gender and victimization risk among young women in gangs.Journal of Research in Crime and Delinquency,35, 429–453. Miller, J. (2001).One of the guys: Girls, gangs and gender. New York: Oxford University Press. Miller, J., & Brunson, R. (2000). Gender dynamics in youth gangs: A comparison of males’and females’ accounts.Justice Quarterly,17, 419–448. Miller, J., & Decker, S. H. (2001). Young women and gang violence: Gender, street offending, and violent victimization in gangs.Justice Quarterly,18,115–140. Moore, J., & Hagedorn, J. (2001).Female gangs: A focus on research. Washington, DC: National Institute of Justice. Mustaine, E. E., & Tewksbury, R. (2000). Comparing the lifestyles of victims, offenders, and victim- offenders: A routine activity theory assessment of similarities and differences for criminal incident offenders.Sociological Focus,33, 339–362. Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. (1998). Using the correct statistical test for the equality of regression coefficients.Criminology,36, 859–866. Pentz, M. A., Dwyer, J. H., & MacKinnon, D. P. (1989). A multicommunity trial for primary prevention of adolescent drug abuse: Effects on drug use prevalence.Journal of the American Medical Association, 261, 3259–3266. Peterson, D., Taylor, T. J., & Esbensen, F. A. (2004). Gang membership and violent victimization.Justice Quarterly,21, 793–815. Rosenbaum, J. L. (1996). A violent few: Gang girls in the California Youth Authority.Journal of Gang Research,3,17–23. Stata Corportation (1999).Stata statistical software: release 6.0. College Station, TX: Stata Corporation. Taylor, C. (1993).Girls, gangs, women, and drugs. East Lansing: Michigan State University Press. Taylor, T. J. (2008). The boulevard ain’t safe for your kids.Journal of Contemporary Criminal Justice,24, 125–136. Taylor, T. J., Peterson, D., Esbensen, F. A., & Freng, A. (2007). Gang membership as a risk factor for adolescent violent victimization.Journal of Research in Crime and Delinquency,44, 351–380. Thrasher, F. (1927).The gang. Chicago: University of Chicago Press. Thornberry, T. (1995). Membership in youth gangs and involvement in serious and violent offending. In R. Loeber, & D. P. Farrington (Eds.),Serious and violent juvenile offenders(pp. 147–166). Thousand Oaks, CA: Sage. Thornberry, T., Krohn, M., Lizotte, A., Smith, C., & Tobin, K. (2003).Gangs and delinquency in developmental perspective. New York, NY: Cambridge University Press. Turley, A. C. (2003). Female gangs and patterns of female delinquency in Texas.Journal of Gang Research,10,1–12. Valois, R. F., & McKeown, R. E. (1998). Frequency and correlates of fighting and carrying weapons among public high school adolescents.American Journal of Health Behavior,22,8–17. Valois, R. F., McKeown, R. E., Garrison, C. Z., & Vincent, M. L. (1995). Correlates of Aggressive and violent behaviors among public high school adolescents.Journal of Adolescent Health,6,26–34. Weisburd, D., & Braga, A. A. (Eds.) (2006).Police innovation: Contrasting perspectives. Cambridge University Press. Am J Crim Just (2009) 34:103–115 115