Sexual Violence Among Middle School Students: The Effects of Gender and Dating Experience

24
Journal of Interpersonal Violence 1–24 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0886260515590786 jiv.sagepub.com Article Sexual Violence Among Middle School Students: The Effects of Gender and Dating Experience Ethan Levine 1 Abstract Sexual violence has been increasingly recognized as a social, rather than strictly individual or family, problem. Unfortunately, providers and policymakers remain divided on the scope and causes of sexual violence, which limits their capacity to develop theory- and evidence-based responses. Such limitations are particularly pronounced in regards to children and adolescents. These youth are rarely addressed in the literature, and when they are, scholars tend to focus on adult victimization of children rather than children’s victimization of their peers. This study investigates the prevalence of unwanted sexual contact among middle students. Data are from a sample of 1,371 students attending New York City public middle schools. Drawing from current antiviolence curricula and scholarly literature on sexual violence, gender and dating experience are used as predictors of victimization and perpetration; race, age, and prior exposure to antiviolence programming are included as controls. Data reveal that, while boys’ violence towards girls comprises a substantial proportion of sexual violence in this population, same-sex violence and girls’ violence towards boys are also prevalent. Analyses conclude with recommendations for future antiviolence programming based on these patterns. 1 Temple University, Philadelphia, PA, USA Corresponding Author: Ethan Levine, Department of Sociology, Temple University, 723 Gladfelter Hall, 1115 W Polett Walk, Philadelphia, PA 19122, USA. Email:[email protected] 590786JIV XX X 10.1177/0886260515590786Journal of Interpersonal ViolenceLevine research-article 2015 at TEMPLE UNIV on June 25, 2015 jiv.sagepub.com Downloaded from

Transcript of Sexual Violence Among Middle School Students: The Effects of Gender and Dating Experience

Journal of Interpersonal Violence 1 –24

© The Author(s) 2015Reprints and permissions:

sagepub.com/journalsPermissions.nav DOI: 10.1177/0886260515590786

jiv.sagepub.com

Article

Sexual Violence Among Middle School Students: The Effects of Gender and Dating Experience

Ethan Levine1

AbstractSexual violence has been increasingly recognized as a social, rather than strictly individual or family, problem. Unfortunately, providers and policymakers remain divided on the scope and causes of sexual violence, which limits their capacity to develop theory- and evidence-based responses. Such limitations are particularly pronounced in regards to children and adolescents. These youth are rarely addressed in the literature, and when they are, scholars tend to focus on adult victimization of children rather than children’s victimization of their peers. This study investigates the prevalence of unwanted sexual contact among middle students. Data are from a sample of 1,371 students attending New York City public middle schools. Drawing from current antiviolence curricula and scholarly literature on sexual violence, gender and dating experience are used as predictors of victimization and perpetration; race, age, and prior exposure to antiviolence programming are included as controls. Data reveal that, while boys’ violence towards girls comprises a substantial proportion of sexual violence in this population, same-sex violence and girls’ violence towards boys are also prevalent. Analyses conclude with recommendations for future antiviolence programming based on these patterns.

1Temple University, Philadelphia, PA, USA

Corresponding Author:Ethan Levine, Department of Sociology, Temple University, 723 Gladfelter Hall, 1115 W Polett Walk, Philadelphia, PA 19122, USA. Email:[email protected]

590786 JIVXXX10.1177/0886260515590786Journal of Interpersonal ViolenceLevineresearch-article2015

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

2 Journal of Interpersonal Violence

Keywordsyouth violence, adolescent victims, sexual assault, adolescents, sexual harassment, sexual assault, GLBT

Since the 1970s, sexual violence has been increasingly recognized as a social, rather than individual or family, problem. The Centers for Disease Control and the National Institute of Justice have identified sexual violence as a pub-lic health issue for the United States and devoted considerable resources to research and program evaluation (Black et al., 2011; Morrison, Hardison, Matthew, & O’Neil, 2004; Tjaden & Thoennes, 2000). Universities have established partnerships with social services agencies to reduce sexual vio-lence and provide quality support for survivors,1 and the Obama administra-tion has publicly committed to addressing rape on campus (Calmes, 2014; Rothman & Silverman, 2007; Wasco, 2012). However, providers and policy-makers remain divided on the scope and causes of sexual violence, which limits their ability to develop theory- and evidence-based responses. Such limitations are particularly pronounced in regard to children and adolescents. These youth are rarely addressed in the literature (Morrison et al., 2004), and when they are, researchers tend to focus on adults’ and older children’s vic-timization of young children rather than children’s victimization of their peers (see C. F. Johnson, 2004; Pèrez-Fuentez et al., 2013). Inattention to sexual violence by peers poses a considerable barrier for prevention work. Program designers and advocates have little choice but to repackage curri-cula originally designed for older students and adults, and simply hope that the practices that work with these populations will work with early adolescents.

This research investigates sexual violence among middle school children, focusing on gender and dating experience as predictors of reported victimiza-tion and perpetration. As will be demonstrated below, these predictors were selected due to their presence in prevention curricula, many of which focus on gendered behaviors and dating violence, as well as in theoretical models that have been used to account for sexual violence among adults. Analyses address the following research questions:

Research Question 1: What is the prevalence of reported unwanted sex-ual contact among middle school students?Research Question 2: How does gender affect reported sexual violence victimization among middle school students?Research Question 3: How does gender affect reported sexual violence perpetration among middle school students?

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 3

Research Question 4: How does dating affect these relationships, and are the effects of dating different for girls and boys?

Literature Review

There have been no systematic investigations of sexual violence among early adolescents. However, research on older students and adults, as well as some data regarding dating violence in middle school, offer foundations from which to speculate. Brownmiller’s (1975) feminist model of rape and sexual script theory (Simon & Gagnon, 1986), which have often been applied to adults, offer theoretical frameworks for identifying factors that may contrib-ute to sexual violence among adolescents. Brownmiller’s feminist model posits patriarchy as the primary cause of sexual violence, whereas sexual script theory points specifically to gendered conventions in dating and sexual communication. These frameworks often implicitly or explicitly inform sex-ual and dating violence curricula for middle school students.

The Scope of the Problem

Recent data indicate that approximately 1 out of every 5 to 6 women and 1 out of every 33 to 71 men in the United States will be raped in their lifetime, and that as many as 1 in 2 women and 1 in 5 men will experience some form of unwanted sexual contact (Black et al., 2011; Tjaden & Thoennes, 2000). Estimates of child sexual assault vary, though they are consistently high. The Adverse Childhood Experiences Study (Centers for Disease Control and Prevention, 2014) found that as many as 1 in 4 girls and 1 in 6 boys may experience sexual abuse by adults and older children. Others have found that 1 out of every 10 to 11 children experience child sexual abuse (Pèrez-Fuentez et al., 2013; U.S. Department of Health & Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau, 2012). Data regarding child sexual assault by peers remain scarce. Researchers from the University of Illinois, Urbana–Champaign found that more than 20% of middle school students report some form of sexual violence on school grounds (Rhinehart, Doshi, & Espelage, n.d.). Although their study received some attention from news media (Dodge, 2014; Klein, 2014), it has yet to be published in a peer-reviewed journal.

These figures should be regarded with some caution. Definitions of sexual violence vary widely, which makes it difficult to compare prevalence esti-mates. Researchers, state officials, and service providers work with their own categories and definitions (see Whitman, 2012, for a review of contemporary sexual violence law in the United States). The FBI recently expanded its

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

4 Journal of Interpersonal Violence

definition of rape to encompass forced vaginal and anal penetration of any person by any body part or object, as well as forced oral penetration of any person by another person’s sex organ; the previous definition was limited to forced penile-vaginal penetration, and presumed a female victim and male perpetrator (Federal Bureau of Investigation, 2012; K. Johnson, 2012). Even if rates of sexual violence remain constant, this shift may produce a substan-tial increase in rapes documented through the Uniform Crime Reporting System.

Furthermore, sexual violence is notoriously challenging to measure. Underreporting is a major concern (Black et al., 2011). Victims may be reluc-tant to disclose their experiences. Perpetrators may be reluctant to admit guilt. Individuals may also disagree with researchers, police, and social ser-vice providers as to whether they have experienced or committed sexual vio-lence, particularly if they adhere to “real rape” stereotypes that privilege scenarios in which a (male) perpetrator violently attacks a (female) victim, who physically resists to her utmost capacity, as the most legitimate form of sexual violence (Jhally, 1994; Peterson & Muehlenhard, 2004). These factors suggest that any research relying on self-reports of sexual violence, including the present study, may understate its occurrence.

Causal Models and Primary Prevention: The Importance of Gender and Dating Experience

The wide scale movement to prevent sexual violence emerged through femi-nist activism in the 1970s. Brownmiller (1975) introduced a groundbreaking feminist model of rape, suggesting that individual men’s violence against women was a consequence of male dominance in society. This model drew attention to the social aspects of sexual violence. Rather than seek out the features of “typical” rapists and victims, it demanded engagement with social forces that contribute to rape culture. More recently, scholars have found a positive relationship between national sexual violence rates and gender inequality (Whaley, 2001; Yodanis, 2004). Others have expanded Brownmiller’s framework to incorporate additional dimensions of social inequality, such as race and class (Eschholz & Vieraitis, 2004). Such work has enabled scholars and activists to conduct more comprehensive analyses of sexual violence, and to provide culturally sensitive support to survivors.

Sexual script theory offers an alternative causal framework. Pioneered by Simon and Gagnon (1973/2005, 1986), it provides a threefold approach to human sexuality that incorporates intrapsychic scripts, comprised of indi-vidual desires and understandings of what sexual encounters can and should look like; interpersonal scripts, through which people negotiate sexual

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 5

interactions; and cultural scenarios, the broader social frameworks through which people learn how to engage (or not to engage) in sexual and romantic encounters. Sexual scripts are central to the negotiation of boundaries and consent. In addition, they often proscribe different behaviors for men and women. By the time that we reach middle school, we have been exposed to a range of cultural messages about sexuality, developed our own desires and understandings, and perhaps engaged in sexual encounters with other youth. Boys will have already engaged with scripts that encourage them to acquire sexual experience, to pressure girls to comply with their advances, and to interpret such behaviors as “teasing” as signs that girls are interested in sex. Girls will have engaged with scripts that encourage them to avoid becoming too sexually experienced, that pressure them to comply with boys’ desires, and that teach them that boys will interpret “teasing” as signs of desire (Anderson, Simpson-Taylor, & Herrmann, 2004).

There is some empirical support for a relationship between sexual scripts and sexual violence, at least among adults. Acceptance of rape myths is posi-tively associated with rape proclivity among straight nontransgender men (Bohner, Jarvis, Eyssel, & Siebler, 2005; Lonsway & Fitzgerald, 1994). Such misconceptions as “women who say ‘no’ are just playing hard to get” func-tion as cultural scenarios that influence individual approaches to sex and dat-ing (Anderson et al., 2004; Ryan, 2011). Peers’ acceptance of rape myths also contributes to men’s likelihood of perpetrating sexual violence (Bohner, Siebler, & Schmelcher, 2006). Shifting desires and boundaries, seduction scripts that emphasize teasing, and pressure to comply with unwanted advances contribute to uncertainty. This increases the likelihood of miscom-munication and the misinterpretation of dating behaviors, which can contrib-ute to sexual violence (Muehlenhard, 2011; Peterson & Muehlenhard, 2007; Sprecher, Hatfield, Cortese, Potapova, & Levitskaya, 1994). Participation in dating and “hookup culture” increases familiarity with dominant sexual scripts, and, may thus, intensify these risks (Stinson, 2010).

Middle school prevention curricula draw heavily on feminist and script-based models of sexual violence. Educators rarely make explicit references to Brownmiller or Simon and Gagnon, but they often appeal to sexism and gen-dered behavioral norms. Many middle school curricula incorporate sexual violence prevention into broader messages against dating violence, with an overt emphasis on boys’ violence toward girls (Sexual Trauma Services of the Midlands, n.d.; Taylor, Stein, Woods, & Mumford, 2012). Such curricula draw on assumptions that require empirical verification, namely that middle school students’ experiences of sexual violence are comparable with adults’, boys are the major perpetrators and girls the major victims of sexual violence, dating puts girls at increased risk of victimization and boys at an increased

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

6 Journal of Interpersonal Violence

risk of perpetration, and prevention education deters perpetration and improves risk reduction skills.

There are reasons to suspect that middle school students differ from older students and adults. Younger people have less experience with dating and sexual encounters (American College Health Association, 2013; De Rosa et al., 2010; Eaton et al., 2012). Middle school students are likely to be less familiar with “hookup culture,” to have had fewer opportunities to explore or challenge sexual norms, and to have had fewer opportunities to explore the gendered stereotypes addressed in rape myth literature (Anderson et al., 2004; Stinson, 2010). Finally, middle school students are unlikely to have lived or attended parties at dorms or fraternity houses, both of which have been associated with an increased risk of sexual violence for college students (Armstrong, Hamilton, & Sweeney, 2006).

This research will explore the empirical relevance of these assumptions to middle school students. Brownmiller’s feminist model connects societal gen-der inequality with individual men’s violence against women. Script-based models emphasize the gendered dynamics of communication in dating and sexual encounters, in which males are presumed to be initiators and females the recipients of advances. Taken together, these theoretical frameworks point to gender and dating experience as key predictors. All adolescents will encounter sexist ideas and institutions. Boys who date will engage with scripts that encourage aggression, perhaps even to the point of violating part-ners’ boundaries. Girls who date will engage with scripts that encourage pas-sivity and pressure them to engage in “token resistance” behaviors. This suggests the following hypotheses:

Hypothesis 1: Overall, boys will report higher rates of perpetration, whereas girls will report higher rates of victimization.Hypothesis 2: Gender and dating should have an interactive relationship such that dating experience will increase boys’ likelihood of perpetration while increasing girls’ risk of victimization.

It should be noted that both frameworks tend to present sexual violence as a heterosexual problem. Brownmiller’s approach presumes male aggressors and female victims. Scholars who use script-based approaches often limit their analyses to men’s violence against women. Consequently, these models have been limited in their capacity to account for same-sex violence or females’ violence against males. This research will address these matters through its broad gendered approach, including analyses of middle school students’ reports of overall victimization, victimization by female peers, vic-timization by male peers, overall perpetration, perpetration against male

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 7

peers, and perpetration against female peers. Unfortunately, due to data set limitations, it will not be possible to determine the effects of sexual orienta-tion or gender identity. Transgender and nontransgender students and stu-dents of all sexual orientations could not be analyzed separately.

Data and Method

This study relied on secondary data from an experimental evaluation of dat-ing violence and sexual harassment prevention programs in New York City middle schools during the 2009-2010 school year (Taylor et al., 2012). Sixth and seventh graders at 30 public schools were randomly assigned to three intervention groups and one control group. All participants were given sur-veys that asked about their beliefs, anticipated behaviors, and past experi-ences of violence and harassment. Surveys were administered immediately before interventions, immediately after, and 6 months later. This research used data from pretests, which had a response rate of 93% (N = 2,655).

Independent Variables

Gender and dating experience were included as independent variables. Both were assessed through individual survey questions: “Are you: female/male”; and “Have you ever been in a boyfriend/girlfriend dating relationship that lasted more than a week?” Admittedly, the gender question conflated sex with gender identity. Students were not asked separately about their legal sex and gender identity, nor were they asked whether they identified as transgen-der. Consequently, any inferences about gender must be based on respon-dents’ answers to this question. There are also potential limitations to the dating experience question. It may be interpreted as somewhat queer-inclu-sive (i.e., “Have you ever had a girlfriend or a boyfriend for more than a week?”), but it may also be interpreted as specific to different-sex relation-ships (i.e., “Have you ever been in a relationship that involved one boyfriend and one girlfriend that lasted more than a week?”).

Control Variables

Prior antiviolence programming, race, and age were included as control vari-ables. Programming was included as a dummy variable, via a survey question asking whether respondents had “ever attended an educational program about sexual harassment, sexual assault/rape, dating violence, and/or family/domestic violence.” The available data on race divided respondents into Asian, Black or African American, White, Multiracial/Other, and “don’t want

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

8 Journal of Interpersonal Violence

to answer” categories. Respondents in the latter group were coded as missing. Racial categories were included as dummy variables, using White as a refer-ence group. Available data on age divided respondents into 11 and younger, 12 years old, and 13 and older categories. These were included as dummy variables, using 11 and younger as a reference group.

Dependent Variables

Reported victimization was assessed through eight questions that addressed direct, nonconsensual sexual contact with peers. This allowed for a broad approach to sexual violence, while also emphasizing the sort of incidents that are most consistently addressed in criminal statutes concerning sexual vio-lence among adults (see Whitman, 2012). Two questions addressed unwanted sexual contact in any environment (e.g., whether a peer had ever touched a respondent’s private parts without consent), the remaining six addressed unwanted sexual contact at school. Students who disclosed victimization were asked to indicate whether they had been assaulted by male or female peers. This allowed for the development of three dichotomous, “yes/no” dependent variables for victimization: overall reported victimization, reported victimization by female peers, and reported victimization by male peers. Given the problem of underreporting, students who answered any of these questions were retained for analysis. If a respondent indicated that a female peer had “ever pushed, grabbed, shoved, or kicked you in your private parts,” and did not answer any other questions about victimization, that person was marked as having experienced victimization in general and victimization by female peers in particular.

Perpetration was assessed through an additional eight questions that addressed direct, nonconsensual contact with peers. These questions addressed the same forms of sexual violence, and asked respondents who disclosed perpetration to indicate whether they had assaulted male or female peers. This allowed for the development of three additional dichotomous, “yes/no” dependent variables for perpetration: overall reported perpetration, reported perpetration against female peers, and reported perpetration against male peers. Respondents who answered any of these questions were retained for analysis. If a respondent indicated having “ever pushed, grabbed, shoved, or kicked” a female peer in the private parts, and did not answer any other questions about perpetration, that person was marked as having engaged in perpetration in general and perpetration against female peers in particular.

All survey questions were developed by Taylor and colleagues (2012) for the baseline survey in their evaluation of violence prevention programs in New York City middle schools. The majority were drawn from survey instruments

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 9

that had been validated in previous research with Cleveland middle schools (Taylor, Stein, Mack, Horwood, & Burden, 2008). Cronbach’s alpha analyses were run for each dependent variable in the present study: overall victimization (α = .81), victimization by male peers (α = .73), victimization by female peers (α = .85), overall perpetration (α = .77), perpetration against male peers (α = .60), and perpetration against female peers (α = .83). Although the value for perpetration against male peers may seem low, it is important to note that these analyses speak to the consistency, rather than the general prevalence, of reported experiences. Higher Cronbach’s alpha values indicate that those who report one form of sexual violence among peers are comparatively likely to report additional forms. For example, in regard to overall victimization, those who report being forced to touch a peer’s private parts in any setting are rela-tively more likely to report having been sexually touched without permission at school. Dependent variables in this study were not designed to assess consis-tency across experiences, but rather to determine whether participants had ever experienced any among several forms of sexual violence among peers. See the appendix for a complete list of items used to create dependent variables, includ-ing bivariate relationships between gender and reported experiences of specific forms of victimization and perpetration.

Data Analysis

Data were analyzed using STATA, version 11.2. Respondents with missing values for gender, dating experience, victimization, or perpetration were excluded from analysis. The resulting sample included 1,371 students. Although this may seem like a drastic decline in sample size, the vast major-ity of dropped cases were excluded due to missing values for gender and dating experience, which in turn likely resulted from an issue in survey design. Participants completed a 12-page survey, and questions concerning personal demographics and dating experience appeared on the very last page. Many had simply stopped answering questions at this point, perhaps due to boredom or general disinterest. Those who completed the survey in full are unlikely to differ from those who did not in regard to sexual violence histo-ries, so much as stamina for lengthy questionnaires. Preliminary analyses with the original sample revealed similar patterns in reported victimization and perpetration (see below), with reductions in overall rates. These changes were likely the result of skipped questions; experiences cannot be docu-mented when respondents decline to complete survey items. It is possible, however, that those who declined to answer questions about gender and dat-ing were less likely to experience (or to report experiencing) sexual violence. Respondents with missing values for control variables (race, age, prior

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

10 Journal of Interpersonal Violence

antiviolence programming) were retained for consistency in sample size. Separate categories were created to substitute for missing values (e.g., “race undisclosed”). Overall sample characteristics are included in Table 1.

The final sample was 46.5% male and 53.5% female. Approximately half (50.3%) of the respondents reported having been in a dating relationship that lasted longer than 1 week. Regarding race, 13.4% of respondents identified themselves as White, 25.5% as Black, 13.3% as Asian, 10.3% as Other/mul-tiracial, and 37.5% skipped this question or declined to disclose their race. Just over one third of respondents (36.3%) were 11 or younger, 51.2% were 12 years old, and 11.7% were 13 or older. Just over one fifth (21.3%) reported having attended some form of antiviolence programming in the past. Rates of reported sexual violence were as follows: 38.1% for overall victimization, 20.6% for victimization by male peers, 21.4% for victimization by female peers, 16.7% for overall perpetration, 8.6% for perpetration against male peers, and 9.0% for perpetration against female peers.

To investigate potential differences between the initial and final sam-ples, univariate analyses of all dependent variables were run with the initial sample of 2,655 (not shown). Results were as follows: 30.7% for overall victimization, 16.3% for victimization by male peers, 17.6% for victimiza-tion by female peers, 13.4% for overall perpetration, 7.0% for perpetration against male peers, and 7.3% for perpetration against female peers. This indicates a decline of approximately 2% to 4% points across all measures with the exception of overall victimization, which declined by approxi-mately 7.5% points.

Three logistic regressions were conducted for each dependent variable. The first models were bivariate, with gender as the sole independent variable. The second models added dating experience and a Gender × Dating interac-tion term. The final models incorporated the control variables. Results for logistic regression analyses appear in Tables 2 to 7.

Results

Bivariate Analyses

Bivariate logistic regressions were performed for each dependent variable, using sex as an independent variable. Each of these was significant (p < .01). Girls were more likely to report victimization by male peers (odds ratio [OR] = 0.34) and perpetration against male peers (OR = 0.56); boys were signifi-cantly more likely to report overall victimization (OR = 1.43), victimization by female peers (OR = 5.00), overall perpetration (OR = 1.64), and perpetra-tion against female peers (OR = 5.83).

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 11

Multivariate Analyses

Each dependent variable was further assessed through two multivariate mod-els: the first included sex, dating experience, and an interaction between these predictors as independent variables; the final model included age, race, and prior antiviolence programming as controls. Sex remained a significant pre-dictor for four dependent variables. Boys were more likely to report overall victimization, victimization by female peers, overall perpetration, and perpe-tration against female peers (p < .01). These relationships persisted with the addition of controls. Dating increased the likelihood of reporting on all but

Table 1. Frequencies of Prior Dating Experience, Race, Age, Prior Antiviolence Programming, and Prior Sexual Violence by Sex.

All Respondents Male Students Female Students

Sex Male 46.46 Female 53.54 Dating experience 50.33 55.89 45.5Race White 13.42 12.24 14.44 Black 25.53 28.57 22.89 Asian 13.27 13.5 13.08 Other/multiracial (missing) 10.28 (37.49) 9.26 (36.42) 11.17 (38.42)Age 11 or younger 36.32 35.48 37.06 12 years old 51.20 49.14 53.00 13 or older (missing) 11.67 (0.80) 14.44 (0.94) 9.26 (0.68)Prior antiviolence

programming (missing)21.30 (4.45) 19.94 (5.49) 22.48 (3.54)

Victimization Any 38.15 42.70 34.20 By male peer 20.64 11.77 28.34 By female peer 21.37 34.85 9.67Perpetration Any 16.70 20.41 13.49 Against male peer 8.61 6.28 10.63 Against female peer 9.04 15.86 3.13n 1,371 637 734

Note. Respondents with missing values for gender, dating, victimization, or perpetration were excluded.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

12 Journal of Interpersonal Violence

Table 2. Logistic Regression Models of Reported Victimization.

Model 1 Model 2 Model 3

OR SE OR SE OR SE

Gender (male) 1.43** 0.16 1.71** 0.33 1.72** 0.33Prior dating experience 6.10** 1.06 5.03** 0.9Gender × Dating 0.61* 0.15 0.56* 0.14Race (White excluded) Black 2.77** 0.59 Asian 0.98 0.26 Multiracial/Other 1.18 0.31Age (11 or younger excluded) 12 years old 1.19 0.16 13 or older 2.02** 0.41Prior antiviolence program 1.18 0.18Likelihood ratio χ2 (df) 10.45 (1) 193.87 (3) 246.30 (12) Log likelihood −906.19 −814.48 −788.26 Pseudo R2 .01 .11 .14 N 1,371 1,371 1,371

Note. OR = odds ratio.*p < .05. **p < .01.

Table 3. Logistic Regression Models of Reported Victimization by Male Peers.

Model 1 Model 2 Model 3

OR SE OR SE OR SE

Gender (male) 0.34** 0.05 1.05 0.24 1.04 0.24Prior dating experience 5.87** 1.08 5.24** 1.01Gender × Dating 0.13** 0.04 0.12** 0.04Race (White excluded) Black 1.90* 0.50 Asian 1.18 0.39 Multiracial/Other 1.61 0.50Age (11 or younger excluded) 12 years old 1.02 0.17 13 or older 1.3 0.31Prior antiviolence

program1.27 0.22

Likelihood ratio χ2 (df) 59.36 (1) 165.25 (3) 177.81 (12) Log likelihood −668.40 −615.45 −609.17 Pseudo R2 .04 .12 .13 N 1,371 1,371 1,371

Note. OR = odds ratio.*p < .05. **p < .01.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 13

Table 4. Logistic Regression Models of Reported Victimization by Female Peers.

Model 1 Model 2 Model 3

OR SE OR SE OR SE

Gender (male) 5.00** 0.75 3.00** 0.84 3.02** 0.86Prior dating experience 3.18** 0.86 2.44** 0.68Gender × Dating 1.98* 0.67 1.97* 0.67Race (White excluded) Black 2.48** 0.65 Asian 0.71 0.26 Multiracial/Other 0.82 0.28Age (11 or younger excluded) 12 years old 1.25 0.21 13 or older 2.08** 0.48Prior antiviolence program

1.13 0.21

Likelihood ratioχ2 (df) 132.39 (1) 252.66 (3) 294.79 (12) Log likelihood −645.13 −584.99 −563.93 Pseudo R2 .09 .18 .21 N 1,371 1,371 1,371

Note. OR = odds ratio.*p < .05. **p < .01.

Table 5. Logistic Regression Models of Reported Perpetration.

Model 1 Model 2 Model 3

OR SE OR SE OR SE

Gender (male) 1.64** 0.24 2.35** 0.73 2.42** 0.77Prior dating experience 6.79** 1.85 5.39** 1.51Gender × Dating 0.53† 0.19 0.49* 0.18Race (White excluded) Black 2.87** 0.82 Asian 0.86 0.35 Multiracial/Other 1.51 0.52Age (11 or younger excluded) 12 years old 1.12 0.20 13 or older 1.49† 0.35Prior antiviolence program 1.67** 0.30Likelihood ratioχ2 (df) 11.72 (1) 111.44 (3) 151.49 (12) Log Likelihood −612.66 −562.81 −542.78 Pseudo R2 .01 .09 .12 N 1,371 1,371 1,371

Note. OR = odds ratio.†p < .10. *p < .05. **p < .01.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

14 Journal of Interpersonal Violence

Table 6. Logistic Regression Models of Reported Perpetration Against Male Peers.

  Model 1 Model 2 Model 3

OR SE OR SE OR SE

Gender (male) 0.56** 0.11 1.58 0.67 1.61 0.69Prior dating experience 8.87** 2.97 8.26** 2.85Gender × Dating 0.22** 0.11 0.21** 0.10Race (White excluded) Black 2.38* 0.93 Asian 1.70 0.86 Multiracial/Other 2.04 0.91Age (11 or younger excluded) 12 years old 1.01 0.23 13 or older 0.89 0.29Prior antiviolence program 1.62* 0.36Likelihood ratioχ2 (df) 8.37 (1) 73.49 (3) 86,31 (12) Log likelihood −397.99 −365.43 −359.02 Pseudo R2 .01 .09 .11 N 1,371 1,371 1,371

Note. OR = odds ratio.*p < .05. **p < .01.

Table 7. Logistic Regression Models of Reported Perpetration Against Female Peers.

Model 1 Model 2 Model 3

OR SE OR SE OR SE

Gender (male) 5.83** 1.39 3.39** 1.56 3.52** 1.63Prior dating experience 2.82* 1.30 1.95 0.91Gender × Dating 1.84 1.00 1.73 0.95Race (White excluded) Black 3.02** 1.17 Asian 0.27 0.22 Multiracial/Other 1.05 0.53Age (11 or younger excluded) 12 years old 1.16 0.28 13 or older 1.86* 0.55Prior antiviolence program 1.52† 0.37Likelihood ratioχ2 (df) 70.73 (1) 119.28 (3) 157.51 (12) Log likelihood −380.82 −356.55 −337.43 Pseudo R2 .09 .14 .19 N 1,371 1,371 1,371

Note. OR = odds ratio.†p < .10. *p < .05. **p < .01.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 15

one of the dependent variables (p < .01); the exception was perpetration against female peers, for which the initial effects of dating experience (OR = 2.82, p < .05) disappeared in the final model. The interaction of gender and dating experience reached or approached significance for all dependent vari-ables except for perpetration against female peers; this persisted with the addition of controls. Dating experience increased the likelihood that girls would report overall victimization (OR = .56, p < .05), victimization by male peers (OR = .12, p < .05), overall perpetration (OR = .49, p < .05), and per-petration against males (OR = .21, p < .01; all values are for the final model). Dating increased the likelihood that boys would report victimization by female peers (OR = 1.97, p < .05 in the final model).

Control variables. Age, race, and prior antiviolence programming all had sig-nificant effects in the final models, although these varied considerably across dependent variables. Black students were significantly more likely than Whites to report on all dependent variables (p < .05, OR ranging from 1.90 for victimization by male peers to 3.02 for perpetration against female peers). Students who identified their race as multiracial or “Other” did not differ significantly from Whites. Relative to students aged 11 or younger, those aged 13 and older were more likely to report overall victimization (OR = 2.02, p < .01), victimization by female peers (OR = 2.08, p < .01), perpetra-tion against female peers (OR = 1.86, p < .05), and were marginally more likely to report overall perpetration (OR = 1.49, p < .10). Prior antiviolence programming had no effects on any of the victimization variables. However, such programming significantly increased the likelihood that students would report overall perpetration (OR = 1.67, p < .01) and perpetration against male peers (OR = 1.62, p < .05), and marginally increased the likelihood of report-ing perpetration against female peers (OR = 1.52, p < .10).

Discussion

Students reported comparable levels of victimization by male and female peers (20.6% and 21.4%, respectively), and comparable levels of perpetra-tion against male and female peers (8.6% and 9.0%, respectively). These pat-terns were further reflected, albeit with somewhat lower rates, in the larger original sample that included those who had declined to disclose sex and/or dating experience. These findings contradict what we would have predicted based on Brownmiller’s feminist model of rape. Yet this is not to say that gender was unimportant. This research revealed a range of gendered patterns in sexual violence among middle school students in New York City. Gender was significant in all bivariate models. Boys were more likely to report

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

16 Journal of Interpersonal Violence

overall victimization, overall perpetration, victimization by female peers, and perpetration against female peers. These effects persisted when controlling for dating experience, age, race, and prior antiviolence programming. Girls were more likely to report victimization by and perpetration against male peers, although these effects disappeared when controlling for dating experi-ence. These findings indicate that much, but not all, of sexual violence in middle school occurs between different-sex peers. They reveal patterns of male aggression, female aggression, male victimization, and female victim-ization within this population.

Dating emerged as a strong predictor of sexual violence, and the effects of dating were different for girls and boys. This is particularly striking, given that survey questions asked about overall experiences, rather than experi-ences within dating relationships (see the appendix). Dating increased stu-dents’ likelihood of reporting on all dependent variables, although the effect on perpetration against female peers disappeared in the final model. The risks associated with dating were heightened for girls in regard to overall victim-ization, victimization by male peers, overall perpetration, and perpetration against male peers. The risks were heightened for boys in regard to victimiza-tion by female peers.

These findings raise questions regarding female perpetration. Respondents reported comparable levels of victimization by male and female peers. However, female students were far less likely than males to report perpetration (13.5%, compared with 20.4%). How should this be interpreted? It may be that female perpetrators are uniquely aggressive, and are assaulting more peers than their male counterparts. An alternative explanation points to students’ classification of their own violent behav-iors. It may be that sexually aggressive male and female students are engaging in similar levels of perpetration, but that boys are more likely to recognize if and when they have violated others’ boundaries. Antiviolence activists, educators, and scholars often emphasize boys’ aggression toward girls in their work (see Men Can Stop Rape, n.d.; Morrison et al., 2004). Exposure to these messages may encourage boys to recognize that they are capable of committing violent acts, and girls to recognize that they are capable of being violated. At the same time, boys may not learn to view themselves as potential victims, and girls may not learn to view them-selves as potential perpetrators.

Similar issues arise regarding the impact of antiviolence programming. This variable had no impact on measures of reported victimization. Yet it consistently increased students’ likelihood of reporting perpetration. This may indicate that antiviolence programming increases the likelihood that middle school students will commit sexual violence. Alternatively, it may

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 17

indicate that antiviolence programming increases the likelihood that students will recognize and/or report whether they have perpetrated. It is also possible that those who have perpetrated are more likely to have been compelled to attend antiviolence trainings.

Data further highlight the significance of race and age. Black middle school students were more likely than Whites to report on all measures for victimization and perpetration. This is consistent with previous research on the relative prevalence of sexual violence among adults, and the need to consider both race and gender in prevention work (Eschholz & Vieraitis, 2004; White, 1999). Curricula that address gender inequality and promote anti-sexist perspectives might be expanded to address racial inequality and promote anti-racist perspectives. Visual aids, such as posters and films, should feature people of multiple races; given the findings of this study, it seems particularly important that Black students are represented. In addition, relative to those aged 11 and younger, teenagers were more likely to report on four of six dependent variables. This suggests a need for primary prevention with younger populations (Morrison et al., 2004). Unwanted sexual contact, already a problem by middle school, may grow increasingly common as students age and begin (or continue) to pursue romantic and sexual relationships. Effective prevention with elementary school students, or first-year middle school students, might mitigate this risk.

These findings have implications for prevention. If we wish to effec-tively address sexual violence among middle school students, we must work with a realistic understanding of the nature and scope of the problem. This means that, although gender should remain a key component of anti-violence programming, gendered violence should be recognized as more complex than Brownmiller’s feminist model suggests. We must address violence among different-sex as well as same-sex peers. We must address male aggression, female aggression, male victimization, and female victim-ization. Although this data set did not allow for investigations of sexual violence against transgender or gender variant students, we must incorpo-rate the needs of this population as well (see Stotzer, 2009; Clements-Nolle, Marx, & Katz, 2006). Furthermore, we must not limit programming to mat-ters of gender and sexuality, but also incorporate race and age. Finally, given the significance of dating experience as a predictor of sexual vio-lence, we may want to work more closely with communication and sexual script-based models in developing prevention programs for adolescents. These data indicate that it may be appropriate to use comprehensive curri-cula that address unwanted sexual contact in general, along with sexual violence in dating situations.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

18 Journal of Interpersonal Violence

Limitations and Future Research

This study had several limitations, many of which may be addressed in future research. As is often the case in this field, this study may have under-stated the prevalence of sexual violence among middle school students. Underreporting was likely an issue. This concern was heightened by some of the survey questions used. Of the eight items used to construct victimiza-tion variables, six referred specifically to sexual violence on school grounds. All eight questions allowed students to report sexually violent experiences with a male or female peer; it was not possible to report having assaulted or been assaulted by both male and female peers. This means that this study could not adequately investigate the scope of sexual violence off of school grounds, and that survey design may have obscured some data regarding gender-specific violent encounters. In addition, several key variables were absent from this analysis due to limitations in the data set. Future research should explore the effects of attitudes toward sexual violence and rape myths, attitudes toward gender roles, overall sexual experience, all dating relationships (including those lasting less than 1 week), sexual orientation, and gender identity and expression on sexual violence among middle school students.

In spite of these limitations, this study demonstrated the need for greater attention to youth in sexual violence research. Theoretical models, particu-larly those that draw on sexual script theory, should be expanded or modified to account for the unique experiences and concerns of adolescents. Future efforts to identify the prevalence of sexual violence in the United States should include adolescents. Finally, further research should address the needs and experiences of diverse adolescent populations. These data were from middle school students in New York City. Patterns found here may reflect or diverge from those in other cities, in suburban and rural communities, in other regions of the United States, in communities with different class and racial compositions, and in communities with different levels of exposure to antiviolence programming.

Appendix

Constructing Dependent Variables

The following tables include all survey questions used to develop victimiza-tion and perpetration variables, as well as bivariate relationships by gender.a

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 19

Reported Sexual Violence Victimization by Sex.

All Respondents Male Students Female Students

Has any girl or boy ever done any of the following to you at school or during a school-sponsored activity when you did not want them to? Touched, grabbed,

or pinched you in a sexual way?

17.21 (5.18 missing) 17.43 (4.08 missing)

17.03 (6.13 missing)

Intentionally brushed up against you in a sexual way?

18.45 (3.50 missing) 19.62 (2.98 missing)

17.44 (3.95 missing)

Pulled at your clothing in a sexual way?**

10.28 (3.79 missing) 13.34 (3.30 missing)

7.63 (4.22 missing)

Pulled your clothing off or down?**

5.40 (3.94 missing) 6.91 (3.61 missing)

4.09 (4.22 missing)

Made you kiss him or her?**

16.34 (4.38 missing) 20.25 (3.45 missing)

12.94 (5.18 missing)

Made you do something sexual, other than kissing?**

6.78 (3.36 missing) 9.58 (3.30 missing)

4.36 (3.41 missing)

Have your male/female peers ever pushed, grabbed, shoved, or kicked you in your private parts?**

13.71 (6.35 missing) 20.09 (6.75 missing)

8.17 (5.99 missing)

Have your male/female peers ever made you touch their private parts or touched yours when you did not want them to?

4.74 (4.23 missing) 5.49 (4.71 missing)

4.09 (3.81 missing)

N 1,371 1,371 1,371

Reported Sexual Violence Perpetration by Sex.

All Respondents Male Students Female Students

Have you ever done any of the following to a girl or a boy at school or during a school-sponsored activity when they did not want you to?

Touched, grabbed, or pinched them in a sexual way?**

5.47 (3.72 missing) 7.54 (3.77 missing)

3.68 (3.68 missing)

(continued)

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

20 Journal of Interpersonal Violence

All Respondents Male Students Female Students

Intentionally brushed up against you in a sexual way?**

5.18 (3.94 missing) 7.54 (4.24 missing)

3.13 (3.68 missing)

Pulled at their clothing in a sexual way?**

3.57 (3.65 missing) 5.34 (3.30 missing)

2.04 (3.95 missing)

Pulled their clothing off or down?

2.19 (3.87 missing) 2.51 (3.77 missing)

1.91 (3.95 missing)

Made them kiss him or her?**

5.84 (4.81 missing) 9.89 (4.24 missing)

2.23 (5.31 missing)

Made them do something sexual, other than kissing?**

2.26 (4.60 missing) 3.77 (5.34 missing)

0.95 (3.95 missing)

Thinking about your male/female peers, have you ever pushed, shoved, or kicked them in their private parts?

6.86 (5.11 missing) 6.59 (4.55 missing)

7.08 (5.59 missing)

Thinking about your male/female peers, have you ever made them touch your private parts or touched theirs when they did not want you to?**

1.82 (4.38 missing) 2.98 (3.61 missing)

0.82 (5.04 missing)

N 1,371 1,371 1,371

aPossible answers included “no,” “yes, male,” and “yes, female” for all questions. Respondents who answered “yes, male” or “yes, female” to an individual question were marked as reporting that aspect of sexual violence.**p < .01 in logistic regression by sex.

Appendix (continued)

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publi-cation of this article.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 21

Note

1. The terms survivor and victim will be used interchangeably to refer to people who have experienced sexual violence.

References

American College Health Association. (2013). National College Health Assessment II: Reference group data report, spring 2013. Hanover, MD: Author.

Anderson, V. N., Simpson-Taylor, D., & Herrmann, D. J. (2004). Gender, age, and rape-supportive rules. Sex Roles, 50, 77-90.

Armstrong, E. A., Hamilton, L., & Sweeney, B. (2006). Sexual assault on campus: A multilevel, integrative approach to party rape. Social Problems, 53, 483-499.

Black, M. C., Basile, K. C., Breiding, M. J., Smith, S. G., Walters, M. L., Merrick, M. T., . . . Stevens, M. R. (2011). The National Intimate Partner and Sexual Violence Survey: 2010 summary report. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.

Bohner, G., Jarvis, C., Eyssel, F., & Siebler, F. (2005). The causal impact of rape myth acceptance on men’s rape proclivity: Comparing sexually coercive and noncoercive men. European Journal of Social Psychology, 35, 819-828.

Bohner, G., Siebler, F., & Schmelcher, J. (2006). Social norms and the likelihood of raping: Perceived rape myth acceptance of others affects men’s rape proclivity. Personality and Social Psychology Bulletin, 32, 286-297.

Brownmiller, S. (1975). Against our will: Men, women, and rape. New York, NY: Ballantine Books.

Calmes, J. (2014, January). Obama seeks to raise awareness of rape on campus. New York Times. Retrieved from http://www.nytimes.com/2014/01/23/us/politics/obama-to-create-task-force-on-campus-sexual-assaults.html?action=click&;module=Search&region=searchResults%230&version=&url=http%3A%2F%2Fquery.nytimes.com%2Fsearch%2Fsitesearch%2F%3Faction%3Dclick%26region%3DMasthead%26pgtype%3DHomepage%26module%3DSearchSubmit%26contentCollection%3DHomepage%26t%3Dqry322%23%2FVAWA%2Fsince1851%2Fallresults%2F1%2Fallauthors%2Fnewest%2F

Centers for Disease Control and Prevention. (2014). Adverse Childhood Experiences Study: Prevalence. Retrieved from http://www.cdc.gov/violenceprevention/aces-tudy/prevalence.html

Clements-Nolle, K., Marx, R., & Katz, M. (2006). Attempted suicide among trans-gender persons. Journal of Homosexuality, 51, 53-69.

De Rosa, C. J., Ethier, K. A., Kim, D. H., Cumberland, W. G., Afifif, A. A., Kotlerman, J., . . . Kerndt, P. R. (2010). Sexual intercourse and oral sex among public middle school students: Prevalence and correlates. Perspectives on Sexual and Reproductive Health, 4, 197-205.

Dodge, J. (2014, April). Study: Middle school sex harassment widespread, goes unnoticed at school. CBS Chicago. Retrieved from http://chicago.cbslocal.com/2014/04/07/study-middle-school-sex-harassment-widespread-goes-unno-ticed-at-school

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

22 Journal of Interpersonal Violence

Eaton, D. K., Kann, L., Kinchen, S., Shanklin, S., Flint, K. H., Hawkins, J., . . . Wechsler, H. (2012). Youth risk behavior surveillance—United States, 2011. Morbidity and Mortality Weekly Report, 61, 1-162.

Eschholz, S., & Vieraitis, L. M. (2004). Race-specific gender equality and rape: A further test of the feminist hypothesis. Critical Criminology, 12, 195-219.

Federal Bureau of Investigation. (2012). UCR program changes definition of rape: Includes all victims and omits requirement of physical force. Retrieved from http://www.fbi.gov/about-us/cjis/cjis-link/march-2012/ucr-program-changes-definition-of-rape

Jhally, S. (Producer and Editor). (1994). The date rape backlash [Motion picture]. United States: Media Education Foundation.

Johnson, C. F. (2004). Child sexual abuse. The Lancet, 364, 462-470.Johnson, K. (2012, January). FBI changes definition of rape to include men as vic-

tims. USA Today. Retrieved from http://usatoday30.usatoday.com/news/nation/story/2012-01-06/fbi-rape-definition-adds-men/52398350/1

Klein, R. (2014, April). Sexual violence among students is a significant problem as early as middle school, says study. Huffington Post. Retrieved from http://www.huffingtonpost.com/2014/04/06/sexual-violence-middle-school_n_5101226.html

Lonsway, K. A., & Fitzgerald, L. F. (1994). Rape myths: In review. Psychology of Women Quarterly, 18, 133-164.

Men Can Stop Rape. (n.d.). What men can do. Retrieved from http://www.mencansto-prape.org/images/stories/PDF/Handout_pdfs/what-men-can-do-final.pdf

Morrison, S., Hardison, J., Matthew, A., & O’Neil, J. (2004). An evidence-based review of sexual assault preventive intervention programs. Research Triangle Park, NC: RTI International.

Muehlenhard, C. L. (2011). Examining stereotypes about token resistance to sex. Psychology of Women Quarterly, 35, 676-683.

Pérez-Fuentez, G., Olfson, M., Villegas, L., Morcillo, C., Wang, S., & Blanco, C. (2013). Prevalence and correlates of child sexual abuse: A national study. Comprehensive Psychiatry, 54, 16-27.

Peterson, Z. D., & Muehlenhard, C. L. (2004). Was it rape? The function of women’s rape myth acceptance and definitions of sex in labeling their own experiences. Sex Roles, 51, 129-144.

Peterson, Z. D., & Muehlenhard, C. L. (2007). Conceptualizing the “wantedness” of women’s consensual and nonconsensual sexual experiences: Implications for how women label their experiences with rape. Journal of Sex Research, 44, 72-88.

Rhinehart, S., Doshi, N., & Espelage, D. (n.d.). Sexual harassment and sexual vio-lence experiences among middle school youth. Retrieved from http://www.aera.net/Portals/38/Newsroom%20-%20Recent%20Research/Sexual%20Harassment%20and%20Sexual%20Violence%20Experiences%20Among%20Middle%20School%20Youth.pdf

Rothman, E., & Silverman, J. (2007). The effect of a college sexual assault prevention program on first-year students’ victimization rates. Journal of American College Health, 55, 283-290.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

Levine 23

Ryan, K. M. (2011). The relationship between rape myths and sexual scripts: The social construction of rape. Sex Roles, 65, 774-782.

Sexual Trauma Services of the Midlands. (n.d.). Youth Violence Prevention Program (YVPP)—For middle and high school youth. Retrieved from http://www.stsm.org/education-amp-outreach/youth-middle-and-high-school-violence-preven-tion-program

Simon, W., & Gagnon, J. (1986). Sexual scripts: Permanence and change. Archives of Sexual Behavior, 15, 97-120.

Simon, W., & Gagnon, J. (2005). Sexual conduct (2nd ed.). New Brunswick, NJ: AldineTransaction. (Original work published 1973)

Sprecher, S., Hatfield, E., Cortese, A., Potapova, E., & Levitskaya, A. (1994). Token resistance to sexual intercourse and consent to unwanted sexual intercourse: College students’ dating experiences in three countries. The Journal of Sex Research, 31, 125-132.

Stinson, R. D. (2010). Hooking up in young adulthood: A review of factors influ-encing the sexual behavior of college students. Journal of College Student Psychotherapy, 24, 98-115.

Stotzer, R. L. (2009). Violence against transgender people: A review of United States data. Aggression and Violent Behavior, 14, 170-179.

Taylor, B., Stein, N. D., Mack, A. R., Horwood, T. J., & Burden, F. (2008). Experimental evaluation of gender violence/harassment prevention programs in middle schools. Washington DC: National Institute of Justice.

Taylor, B., Stein, N. D., Woods, D., & Mumford, E. (2012). Experimental evaluation of a youth dating violence prevention program in New York City Middle Schools, 2009-2010. Ann Arbor, MI: Inter-University Consortium for Political and Social Research [distributor], 2012-05-18. doi:10.3886/ICPSR32901.v1

Tjaden, P., & Thoennes, N. (2000). Full report of the prevalence, incidence, and consequences of violence against women. Washington, DC: Office of Justice Programs, U.S. Department of Justice.

U.S. Department of Health & Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau. (2012). Child maltreatment 2011. Washington, DC: U.S. Department of Health & Human Services. Retrieved from http://www.acf.hhs.gov/programs/cb/research-data-technology/statistics-research/child-maltreatment

Wasco, S. (2012). Assessing campus readiness for prevention. Enola: Pennsylvania Coalition Against Rape.

Whaley, R. B. (2001). The paradoxical relationship between gender inequality and rape: Toward a refined theory. Gender & Society, 15, 531-555.

White, A. M. (1999). Talking feminist, talking black: Micromobilization processes in a collective protest against rape. Gender & Society, 13, 77-100.

Whitman, C. (2012). Rape and sexual assault analyses and laws: Current as of July 2012. Washington, DC: AEquitas: The Prosecutors’ Resource on Violence Against Women. Retrieved from http://www.ncdsv.org/images/AEquitas_RapeAndSexualAssaultAnalysesAndLaws_7-2012.pdf

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from

24 Journal of Interpersonal Violence

Yodanis, C. L. (2004). Gender inequality, violence against women, and fear: A cross-national test of the feminist theory of violence against women. Journal of Interpersonal Violence, 19, 655-675.

Author Biography

Ethan Levine received his master’s in sociology from the New School for Social Research in 2010. He is currently a PhD student in sociology at Temple University, where his research focuses on gender, sexuality, and social policy. He has previously published work concerning the capacity for United Nations policy to advance intersex rights. Outside of academia, he has a combined 8 years’ experience working with survivors of sexual and domestic violence.

at TEMPLE UNIV on June 25, 2015jiv.sagepub.comDownloaded from