Interpersonal Violence Victimization Experiences of Middle School Youth: An Exploration by Gender and Sexual/Romantic Attraction

Abstract

Sexual minority youth (SMY) are at increased risk for interpersonal violence victimization compared to heterosexual youth. The current study examined how self-reported victimization (i.e., bullying, sexual harassment, dating violence) among middle school youth varied as a function of sexual/romantic attraction as well as gender identity. Cross-sectional data were gathered from students at seven middle schools in New England (n = 2245). Mean comparisons with post-hoc Tukey tests determined differences in rates of past six-month and lifetime interpersonal violence victimization by sexual/romantic attraction and the intersection of gender and attraction. As hypothesized, interpersonal violence victimization among middle school youth differed as a function of sexual/romantic attraction as well as gender. To date, most research has focused on older samples, particularly high school youth and young adults. These data are consistent with these prior studies documenting increased risk for interpersonal violence victimization among youth who indicate same-gender attraction but add to the literature in demonstrating the expansive forms of peer victimization that same-gender-attracted youth already experience by early adolescence. Given that victimization is associated acutely and longitudinally with many deleterious outcomes, including poorer mental health and increased risk for subsequent victimization, greater structural supports are needed for early adolescent SMY.

Introduction

Despite positive cultural shifts indicating greater support and legal protection for sexual minority youth (SMY; individuals who are attracted to or have sexual contact with people of the same gender and/or those whose gender identity or expression is different from their sex assigned at birth; ^{CDC, 2019}), population-based studies continue to reveal that SMY report significantly poorer mental health than heterosexual youth (Lucassen et al., 2017; Marshal et al., 2011). A growing percentage of youth report a sexual minority identity (e.g., lesbian, gay, bisexual, or other) compared to prior generations (Jones, 2021), and these youth have access to more supportive school systems than in prior generations, including sexuality-specific supports such as Gender-Sexuality Alliances (GSAs; Fetner & Kush, 2008) as well as legal protection from discrimination in schools (U.S. Department of Education, 2021). Yet greater visibility and school supports are not associated with global improvements in well-being, reflecting the myriad of stressors SMY face by a young age. In line with the minority stress model (Hatzenbuehler, 2009; Meyer, 1995), SMY continue to face markedly high rates of victimization (Mereish et al., 2021; Meyer et al., 2021). In research comparing rates of victimization and mental health across generational cohorts, Meyer and colleagues (2021) found that younger cohorts report similarly high levels of victimization, but poorer mental health compared to older generational cohorts (e.g., individuals born in 1990–1997 compared to those born in 1956–63 and 1974–81). A recent daily diary study of 12–18-year-old SMY demonstrated the markedly high rates of stigma they face. In just a 21-day window, 67% of SMY reported that someone had made them feel uncomfortable or unsafe due to their identity (Mereish et al., 2021).

Interpersonal violence victimization can be physical, sexual, or psychological in nature, and involves the intentional use of physical force or power against another person (Dahlberg & Krug, 2002). Among youth, interpersonal violence victimization forms include bullying, dating violence, and sexual violence (Kann et al., 2014; Pearlman, 2016). Adolescent interpersonal violence victimization, a form of adverse childhood experiences, is understood to be, in part, the product of cultural norms that promote rigid adherence to traditional notions of masculinity (Katz et al., 2011; Miller, 2018). Individuals who deviate from heteronormative or rigid gender norms may be the targets of peer aggression, which may explain why SMY are more likely to experience interpersonal violence victimization across contexts and levels of the social-ecological model (e.g., home, school), different types of violence (e.g., bullying, sexual victimization), and general and bias-based harassment due to their perceived sexual orientation or gender (Lessard et al., 2020; Norris & Orchowski, 2020; Sterzing et al., 2017). Elucidating the scope of interpersonal violence is important for contextualizing the many health disparities SMY face (e.g., minority stress; Hatzenbuehler, 2009^; Meyer, 2003). In particular, experiences of interpersonal violence among SMY have been linked to increased rates of sexual risk behaviors (Robinson & Espelage, 2013), subsequently poorer mental health (Mittleman, 2019), and elevated risk of experiencing other forms of violence (Martin-Storey et al., 2021).

Middle school is an important time to assess peer victimization among SMY, and yet most peer victimization research has focused on the experiences of high school students exclusively or collapsed the experiences of middle and high school students (Dank et al., 2014; Norris & Orchowski, 2020; Poteat & Espelage, 2005; Sterzing et al., 2017; Zaza et al., 2016). This focus on older SMY reflects a broader emphasis on older adolescents in victimization research. In a meta-analysis of school-based victimization among middle and high school students, the overwhelming majority of studies had a mean age in the high school range (Toomey & Russell, 2016). Specifically, only one out of 24 studies had a mean age under 14.5 (Toomey & Russell, 2016). Many forms of peer interpersonal violence peak (e.g., bullying, bias-based harassment) or initiate (e.g., dating violence) in middle school (e.g., Espelage et al., 2012^; Mittleman, 2019). For example, bullying is more prevalent during middle school than high school, such that bullying declines from 6^th grade through the end of high school (Hymel & Swearer, 2015; Menesini & Salmivalli, 2017). Understanding the scope of violence among middle school youth can play a vital role in understanding the timing and focus of prevention efforts.

However, the limited research that has examined sexual orientation-based disparities among younger samples demonstrates that peer victimization disparities appear much earlier than high school. In a population-based birth cohort study following youth from age 5 through ages 9 and 15, SMY report higher rates of bullying than heterosexual youth by age 9 (Mittleman, 2019). More than two-thirds of nine-year-olds reporting a same-sex attraction (68%) endorsed past-month bullying (Mittleman, 2019). Second, early adolescent SMY have less access to school supports afforded to SMY in high school. For example, in GLSEN’s 2019 National School Climate Survey (Kosciw et al., 2020), only 34% of middle school SMY, compared to 74% of high school SMY reported their school had a Gender-Sexuality Alliance (GSA). Structural factors are incredibly important in buffering the impacts of victimization on SMY. The presence of and engagement with GSAs is associated with lower levels of depression and anxiety among SMY (Walls et al., 2013), including prospectively (Poteat et al., 2020). Therefore, middle school SMY might be particularly vulnerable to peer victimization and its deleterious effects compared to their high school peers. This paper seeks to expand the understanding of the scope of peer victimization that middle school youth experience based on (a) their sexual and/or romantic attraction and (b) how sexual and/or romantic attraction intersects with gender identity.

Interpersonal Violence Victimization in Middle School: Gaps in the Knowledge Base

Although sexual orientation-based disparities in rates of bullying are already evident by middle school (e.g., Mittleman, 2019), much less is known about the broader scope of peer victimization that SMY in middle school experience. Most research on middle school peer victimization focuses on bullying, with bullying behavior and attitudes becoming more positive over middle school (Swearer & Cary, 2003). Research samples that collapse middle school and high school students demonstrates that SMY do experience substantially higher rates of numerous forms of interpersonal violence victimization than do heterosexual youth, including bullying (Gruber & Fineran, 2008), sexual harassment (Gruber & Fineran, 2008; Mitchell et al., 2014), and dating violence victimization (Dank et al., 2014). There is little research examining these disparities in samples of only middle school students. However, there is strong reason to suspect that the disparities observed by high school would extend to early adolescence as well.

Peer victimization in adolescence is rooted in society’s dominant values and expectations for heteronormative and gendered appearance, interactions, and romantic relationships (e.g., Birkett & Espelage, 2015^; Poteat et al., 2011). Difference, including in appearance, promotes peer victimization in middle school (Swearer & Cary, 2003). Further, in line with the social-ecological model of peer aggression (Espelage & Swearer, 2004), the experience of interpersonal violence victimization depends on the peer context, which varies by age and developmental stage (Espelage & Swearer, 2003; Swearer & Hymel, 2015). Peer salience and peer groups are central to the development of peer victimization in middle school (Birkett & Espelage, 2015; Merrin et al., 2018). Notably, many of these peer processes are pronounced in middle school, including peer influence and susceptibility (e.g., Chein et al., 2011^; Gardner & Steinberg, 2005), as well as the need for youth to negotiate social status, form new social connections, and navigate new types of relationships (e.g., mixed-gender friendships and romantic relationships: Connolly et al., 2004). For example, early adolescence is characterized by increasing mixed-gender peer affiliation from the predominantly same-gender affiliation of childhood, such that Grade 7 is the first grade associated with increases in both mixed-gender affiliation and dating activities (Connolly et al., 2004). Most students begin dating before high school (Connolly et al., 2004; Orpinas et al., 2013) and research demonstrates the presence of these forms of peer victimization do appear in these younger early adolescent samples (Goncy et al., 2018; Goncy et al., 2017; Lormand et al., 2013; Niolon et al., 2015). In a study of 7^th grade students from 10 Texas middle schools, researchers found that many students who had a dating relationship reported experiencing physical (21%) or nonphysical dating violence (48%; Lormand et al., 2013). Similarly, in a sample of more than 1,000 middle school students attending 3 urban public middle schools, 46% of students in the fall of 7^th grade endorsed at least one item of dating victimization in the past 3 months alone.

In line with the minority stress model, SMY in middle school might be vulnerable to marginalization and victimization if they do not conform to norms and expectations for peer relationships and interests (e.g., heteronormative) that are gaining power over early adolescence. Peer victimization can be one way, consciously or not, that youth reinforce perceived expectations of heteronormativity and the feminine/masculine dichotomy (Birkett & Espelage, 2015; Poteat et al., 2011). These gender roles, norms, and expectations become increasingly evident over the middle school period given that outward markers of the male/female binary and key pubertal development milestones coincide with middle school (e.g., average age of menarche is 12–13 years old and genital development among boys; Biro et al., 2012^; Sun et al., 2005). Given that bullying, homophobic teasing, and sexual harassment tend to co-evolve (Espelage et al., 2012; Merrin et al., 2018), middle school SMY might be more likely to experience the bullying that peaks in middle school, but also other forms of relational aggression (i.e., sexual harassment, dating violence) that become more pronounced in middle school.

The second gap in the understanding of interpersonal violence victimization among SMY is the conceptualization of SMY as one monolithic group across different sexual minority subgroups and genders (Birkett et al., 2009; Gruber & Fineran, 2008). Yet although all SMY are likely at increased risk for most forms of peer interpersonal violence victimization in middle school, risk might be particularly pronounced for particular SMY x gender subgroups. Regarding bullying, middle school is marked not only by high levels of general bullying, but also of sexual orientation-specific bullying, such as homophobic name-calling and teasing. Middle school students who engage in more bullying and homophobic teasing specifically are associated with increased popularity (Merrin et al., 2018). Among high school samples, all sexual minority subgroups are more likely to report homophobic bullying than heterosexual youth (Pollitt et al., 2018). Given that young boys are more likely to engage in and experience this type of bullying (Birkett & Espelage, 2015; Espelage et al., 2015; McMaster et al., 2002; Miller et al., 2020; Poteat & Espelage, 2005), and that this type of bullying is more common during middle school than high school, it might be that boys who are marked as different (e.g., sexual minority boys) are even more likely to be punished than girls with bullying during this developmental period.

To our knowledge, research with early adolescents has not examined sexual harassment and dating violence at the intersection of gender and sexual orientation. For example, in the studies on sexual harassment and dating violence victimization that at least include middle school students collapsed with high school students (Dank et al., 2014; Gruber & Fineran, 2008; Mitchell et al., 2014), most of these studies examined differences based on sexual orientation and differences based on gender (Dank et al., 2014; Gruber & Fineran, 2008), but not the intersection of gender and sexual orientation. These studies found that SMY report higher rates of sexual harassment (Gruber & Fineran, 2008) and all forms of dating violence with their most recent partner (physical, psychological, cyber, sexual; Dank et al., 2014) than heterosexual youth. Regarding gender, although boys and girls reported similar rates of sexual harassment (36% and 34%; Gruber & Fineran, 2008), female youth reported higher rates of some forms of dating violence (sexual coercion) whereas male youth reported higher rates of other forms (physical violence; Dank et al., 2014). However, the authors did not examine whether these effects held across both sexual orientation and gender.

Although we could not identify research with middle school youth at the intersection of sexual orientation and gender, some research has explored these nuances in older adolescent samples. In one study that compared sexual harassment among specific subgroups of SMY by gender, Mitchell et al. (2014) found that all sexual minority groups reported more sexual harassment than heterosexual boys. However, bisexual girls (66%), gay/queer boys (66%), and lesbian/queer girls (72%) were the most likely to report past-year sexual harassment compared to heterosexual boys (23%), heterosexual girls (43%), bisexual boys (50%), questioning boys (47%), and questioning girls (53%). These findings echo research with high school samples suggesting that both gay/lesbian and bisexual boys and girls report higher rates of sexual harassment and dating violence compared to heterosexual boys and girls (Martin-Storey, 2015; Norris & Orchowski, 2020), but sexual minority girls appear to have the highest rates of sexual harassment (Mitchell et al., 2014; Norris & Orchowski, 2020).

Finally, some work with high school students highlights that some forms of violence appear nuanced by gender and sexual orientation subgroup. For example, rates of particular forms of violence appear particularly pronounced among female SMY in high school, such as sexual violence and physical dating violence (Norris & Orchowski, 2020; Semprevivo, 2021; Whitton et al., 2019).

Purpose of the Current Study

The purpose of the current study was to contextualize the experiences of recent (past six months) and lifetime interpersonal violence victimization based on middle school youths’ sexual/romantic attraction, and further by how that varied by gender. To do so, it is necessary to contextualize differences by sexual/romantic attraction separately before exploring differences based on gender and sexual attraction combined. To date, few studies examine interpersonal violence victimization at the intersection of gender and sexual/romantic attraction in younger samples, with most research focusing on either gender or sexual orientation (Dank et al., 2014) and very few studying middle school youth specifically. By focusing on middle school students, we can help expand the literature on victimization among an age group that differs from others, namely high schoolers, in two important developmental ways: sexual and gender identity formation and violence behaviors. In terms of sexual and gender identity formation, this age likely marks the beginning of their identity formation and for many SMY youth as formation often continues into later ages (D’augelli & Grossman, 2001). Additionally, middle school holds important patterns for violence behaviors, such as the peak of some forms of peer aggression and the entry into romantic aggression. Therefore, studying violence experiences among SMY middle school students can provide important insight into later violence experiences that may occur.

Though our primary aim is to explore victimization at the intersection of gender and sexual/romantic attraction, we first will explore the main effects of sexual/romantic attraction in order to contextualize our findings.

Hypothesis 1:

Male and female youth who endorse attraction exclusively to the other gender will endorse lower rates of bullying, dating violence victimization, sexual harassment, and polyvictimization compared to youth reporting any degree of same-gender attraction (i.e., only same-gender attraction; mixed-gender attraction).

Hypothesis 2:

Boys who report any same-gender attraction will be most likely to be bullied by their peers in middle school.

Hypothesis 3:

Girls who report any same-gender attraction will be most likely to experience sexual harassment and will have a greater number of victimization experiences.

Materials and Methods

Participants

The present study uses baseline data from a large evaluation of a school-based sexual and gender-based violence prevention program for middle school youth (See NIH abstract: https://reporter.nih.gov/search/4GG3M0P910CfVHG4WXSI7Q/project-details/9535689 and clinical trials: https://clinicaltrials.gov/ct2/show/NCT03473067.). Sixth-, seventh-, and eighth-grade youth enrolled at seven middle schools in one New England state during the 2017–2018 school year were invited to enroll in the research. The state’s Department of Education did not permit the study team to assess participant race and ethnicity to minimize the potential for individual identification based on self-reported race and ethnicity. However, based on publicly available state-level data regarding student demographics within each school, representation of racial and ethnic minority students across schools in the sample ranged between 7% to 92%. Using these publicly available data, we estimated that approximately 35% of youth in the study identified as a racial or ethnic minority. In Year 1, the baseline assessment was completed by 3,017 participants. A series of questions were included in the study to ensure attentive responsiveness (i.e., “This question is to make sure the survey is working OK. Please pick the answer below that says CAT”). Youth who did not pass attention checks (N = 772) were excluded from the analyses.

Descriptive statistics for the study sample are presented in Table 1. The analytic sample included 1884 participants (46% boys, 52% girls, <1% who identified as transgender youth, and 2% listed “prefer not to answer”). The average age of the study sample was 12.22 years old (SD = 0.87). Just over one-third of all youth reported only being attracted to boys (38%) or only being attracted to girls (36%). Approximately 6% of youth reported attraction to both boys and girls, and another 6% of youth reported no attraction to either boys or girls. The final 13% of youth preferred not to answer to the question about attraction. Just over one-third of the sample (36%) reported having ever dated.

Table 1.

Univariate Statistics of Demographic Variables and Lifetime and Past 6 Month Interpersonal Violence Victimization among Middle School Youth (n = 1,884)

	n	Mean/n	SD/%
Age	1871	12.22	0.87
Boys	1884	862	45.8%
Girls	1884	982	52.1%
Transgender	1884	8	0.4%
Prefer not to Answer for Gender	1884	32	1.7%
Attracted to Boys	1884	716	38.0%
Attracted to Girls	1884	678	36.0%
Attracted to Boys and Girls	1884	117	6.2%
Not Attracted to Boys or Girls	1884	120	6.4%
Prefer not to Answer for Attraction	1884	253	13.4%
Ever Dated	1507	682	36.2%
Lifetime IVV
Bullying	1281	1070	83.5%
Sexual Harassment	1296	501	38.7%
Dating Violence	685	116	16.9%
Polyvictimization Score	1534	1.10	0.83
Past 6 months IVV
Bullying	1537	1128	73.4%
Sexual Harassment	1609	504	31.3%
Dating Violence	863	128	14.8%
Polyvictimization Score	1732	1.02	0.84

Note: SD = standard deviation; IVV = interpersonal violence victimization

Procedure

The school board as well as superintendent (or head of school) for each study site provided approval for the research. The study was also approved by the local Institutional Review Board as well as the state Department of Education. Only baseline data from Year 1 of the study was included in the current analyses. Schools enrolled in the research study were assigned to treatment and control conditions. The baseline survey was administered prior to the implementation of the intervention. As youth in this sample were minors (<18 years old), parents were provided an opportunity at the start of the school year to opt their child out of the study. Youth additionally provided verbal assent for study participation prior to the administration of the assessment. The survey was anonymous and was administered to students by trained research staff in their homeroom, health, or gym class. The survey was designed to be completed within one 50-minute academic class period and was administered via a paper and pencil questionnaire or via computer when possible. Students whose parents opted them out of the research, or who did not provide adolescent assent to participate, were provided with alternative activities to complete during the survey administration (less than 4%). To facilitate honest responding, students were provided with ample space to complete the questionnaire (e.g., far enough away that others could not see their responses) and reminded of the anonymous nature of the study.

Measures

Identity Variables

Gender Identity.

Gender identity is defined as an individual’s sense of their self as man/boy, woman/girl, transgender, or something else, and is closely related to concepts of gender role and gender expression (CDC, 2019). Gender identity was assessed with one question probing current gender identity, rather than a two-step approach for assessing gender identity (e.g., initial question probing sex assigned at birth; second question probing current gender identity). Youth were thus asked “What is your gender” and provided the following four response options: boy, girl, transgender, and prefer not to answer.

Sexual/Romantic Attraction.

Previous research with younger samples suggests that youth find questions about sexual attraction the clearest to answer (Austin et al., 2007). Accordingly, in this study, sexual/romantic attraction was measured with a single item. Youth were asked whether they would describe themselves as attracted to boys, attracted to girls, attracted to both boys and girls, attracted to neither boys or girls, or if they preferred not to answer. When we explore the intersection between sexual/romantic attraction and gender, youth who reported attraction to both boys and girls and youth who only had same-gender attraction (based on reported gender) were collapsed into a single category due to sample size.

Outcome Variables

Sexual Harassment Victimization.

Experiences of sexual harassment were assessed using a four-item modified version of the Shifting Boundaries Sexual Harassment Scale (Taylor et al., 2011). Youth were asked how many times in their lifetime and in the past six months another student had perpetrated harassing behaviors against them (e.g., “How many times has another student made sexual comments, jokes, gestures, or looks about/to you?”). Response options for six-month victimization experiences ranged from 0 times to 10 or more times and lifetime sexual harassment victimization was assessed with a dichotomous yes/no response. The six-month sexual harassment victimization was then dichotomized (0 = none, 1 = at least one experience) in order to make comparisons between lifetime (only dichotomous measures) and six-month violence rates.

Dating Violence Victimization.

Participants’ experiences of dating violence victimization were assessed using seven items adapted from Shifting Boundaries (Taylor et al., 2011). Items queried whether youth had been physically hurt or threatened by a dating partner in their lifetime or in the past 6 months (e.g., “How many times have people you have dated pushed, grabbed, shoved, or kicked you?”) as well as whether they had perpetrated these same behaviors against a dating partner. Respondents answered each item using the following 5-point scale: 0 (“0 times”), 1 (“1 to 3 times”), 2 (“4 to 6 times”), 3 (“7 to 9 times”), or 4 (“10 or more times”). Responses across all victimization items were examined to create a dichotomous variable representing exposure to dating violence victimization scores for both the past six-months and lifetime (0 = none, 1 = at least one experience).

Bullying Victimization.

Participants’ bullying experiences were assessed using a modified version of the School Crime Supplement (DeVoe & Kaffenberger, 2005) and the Illinois Bully Scale (Espelage & Holt, 2001). Youth were asked five questions about whether they had been bullied by another student in their lifetime or in the past six months (e.g., “How many times has another student made fun of you, called you names, or insulted you in a hurtful way?”) and whether they had bullied other youth. Respondents answered each item using the following 5-point scale: 0 (“0 times”), 1 (“1 to 3 times”), 2 (“4 to 6 times”), 3 (“7 to 9 times”), or 4 (“10 or more times”). Responses across all five victimization items, for both six-month and lifetime, were summed to create bullying victimization scores, which were then dummy coded (0 = none, 1 = at least one experience).

Polyvictimization.

Both a lifetime and six-month cumulative score were created based on the three types of interpersonal violence victimization measured above: sexual harassment, dating violence, and bullying. Participants scores ranged from zero (did not experience any victimization) to three (experienced all three forms of interpersonal violence victimization). This scale, ranging from 0–3, represents the number of interpersonal violence victimization forms experienced by youth either in the past six-months or in their lifetime.

Data Analysis

All analyses were conducted in Stata 17. We first present descriptive statistics to provide summaries about the sample. In line with our hypotheses, we first performed mean comparisons of interpersonal violence victimization by sexual/romantic attraction and then we examined how victimization experiences differ at the intersection of gender and sexual/romantic attraction for cisgender boys and girls. We limited our analyses to those who provided a response for both their gender identity and sexual/romantic attraction (n = 1884), but due to the gated nature of some of our violence variables (e.g., only those who reported dating were asked questions about dating violence) and missing data the sample sizes for each interpersonal violence victimization outcome comparison vary. Missing data on interpersonal violence victimization variables differed by gender identity, sexual/romantic attraction, and age. Specifically, those who prefer not to answer about their gender identity and sexual/romantic attraction were more likely to not answer questions about their interpersonal violence victimization experiences. Additionally, older students were more likely to be dating and therefore more likely to report about their dating violence experiences. Notably, in some cases we have a larger sample of students who reported six-month measures compared to lifetime measures, and this was likely due to the structure of the survey where lifetime measures were an added yes/no question beside each six-month measure. Mean comparisons using the pwmean command in Stata were employed, and comparisons between groups were done using the bonferroni correction. Results were considered significant at p ≤ 0.05.

Results

Experiences of Victimization by Sexual/Romantic Attraction

Youth attracted to both boys and girls reported the highest lifetime prevalence of bullying victimization (97%), significantly higher than those attracted to girls only (82%), those attracted to neither boys nor girls (73%), and those who preferred not to answer (80%) (see Table 2). Lifetime prevalence of sexual harassment victimization was significantly higher among youth reporting attraction to both girls and boys (62%) compared to youth only attracted to boys (45%), only attracted to girls (36%), attracted to neither boys nor girls (19%), and youth who preferred not to answer (22%). Among daters, youth attracted to both boys and girls reported both the highest lifetime prevalence dating violence (33%) and polyvictimization scores (1.48), while youth who are not attracted to boys nor girls or prefer not to answer reported similarly low rates of dating violence victimization and polyvictimization scores.

Table 2.

Lifetime and Past 6 Month Rates of Violence Victimization by Attraction

	n	Only Boysa	Only Girlsb	Both Boys and Girlsc	Neither Boys nor Girlsd	Prefer Not to Answere
Lifetime
Bullying	1286	451 (86%)d	362 (82%)c	76 (97%)b,d,e	62 (73%)a,c	124 (80%)c
Sexual Harassment	1300	236 (45%)b,c,d,e	163 (36%)a,c,d,e	51 (62%)a,b,d,e	16 (19%)a,b,c	37 (22%)a,b,c
Dating Violence	686	36 (12%)b,c	54 (21%)a	20 (33%)a,e	2 (9%)	4 (8%)c
Polyvictimization Score	1539	1.17c,d,e	1.10c,d,e	1.48a,b,d,e	0.81a,b,c	0.84a,b,c
Past 6 months
Bullying	1541	473 (80%)b,e	385 (70%)a,c	86 (86%)b,d,e	63 (67%)c	125 (62%)a,c
Sexual Harassment	1614	249 (40%)b,d,e	152 (27%)a,c,e	54 (50%)b,d,e	19 (19%)a,c	32 (16%)a,b,c
Dating Violence	866	42 (12%)	65 (18%)	16 (23%)	2 (8%)	4 (7%)
Polyvictimization Score	1738	1.14b,d,e	0.97a,c,e	1.38a,b,d,e	0.79a,c	0.70a,b,c

^aMean is significantly different from the mean for only boys (p < .05)

^bMean is significantly different from the mean for only girls (p < .05)

^cMean is significantly different from the mean for both boys and girls (p < .05)

^dMean is significantly different from the mean for neither boys nor girls (p < .05)

^eMean is significantly different from the mean for prefer not to answer (p < .05)

Attraction to both boys and girls was associated with the highest prevalence of past six-month bullying victimization (86%), significantly higher than youth only attracted to girls (70%), attracted to neither boys nor girls (67%), and those who preferred not to answer (62%). Youth attracted to only boys also experienced higher prevalence of past six-month bullying victimization (80%). The highest prevalence of past six-month sexual harassment victimization was reported by youth attracted to both boys and girls (50%) and attracted to boys only (40%); prevalence among these groups was higher than among youth only attracted to girls (27%), attracted to neither (19%), and those who preferred not to answer (16%). Youth attracted to both boys and girls (1.38) followed by those attracted to boys only (1.14) reported higher polyvictimization scores than youth reporting attraction to girls only (.97) and to neither boys nor girls (.79) reported counts of violence than those who chose not to answer about attraction (.70).

Experiences of Victimization by Gender and Attraction.

Girls who reported any attraction to girls reported the highest lifetime prevalence of bullying (95%); this was significantly higher than the lowest reported group, girls attracted to neither boys nor girls (68%) (Table 3). Lifetime prevalence of sexual harassment was significantly higher among girls reporting attraction to girls (61%) and girls only attracted to boys (46%) than boys only attracted to girls (35%), boys and girls attracted to neither (21% and 17%, respectively), and boys and girls who prefer not to answer (21% and 25%, respectively).

Table 3.

Lifetime and Past 6 Month Rates of Violence Victimization by Attraction for Boys and Girls

		Boys				Girls
Attraction to:	n	Only Girlsa	Ever Boysb	Neither Boys Nor Girlsc	Prefer Not To Answerd	Only Boyse	Ever Girlsf	Neither Boys Nor Girlsg	Prefer Not To Answerh
Lifetime
Bullying	1256	339 (82%)	23 (92%)	34 (79%)	58 (78%)	440 (86%)	70 (95%)g	28 (68%)f	60 (85%)
Sexual Harassment	1267	147 (35%)e,f	8 (40%)	9 (21%)e,f	16 (21%)e,f	234 (46%)a,c,d,g,h	48 (61%)a,c,d,g,h	7 (17%)e,f	19 (25%)e,f
Dating Violence	669	48 (20%)	7 (41%)e,h	1 (8%)	3 (15%)	35 (12%)b	13 (25%)	1 (10%)	1 (4%)b
Polyvictimization Score	1502	1.07f	1.36d,g	0.88f	0.80b,e,f	1.18d,g	1.42a,c,d,g,h	0.75b,e,f	0.91f
Past 6 months
Bullying	1505	360 (69%)e,f	26 (87%)	30 (67%)	57 (60%)e,f	459 (79%)a,d	77 (86%)a,d	33 (67%)	61 (66%)
Sexual Harassment	1577	141 (26%) e,f	11 (39%)	12 (25%)	11 (12%) e,f	243 (40%)a,d,g,h	47 (49%) a,d,g,h	7 (13%) e,f	19 (21%) e,f
Dating Violence	846	57 (17%)	6 (30%)	2 (13%)	3 (11%)	40 (11%)	14 (23%)	0 (0%)	1 (4%)
Polyvictimization Score	1697	0.95d,e,f	1.39d,g,h	0.85f	0.64a,b,e,f	1.13a,d,g,h	1.33a,c,d,g,h	0.74b,e,f	0.79b,e,f

^aMean is significantly different from the mean for boys only attracted to girls (p < .05)

^bMean is significantly different from the mean for boys ever attracted to boys (p < .05

^cMean is significantly different from the mean for boys attracted to neither boys nor girls (p < .05)

^dMean is significantly different from the mean for boys who prefer not to answer (p < .05)

^eMean is significantly different from the mean for girls only attracted to boys (p < .05)

^fMean is significantly different from the mean for girls ever attracted to girls (p < .05)

^gMean is significantly different from the mean for girls attracted to neither boys nor girls (p < .05)

^hMean is significantly different from the mean for girls who prefer not to answer (p < .05)

Note: Ever boys = boys who reported that they were attracted to both boys and girls or boys who were only attracted to other boys; Ever girls = girls who reported that they were attracted to both boys and girls or girls who were only attracted to girls

Boys reporting any attraction to boys had the highest lifetime prevalence of dating violence victimization (41%), and significantly higher than girls attracted to boys (12%) and girls who prefer not to answer (4%). Girls who are ever attracted to girls reported the highest lifetime polyvictimization scores (1.42), significantly higher than boys only attracted to girls (1.07), boys and girls attracted to neither (0.88 and 0.75, respectively), and boys and girls who preferred not to answer (0.80 and 0.91, respectively).

Girls reporting ever being attracted to girls and being attracted to boys only reported higher prevalence of past six-month bullying victimization (86% and 79%, respectively) than boys only attracted to girls (69%) and boys who preferred not to answer (60%). Girls who are only attracted to boys (40%) and girls who are ever attracted to girls (49%) reported higher prevalence of past six-month sexual harassment than boys only attracted to girls (26%), girls attracted to neither (13%), and boys and girls who preferred not to answer (12% and 21%, respectively). Boys who are ever attracted to boys reported the highest past six-month polyvictimization scores (1.39), significantly higher than girls attracted to neither boys nor girls (0.74), and boys and girls who preferred not to answer (0.64 and 0.79, respectively). Girls who are ever attracted to girls (1.33) and girls attracted only to boys (1.13) also reported higher polyvictimization scores than boys only attracted to girls (.95). Of note, the sample sizes were too small to examine the extent to which rates of violence victimization varied among transgender youth as a function of romantic attraction.

Discussion

The current study adds to the literature by documenting rates of varying forms of interpersonal violence victimization in a sample of young adolescents in grades 6–8. We explored to what extent rates of interpersonal violence victimization differ by sexual/romantic attraction; and the intersection between sexual/romantic attraction and gender for boys and girls. Data revealed notable findings across these comparisons, with differences in interpersonal violence victimization among middle school youth by attraction, and the intersection between attraction and gender.

Victimization by sexual/romantic attraction, specifically attraction to both boys and girls among this sample of middle school youth was associated with the highest prevalence of all forms of interpersonal violence victimization, and this difference was significant across all forms of victimization other than past-six month dating violence. This extends previous research in high school samples pointing to higher rates of sexual violence and dating violence among sexual minority students (Martin-Storey et al., 2021; Norris & Orchowski, 2020). Among older populations, bisexual individuals tend to experience the highest rates of violence victimization, especially in terms of sexual violence and dating or intimate partner violence (Bermea et al., 2018; Dickerson-Amaya & Coston, 2019; Flanders et al., 2019), but additional research is needed exploring when these patterns emerge. Attraction to only boys, regardless of youth gender, was associated with higher prevalence of recent bullying and sexual harassment victimization than attraction to only girls. This, in part, contradicts previous research that shows that boys (Birkett & Espelage, 2015; Espelage et al., 2015) are more likely to engage in bullying, both as perpetrators and as victims, and future research is needed to understand how bullying experiences in middle school vary as a function of attraction.

Regarding the occurrence of violence at the intersection of gender and sexual attraction, data revealed that girls who report attraction to boys or girls and boys attracted to other boys reported similar patterns in their experiences of victimization. In general, boys and girls who report any same-gender attraction report the highest rates of victimization. This aligns with research on bullying among middle schoolers (Mittleman, 2019), however, this pattern has been underexplored to date among middle schoolers for other forms of violence. In most cases, girls with same-gender attraction report the highest victimization rates, but this pattern reverses for dating violence where boys with same-gender attraction report higher rates. Studies of interpersonal violence often examine victimization experiences as a function of gender or sexual orientation separately (i.e., Dank et al., 2014). Patterns of risk that have been previously established by gender or sexual orientation separately may function differently when both are accounted for simultaneously. This approach may be especially important for youth attracted to both boys and girls and boys with same gender attractions as their unique patterns of risk may be especially pronounced. These data highlight the importance of attending to gender as well as sexual/romantic attraction when characterizing the interpersonal violence experiences of youth.

Limitations

Whereas the current study adds to the literature in several ways, the data should be interpreted considering several study limitations. Specifically, because the research enrolled students within school settings, the assessment of race and ethnicity was not permitted by the local Board of Education. Whereas the opt out rate for the study was low (<4%), a portion of students were excluded from the dataset due to inconsistent responses on questions designed to ensure students were paying attention to the study questions. Thus, the sample may not be representative of all students in the schools surveyed. It should also be noted that the assessment of gender identity in the current sample was limited to four choices: boy, girl, transgender, or prefer not to answer. Future research may utilize broader categories of gender identity, which allow students to describe their gender identity using their own terms, which may not have been captured through these questionnaire options. Specifically, because there was not a question that assessed sex assigned at birth, we cannot guarantee that all students that selected boy or girl were cisgender, which creates an imprecise measurement of transgender students. We have limited information to understand how middle school youth understand the distinction between sex and gender, and therefore future research should investigate how best to accurately represent gender expression and gender identity among youth. Further, the study sample included a relatively small number of transgender youth and that coupled with the imprecise measurement of transgender students led us to exclude them from analyses. Indeed, when exploring the intersection of gender, we were unable to include transgender youth both because of the relatively small sample size, but also through the inability to categorize youth as having same-gender attraction. Moving forward, it is essential for research addressing intimate partner violence and sexual violence to attend to multiple intersections of identity (i.e., race, gender, sexual orientation) while maintaining confidentiality, and to ensure sufficient representation of transgender and gender diverse youth when seeking to understand experiences of violence victimization (Dills & Brown, 2019).

There are also important considerations regarding the operationalization of some constructs in the current survey. We implemented a measure of sexual/romantic attraction which allowed participants to report whether they would describe themselves as attracted to: boys, girls, both boys and girls, neither boys nor girls, or if they preferred not to answer. It was unclear whether this group of students preferred not to answer because they were uncertain about how to classify their sexual/romantic attraction, they did not yet have feelings of sexual romantic attraction at all, or for other reasons (i.e., confidentiality). We believe it was important to include these students and their experiences of victimization, while acknowledging the uncertainty of who is in this group. Future research may seek to characterize why students prefer not to answer items regarding sexual/romantic attraction, and whether these questions are potentially avoided due to internalized stigma, a lack of privacy in the administration of a survey, or because some youth do not yet feel comfortable labeling their sexual/romantic attraction at this age.

The current data also do not allow the opportunity to characterize the experiences of victimization experienced among this sample of middle school youth. Future research may seek to garner information on the context of violence; for example, whether it is perpetrated by peers, occurred on school grounds, or occurred in the presence of bystanders who might have the opportunity to intervene. The utilization of surveys in the current research is also a limitation and may incur recall bias.

Future Research Directions

Notably, public health strategies for violence prevention, including other forms of ACEs, are shifting towards addressing risk and protective factors at the community- and societal-levels of the social ecology, which are likely to have a synergistic and broad-scale impact preventing multiple types of violence simultaneously (Armstead et al., 2018). The increased risk for multiple forms of violence among same-gender attracted girls and boys highlights the need to uncover risk and protective factors that – if targeted – could have a cross cutting effect on more than one form of victimization. Numerous studies support the role of rigid gender role expectations and norms that support aggression as a societal-level influence on the occurrence of multiple forms of interpersonal violence (Armstead et al., 2018; Wilkins et al., 2014). Despite recognition of increased rates of interpersonal violence among SMY, the social stressors experienced by minorities are not routinely addressed in the context of prevention. Stressors experienced by members of marginalized groups are often additive, contributing to increased psychological stress, and worse recovery following an experience of trauma (Binion & Gray, 2020; Diaz et al., 2001).

From a public health standpoint, advancing violence prevention efforts requires a strong understanding of the scope of the problem. The current research adds to the literature by documenting how various types of interpersonal violence victimization among middle school youth differed as a function of attraction and it’s intersection with gender. These findings are consistent with other studies among older adolescents and adults that document increased risk for interpersonal violence victimization among youth who indicate same-gender attraction (D’augelli & Grossman, 2001; Dank et al., 2014; Norris & Orchowski, 2020). It is vital that future efforts to advance violence prevention seek to explicate the pathways through which SMY are at greater risk for multiple forms of interpersonal violence victimization and other ACEs, given the increased rates of multiple forms of victimization among SMY in this research. Given that many prevention programs use a heteronormative lens, future work should aim to provide examples and programming that utilizes a broader pool of examples and address issues that are more specifically related to SMY. These future efforts should contain less rigid measurement of gender, attraction, identity to more accurately capture the nuance in terms of risk that may exist. Tailored population-targeted in-depth programs or adaptations for SMY youth to existing evidence-based prevention efforts like bystander approaches, norms-based approaches to foster protective environments, and increasing knowledge about healthy relationships and social-emotional learning (Basile et al., 2016) for multiple forms of violence (Niolon et al., 2019; Vivolo-Kantor et al., 2019) can potentially have a cross-cutting effect on multiple forms of violence simultaneously.

Conclusion

As the number of youth who identify with a sexual minority identity is increasing (Jones, 2021) it is imperative that we identify and address the disproportionate risk for violence that they experience. These experience of minority stress (Meyer, 1995) have been linked with multiple deleterious outcomes for youth (Hatzenbuehler, 2009; Meyer, 2003), and efforts to prevent violence should reflect the experiences of SMY. While creating violence prevention programs that are conscious and inclusive of SMY, along with promoting sexuality-specific supports such as GSAs are important steps, many of these efforts are focused among high schoolers. This research builds on a body of work motivating these efforts among middle school populations. As research has shown that certain types of violence peak (e.g., bullying) or begin (e.g., dating violence) among middle schoolers (Espelage et al., 2012; Mittleman, 2019), it is necessary to address these issues prior to their onset. Future research should continue to explore how these experiences of violence vary both by gender and sexual orientation, and especially at their intersection, to better understand how to prevent it from happening in the first place.