sexual harassment in college essay

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• HHS Author Manuscripts

Prevention of sexual violence among college students: Current challenges and future directions

Erin e. bonar.

a Addiction Center, Department of Psychiatry, University of Michigan, Ann Arbor, MI, USA

b Injury Prevention Center, University of Michigan, Ann Arbor, MI, USA

Sarah DeGue

c Division of Violence Prevention, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, Atlanta, GA, USA

Antonia Abbey

d Department of Psychology, Wayne State University, Detroit, MI, USA

Ann L. Coker

e Department of Obstetrics & Gynecology, School of Medicine, University of Kentucky, Lexington, KY, USA

Christine H. Lindquist

f Division for Applied Justice Research, RTI International, Research Triangle Park, NC, USA

Heather L. McCauley

g School of Social Work, Michigan State University, East Lansing, MI, USA

Elizabeth Miller

h Division of Adolescent and Young Adult Medicine, University of Pittsburgh Medical Center Children’s Hospital of Pittsburgh & Department of Pediatrics, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

Charlene Y. Senn

i Department of Psychology and Women’s & Gender Studies Program, University of Windsor, Windsor, Ontario, Canada

Martie P. Thompson

j Department of Psychology, Clemson University, Clemson, SC, USA

Quyen M. Ngo

k Department of Emergency Medicine, University of Michigan, Ann Arbor, MI, USA

Rebecca M. Cunningham

l School of Public Health, University of Michigan, Ann Arbor, MI, USA

Maureen A Walton

Preventing sexual violence among college students is a public health priority. This paper was catalyzed by a summit convened in 2018 to review the state of the science on campus sexual violence prevention. We summarize key risk and vulnerability factors and campus-based interventions, and provide directions for future research pertaining to campus sexual violence.

Results and Conclusions:

Although studies have identified risk factors for campus sexual violence, longitudinal research is needed to examine time-varying risk factors across social ecological levels (individual, relationship, campus context/broader community and culture) and data are particularly needed to identify protective factors. In terms of prevention, promising individual and relational level interventions exist, including active bystander, resistance, and gender transformative approaches; however, further evidence-based interventions are needed, particularly at the community-level, with attention to vulnerability factors and inclusion for marginalized students.

Introduction

Sexual violence remains a critical public health concern for students attending colleges and universities. The Centers for Disease Control and Prevention (CDC) defines sexual violence as “a sexual act that is committed or attempted by another person without freely given consent of the victim or against someone who is unable to consent or refuse.” 1 (p.11) Such acts range from non-contact unwanted experiences (e.g., verbal sexual harassment) to forced penetration. The current paper generally focuses on contact-related behaviors. Unwanted sexual touching is the most prevalent form of campus sexual violence experienced by college students, followed by incapacitated (i.e., due to drugs/alcohol) rape, and attempted forced rape. 2 A clear understanding of the scope of this problem is hindered by: 1) use of official statistics to estimate its magnitude, as there is underreporting of sexual violence to campus authorities, 2) variation in definitions and measures of sexual violence in self-report surveys 3 – 7 , and 3) missing data on surveys. 8 Nonetheless, a recent review suggested that prevalence of sexual violence victimization on college campuses is approximately 5% for men and 25% for women. 8 These rates remain comparable to what has been found since researchers began systematically measuring college women’s self-reported victimization in the 1980s. 9

College students are an important focus for sexual violence prevention as 18 to 24 year-old females have the highest rate of sexual violence victimization compared to females of other ages 10 and 81.3% of female victims experience a first rape before age 25. 11 The college context positions students to be at risk through decreased parental monitoring, increased alcohol use, increased sexual activity, and exposure to peer norms about sexual violence. 12 , 13 Thus, sexual violence affects many college students, takes place in a high risk context of increasing independence and risk behaviors, and can result in many negative sequelae (e.g., psychiatric distress, academic problems). 14 – 16

It is critical to provide appropriate support for survivors 6 while also engaging in upstream prevention approaches to eliminate sexual violence. In 2018 the University of Michigan Injury Prevention Center hosted a day-long event titled, the “Summit on the Prevention of Campus Sexual Assault.” The purpose of this summit was to better understand the state of campus sexual violence prevention science and to identify next steps for prevention under a public health model. Leaders in the field presented research data pertaining to sexual violence prevention for college students across levels of the social ecology. 17 Presentations focused on the public health framework for prevention, risk factors for campus sexual violence based on experimental and survey research, special considerations for marginalized students (i.e., sexual and gender minorities, racial/ethnic minorities), and evidence-based prevention programs delivered on campuses and in college communities. Interactions occurring during this summit catalyzed this subsequent summary paper, which included summaries provided by speaker authors, supplemented by additional literature. Key video excerpts are available on the University of Michigan Injury Prevention Center’s website: https://injurycenter.umich.edu/event/summit-on-the-prevention-of-campus-sexual-assault/ .

Several recent reviews summarize the sexual violence literature, many including college students, focusing on: prevention programs, 18 – 22 violence against women, 23 risk factors for victimization and perpetration in college and general populations, 24 – 27 and, specifically, male-targeted sexual violence programs. 28 Reviews specific to campus sexual violence prevention have focused on topics such as campus response/policy 29 or more narrowly on prevalence, 2 , 5 or bystander programs, 19 , 20 with a notable exception addressing prevalence, prevention, response, and policy, broadly. 30 In light of these prior reviews and given the Summit’s focus, this paper broadly highlights key research findings to inform current campus sexual violence prevention efforts under a social ecological framework while recognizing that factors at a given level of the social ecology may also exert influence across levels. First, we summarize risk factors at the individual, relationship, and community levels, with particular attention to vulnerability among marginalized students (i.e., racial/ethnic minority students, LGBTQ students). Second, we describe key interventions across these levels. Finally, we provide directions for next steps in sexual violence prevention research and practice for colleges and universities, highlighting key technical packages. 31 – 33

Understanding risk factors for perpetration and victimization

To prevent sexual violence among college students, it is important to understand risk factors for perpetration and victimization identified in studies using survey and experimental research designs, across individual, relationship, and community levels. Note that although individual behaviors contribute to victimization risk when a perpetrator is present, sexual violence is never the fault of the victim as it cannot occur without a perpetrator.

Individual level risk factors

Individual level risk factors described here are summarized in Table 1 . Data suggest that men are more likely to perpetrate sexual violence with research lacking on risk factors for female perpetration, likely given the low rates of this behavior. 34 , 35 For men, risk factors for sexual violence perpetration with strong support include prior perpetration, sexual behaviors (e.g., multiple sexual partners, impersonal sex, motivation for sex, exposure to sexual media), and sexual- and violence-related cognitions (e.g., hostility toward women, hypermasculinity, rape myth acceptance). 24 Consistent with a prior review, 24 one recent study found that 19% of men reported sexual violence perpetration before college, with risk factors including sexual media consumption, alcohol misuse, and hypermasculinity. 36 Longitudinal research of sexual violence perpetration is infrequent, yet existing literature suggests that risk factors for perpetration, such as hostile masculinity, rape supportive beliefs, and pornography exposure, can change over the college years, and that these changes are associated with changes in the likelihood of sexual violence. 37 – 39 Studies also suggest protective factors that buffer against sexual violence perpetration. For example, empathy, an individual level construct, may be protective via moderating the relationship between some risk factors (e.g., perceived peer approval of forced sex, high risk drinking) and sexual violence perpetration. 40 Another study found that college men who felt they learned from committing prior sexual violence and felt remorse were less likely to perpetrate again within one year. In contrast, men who spoke about women in callous ways were more likely to perpetrate over time. 41

Summary of individual-level risk factors for sexual violence perpetration and victimization discussed

In addition to female sex, research suggests several individual risk factors for sexual violence victimization including indicators of lower socio-economic status (e.g., difficulty paying for basic needs). 15 , 35 Further, one study of college women found that over half of sexual violence cases occur in the context of victim and/or perpetrator intoxication 42 supporting acute intoxication as a risk factor. 24 Emerging research also suggests that sexual violence victimization in the year before college entry relates to increased hazardous drinking during the first year of college 43 which can further increase risk.

Consistent with these surveys, experimental studies provide clues about risk factors for perpetration. 44 , 45 Given the extensive literature linking alcohol consumption with sexual violence perpetration, 46 alcohol is the most frequently manipulated variable in laboratory studies. 27 Typically, men are randomly assigned to consume non-alcoholic or alcoholic beverages and are asked to read, listen to, or watch a sexual violence scenario before answering questions about how they would think, feel, and act in that situation. Compared to sober men, intoxicated men often have higher scores on measures of the woman’s sexual arousal and their own sexual arousal, anger, belief that the woman “owed them” sex, perceptions that the man in the scenario acted appropriately, and willingness to act similarly in that situation. 47 – 51 Men higher in hostility and sexual dominance are the most likely to respond in a sexually aggressive manner when intoxicated. 52 , 53 Thus, the effects of alcohol on cognitive processing appear to increase the likelihood that a man who is predisposed to be sexually aggressive will act in such a manner when intoxicated.

Beyond these general risk factors, marginalized students may differ in their experience of sexual violence vulnerability; thus, we highlight key topics of race and ethnicity and gender and sexual diversity below.

Racial and Ethnic Minority Students.

The association between race/ethnicity and risk of sexual violence victimization is complex given that the socio-historical context of race in the United States has had a long-standing impact on the economic, social, and health consequences experienced by marginalized communities. Racial categories intersect and are influenced by other variables across levels of social ecology (e.g., socioeconomic status, campus environments, cultural norms, alcohol consumption), the unique effects of which are difficult to isolate. Measurement differences across studies increase this complexity. Although small numbers make precise estimates difficult, American Indian/Alaska Native students appear to experience higher rates of sexual violence than other students. 54 Asian and Pacific Islander students appear to be at lowest risk. 55 , 56 One study found that Latino students (the largest minority group enrolled in U.S. post-secondary institutions 57 ) had lower rates of victimization than White students. 55 However, another study found generally comparable rates for Hispanic (12.2%) and non-Hispanic (11.6%) students, 56 making it difficult to discern a consistent pattern.

Regarding Black students (the second largest minority group in U.S. post-secondary institutions 57 ) results are mixed. One study found that they had higher odds of past-year sexual violence compared to White students, 55 another found comparable rates for White and Black students, 56 and others have reported lower rates for Black students compared to White students. 9 Data from the Campus Sexual Assault study (conducted at two traditionally White public universities), and a similar study at four Historically Black Colleges and Universities (HBCU), suggest that racial differences in alcohol consumption and the extent to which survey questions account for alcohol-related incapacitation may clarify these inconsistencies. HBCU undergraduate women, versus those at non-HBCUs, had lower rates of incapacitated sexual violence (and overall sexual violence) yet comparable rates of physically forced sexual violence since entering college. 58 This difference appeared to be due to lower alcohol consumption among Black women (not unique factors of HBCUs), because no differences in sexual violence were found between Black women attending HBCUs and non-HBCUs and alcohol consumption was lower for Black women across school type. 58 Therefore, measurement nuances, particularly whether surveys query sexual violence when incapacitated due to substances, are important when attempting to understand racial/ethnic differences. Survey questions that do not assess sexual violence when incapacitated due to substances may undercount such experiences, producing estimates that primarily reflect physically-forced incidents.

Sexual and Gender Minority Students.

Sexual and gender minority college students are at elevated risk for sexual violence victimization prior to 59 , 60 and during college, 12 , 56 , 60 , 61 compared with heterosexual and cisgender students. Bisexual students experience disproportionate risk, with more than a quarter reporting sexual violence victimization, compared to 14% of gay and lesbian students and 11% of heterosexual students. 56 Students with multiple marginalized identities experience elevated victimization risk. A recent National College Health Assessment study found that Black transgender students had the highest predicted probability of sexual violence (58%) compared to the Latino (27%) and White (14%) transgender students. 55

Numerous social and contextual factors shape vulnerability to sexual violence among sexual and gender minority students. They often experience discrimination related to these aspects of their identities, 62 – 64 which fuels violence perpetration against them, 65 fosters feelings of internalized homonegativity 66 – 68 and hinders disclosure of victimization. 69 Social norms and attitudes also shape these students’ vulnerability to sexual violence. For example, bisexual students confront sexual objectification, 70 pressure to “prove” their sexuality, 71 and cultural narratives of hypersexuality. 70 These experiences are compounded by potential exclusion from both heterosexual and queer communities, 70 , 72 which makes care-seeking difficult. Male-identified survivors of sexual violence may be silenced by hypermasculinity norms that are incongruent with narratives of survivors as feminine and defenseless. 69 , 73 , 74 Such barriers are compounded for gay and bisexual male survivors, who face pressures to conform to hegemonic masculine norms. 75

Relationship level risk factors

Relationship level risk factors described here are summarized in Table 2 . Risk factors for sexual violence perpetration at the relationship level include all male peer affiliation. 24 A meta-analysis suggested that males’ athletic and fraternity involvement increases risk for sexual violence supportive attitudes and perpetration. 76 As influenced via peer relationships, perceptions of peer attitudes and behaviors (e.g., peer approval of forced sex, peer sexual aggression, peer pressure for sex) are also associated with perpetration. 24 , 36 , 38 , 39 , 77 Longitudinal research shows that as perceptions of peer approval of forced sex and sexual coercion increase, so does likelihood of perpetration, whereas decreases in rape supportive peer norms are associated with a declining trajectory of perpetration. 38 , 39 Peer norms have also been manipulated in lab-based experiments, in which men are asked to select a film clip (sexually aggressive or not) to show to a confederate woman whom they believe is another study participant. Men are more likely to choose the sexually aggressive film when they are randomly assigned to first watch another man (a confederate) choose the film clip. 78 In simulated bystander situations, men randomly assigned to hear male confederates objectifying women are less likely to stop a female confederate’s exposure to a sexual video, as compared to men assigned to hear neutral statements about women. 79 Being in a casual or committed relationship (vs. single) is associated with more sexually aggressive behaviors among college men, particularly for those using alcohol prior to sex. 80 At the family relationship-level, pre-college factors such as child abuse and exposure to parental violence are risk factors for perpetration. 24

Summary of relationship-level risk factors for sexual violence perpetration and victimization discussed

Social situations involving drinking peers are important risk factors for perpetration and victimization. In lab studies 81 , 82 , intoxication can reduce bystanders’ recognition of dangerous situations. In surveys, alcohol consumption is associated with lower likelihood of bystander intervention. 83 , 84 Regarding victimization risk, being in a dating relationship confers risk as a large portion of sexual violence occurs with dating partners. For example, a recent study of 361 students found that 35% had experienced sexual violence from a dating partner 85 and that exposure to deviant peers (e.g., involved in crime) was associated with attempted rapes. Recent research has also identified meeting potential partners via dating apps and having “hook-up” type relationships as risk factors for victimization, 15 , 86 with more research needed to understand what factors drive these associations (e.g., are individuals who use dating apps also more likely to perpetrate sexual violence?). Prior victimization is a risk factor for future victimization as well. 15 , 35 Related to re-victimization, positive peer norms about hook-ups are associated with re-victimization among college women with prior sexual violence victimization. 87 Fraternity/sorority involvement is also a relationship-level risk factor for victimization. 15

Community, contextual, and institutional level risk factors

Given that sexual violence reporting rates vary across institutions, 56 an emerging focus in prevention-related research involves identifying risk factors at the campus and/or community level. 24 , 88 , 89 Such risk factors are typically measured via locations/policies related to alcohol consumption (e.g., drinking settings) that increase risk for sexual violence, individual perceptions of campus climate (e.g., inclusivity), and institutional characteristics (e.g., public vs. private). Research using these proxy-measures for community level factors provide direction for community-level interventions to address risk for sexual violence via policy, environmental changes, or institution-wide practices to change the culture of the campus and/or students’ broader community. To begin, although data from the National Crime Victimization Survey indicate that most incidents of female sexual violence victimization among college students occur at or near their home (38%) or the home of someone they know (29%), 10 campus environments that facilitate high-risk drinking are associated with increased risk. Longitudinal research shows that college men’s attendance in high risk drinking settings (bars/parties) is associated with perpetration. 90 A recent study from two New York campuses found that in cases of incapacitated sexual violence victimization, about half of women had just been at a party with the perpetrator before the event (versus 15.8% for non-incapacitated victimization), and nearly half (46.5%) said the incident occurred in a dorm (fraternity house: 10.5%; off-campus party/bar: 14.9%; other location: 28.9%). 86 Next, while athletic and Greek life involvement are identified as relationship-level risk factors, particularly because they are associated with attitudes that predict sexual violence perpetration (e.g., hypermasculinity, rape myth acceptance 76 ), more research is needed to clarify which types of events and/or features of the environment of these communities (e.g., parties, formal events) affect risk. 89

Second, while research documents elevated risk among sexual and gender minority students, 12 , 56 , 60 , 61 , 91 studies examining community-level factors, such as how campus climates may drive these disparities, are lacking. Studies using proxy-measures of campus inclusivity, namely sexual and gender minority students’ perceptions, provide clues to inform campus-level prevention interventions. For example, among students from 478 higher education institutions, perceived inclusivity of sexual and gender minority people on campus was associated with significantly lower odds of sexual violence for these individuals. 92 Inclusive climates may operate by reducing perpetration against sexual and gender minority students, increasing bystander intervention, or empowering students to use harm reduction strategies. 92

Finally, several studies examined institutional characteristics in order to infer community-level factors. One study found that institutional characteristics (public/private, 2-/4-year; metro/non-metro location) were not significantly associated with perpetration rates, 34 whereas another found significant differences with a relatively small magnitude (e.g., higher victimization rates at small vs. large, and private vs. public institutions). 56 Another study found that geographic region is associated with increased risk in some cases (e.g., Midwestern and Southern vs. Northeastern campuses), as was campus size (10,000–20,000 students versus <2,500), while research institutions had lower rates than bachelor’s institutions. 93 This study also identified several student body features associated with campus sexual violence rates, including higher campus-wide binge drinking rates, lower proportions of heterosexual students and higher proportions of younger students and females, higher number of sexual partners, and greater proportions reporting discrimination.

As most studies examine traditional universities, research is needed to examine community college characteristics (e.g., frequent evening classes, commuter factors such as walking to parking lots). Recent data from seven northeastern U.S. community colleges (N=800 students) suggest concerning rates of sexual violence victimization since enrollment (11% victimized via unwanted sexual contact, coercion, attempted and completed rape; 48% when including sexual harassment, relationship violence, stalking, etc.). 94 Students who were female, sexual and racial minorities, and under age 26 were more likely to report any victimization. 94 These data highlight unique characteristics of community colleges, which may reflect physical environments and/or differences in programs or policies related to sexual violence requiring further study.

Future directions

Table 3 summarizes future directions for research in regarding risk factors for campus sexual violence based on gaps in the above literature. These include: (1) conducting additional longitudinal studies including time-varying risk factors and evaluating victimization and perpetration trajectories for students of all genders; (2) using a developmental lens to examine malleable risk factors (e.g., alcohol consumption, peer norms), that can change during college; and (3) focusing on understudied risk factors at the community and contextual levels (e.g., physical and social/contextual environment), including in community colleges. Understanding community-level factors could inform implementation of effective prevention strategies across levels of the social ecology, particularly for marginalized students. Further, given variations in risk across sexual, gender, and racial/ethnic minority sub-groups, more research is needed to better understand how sexual violence manifests in these marginalized groups to inform prevention. We note that specific racial/ethnic groups should not be presumed to be at higher or lower risk without more consistent evidence. Further, students with disabilities comprise another marginalized population warranting future attention, given research suggesting that individuals with disabilities experience increased risk for sexual violence. 95 , 96

Summary of next steps for future research pertaining to risk factors and interventions

Finally, regarding experimental work, researchers are developing virtual reality paradigms which may allow for more nuanced and realistic responses. 97 , 98 The peer studies above demonstrate how constructs under consideration for interventions could be piloted in experimental studies. Some researchers find unexpected, harmful effects with high-risk individuals, hardening their attitudes when presented with messages intended to challenge beliefs about women or masculinity. 99 , 100 Thus, using experimental paradigms with updated technology to pilot potential prevention interventions may help identify and avoid unintended negative consequences.

Efficacious and Promising Prevention Programs

A remaining gap in campus sexual violence prevention work involves needing to expand and implement efficacious prevention approaches addressing risk for perpetration and victimization across genders and levels of the social ecology. Few approaches have been rigorously evaluated in college populations (i.e., with well-controlled designs, such as randomized controlled trials [RCTs]) and shown to reduce sexual victimization or perpetration or sexual violence risk factors. 18 , 19 , 22 We highlight key programs below targeting different levels of the social ecology (although some inherently address multiple levels).

Individual level

Few prevention programs focusing on sexual violence by potential perpetrators in college populations have strong evidence of effectiveness in reducing men’s perpetration behavior, with two promising exceptions. 101 , 102 RealConsent is an interactive, web-based program designed for college-age men to increase prosocial intervening behaviors, change attitudes and normative beliefs about sex, rape, and masculine gender roles, and increasing knowledge of consent. An RCT found that RealConsent was effective in decreasing sexual violence perpetration and increasing positive bystander behavior over 6 months. Similarly, the group-based Sexual Assault Prevention Program (1.5 hour program, 1 hour booster) which seeks to increase men’s empathy about sexual violence, decrease rape myth acceptance, increase consent knowledge, promote bystander intervention, and reduce normative misperceptions holds promise. 103 In addition to demonstrating efficacy on several risk factors (e.g., exposure to sexual media), relative to a control group, program recipients were less likely to perpetrate sexual violence over 4-months.

Other interventions focus on changing individual-level attitudes and behaviors to reduce the victimization risk 18 by teaching women how to recognize threat and bolstering self-efficacy to use self-defense strategies. 104 – 110 Few individual-focused programs demonstrate sustained behavioral outcomes. The Enhanced Assess, Acknowledge, Act (EAAA) Sexual Assault Resistance program 105 is a notable exception that also includes curricula on positive sexuality. Designed for women of all sexual identities, based on feminist and social psychological theory 111 , 112 and prior research, 113 – 115 EAAA includes four group sessions delivered by female facilitators to: 1) decrease the time needed to assess a situation as dangerous and take action, 2) reduce emotional obstacles to taking the action necessary to get away, and 3) maximize use of verbal and physical self-defense tactics most likely to be effective. The program script asserts that perpetrators are entirely responsible for their behavior. After pilot studies showed promise, 116 – 118 a multi-site RCT 106 showed that EAAA increased women’s perception of their risk of acquaintance rape, knowledge, self-efficacy, and willingness to use self-defense strategies in hypothetical situations. EAAA also reduced the relative risk of attempted and completed rape (non-consensual oral, vaginal, or anal penetration) by 50% in the year following participation, as well as non-penetrative sexual violence. Program benefits persisted at least two years later. 107 Despite concern that programs for women may inadvertently increase perceptions of women’s responsibility for sexual violence, 119 participation in EAAA led to sustained decreases in rape myth acceptance and beliefs in female provocation or responsibility for rape. 107 The program worked both for rape survivors and women with no prior rape experience, and decreased self-blame if women were raped after EAAA. 120 A multi-site implementation study of EAAA is underway in Canada.

As called for in a recent review, 16 comprehensive prevention strategies should target multiple settings and risk factors, 121 with campus healthcare settings being important locations (e.g., health centers, counseling centers, Emergency Departments), potentially due to the medical and mental health consequences of sexual violence. 14 – 16 , 122 , 123 Therefore, we highlight ongoing work using a cluster RCT testing a brief trauma-informed, harm reduction intervention to reduce risk for alcohol-related sexual violence among male and female students receiving care from 28 college health centers. 124 The Giving Information for Trauma Support and Safety (GIFTS) intervention is being compared to a brief alcohol risk reduction counseling intervention. GIFTS uses a palm-size safety card with information about sexual violence and provides: (a) education and assessment regarding sexual violence (regardless of disclosure); (b) discussion of harm reduction behaviors to reduce risk of alcohol-related sexual violence for self and peers (including bystander intervention); and (c) supported referrals to survivor services. GIFTS builds on a growing evidence-base demonstrating effectiveness of survivor-centered clinic-based interventions in increasing recognition of abusive behaviors, knowledge of resources, and self-efficacy to enact harm reduction strategies, and among specific populations, reducing physical, sexual, and cyber relationship abuse victimization and reproductive coercion. 125 – 127 Although primarily targeted at the individual level as primary prevention, GIFTS is also intended to amplify campus prevention efforts by encouraging students who have witnessed sexual violence, to be more likely to intervene to interrupt a peer’s harmful behaviors and to provide support to peers who have been harmed. As secondary prevention, students with prior sexual violence exposure receiving GIFTS are expected to be more likely to disclose sexual violence during their clinic visit and report greater use of sexual violence-related services. 124

Relationship level

Bystander training addresses the relationship level of the social ecology by training individuals to intervene with others when they witness sexual violence or behaviors that increase sexual violence risk, such as use of sexist or objectifying language, endorsement of violence-supportive attitudes, or violence risk behaviors (e.g., heavy alcohol use). Engaged bystander programs seek to decrease sexual violence acceptance and build skills to increase bystander actions that can reduce sexual violence. As such, when well-implemented across a community (i.e., college campus), bystander programs can change perceived social norms around violence and encourage both men and women to take action to prevent it. The Campus SaVE legislation required training on college campuses that teaches “safe and positive bystander intervention that may be carried out by an individual to prevent harm or intervene” when there is a risk of violence. 128 A number of bystander programs have focused on men and women (e.g., Green Dot, 129 – 131 Bringing in the Bystander, 132 – 134 Know Your Power 135 ), while others use sex-specific training [e.g., the Men’s Program 136 – 138 and Women’s Program 139 , 140 ]. 21 Rigorously evaluated and effective bystander programs for colleges include Bringing in the Bystander 132 , 134 , 141 and Green Dot 129 – 131 ; we specifically highlight Green Dot below, as it was presented in our prevention Summit.

The Green Dot program for colleges and high schools includes two phases: 1) a 50-minute motivational speech including definitions of sexual violence, its frequency, risk factors, and opportunities for prevention and intervention, designed for delivery to all students at a campus; and, 2) an intensive, interactive skill development bystander training, ideally delivered in groups of 20–25 over 4–6 hours using a Peer Opinion Leaders strategy (e.g., training by those whom others emulate or respect). For college students, Green Dot significantly reduced sexual violence acceptance and increased active bystander behaviors relative to those who did not receive Green Dot. 129 One college campus using Green Dot, relative to two with no bystander program, had significantly lower rates of sexual violence accounted for by a reduction in alcohol- or drug-facilitated sexual violence. 131 Similarly, over 4 years, rates of sexual violence were 25% lower in the Green Dot campus versus the two with no bystander program 131 ; this was primarily accounted for by reduced alcohol- and drug-facilitated sexual violence. This program also reduced sexual harassment, stalking, and psychological dating violence victimization and perpetration. 130 , 131

Community level

To date, no community-level interventions for campus sexual violence have been rigorously tested. A review 18 of over 140 sexual violence prevention programs (not restricted to college campuses) found that very few (<10%) addressed campus climate or policies. A study 142 of 24 four-year colleges in Georgia, found that only 14% had policies and practices that met compliance criteria for the Clery Act in 2014. Policies varied widely across institutions, despite increased public attention and federal guidance surrounding campus sexual violence after a 2011 Dear Colleague letter from the Department of Education.

Future Directions

Table 3 summarizes future directions for research on prevention interventions, based on gaps in the literature above. More research is needed to continue identifying effective primary prevention programming for campus sexual violence. Although bystander programs are effective at changing bystander attitudes and behaviors, 129 , 132 , 134 bystanders are present in few situations where sexual violence directly occurs (no more than 17%). 143 Combining bystander programs with programs helping students build knowledge and skills to intervene on their own behalf is warranted. 144 – 147 Similarly, programs targeting high-risk groups that address norms related to masculinity and sexual aggression in addition to bystander training may be suitable for college settings. For example, the coach-delivered Coaching Boys Into Men program for male athletes, was efficacious among middle school students in increasing bystander behaviors and decreasing relationship violence among those with a history of dating. In addition to potentially adapting such programs for college campus delivery, there is a need to examine whether the effects of this and other early prevention programs persist into the college years. 148 Next, implementation research that seeks to implement and identify essential elements of efficacious programs (e.g., using dismantling designs 149 ) and optimal program dose, across specific settings (e.g., dorms, online, classroom; community colleges, commuter or residential schools) and populations, would be an important contribution. Conducting cost-effectiveness analyses of efficacious programs can help ensure that prevention resources are well-utilized, and can assist communities in making informed implementation decisions. Collecting cost-related data during an evaluation can permit such analyses with limited need for additional resources. Finally, research identifying optimal implementation and dissemination strategies for efficacious programs is needed. Although a challenging task, evaluating comprehensive prevention strategies that include combining evidenced-based programming with broader initiatives to improve campus climate for marginalized students (e.g., programs that enhance inclusivity for sexual and gender minorities such as those targeting micro-aggressions, tailored prevention services for minority students at higher risk), could move the field forward.

As a limitation of the field, prior interventions often focused on men’s perpetration of sexual violence toward (presumably heterosexual) women, with future studies needed to enhance prevention for gender and sexual minorities. Moreover, research is needed regarding male victims of sexual violence who often have few resources and potentially more stigma (e.g. rape myths around male victims), and female perpetrators of sexual violence who are rarely studied (e.g., risk factors, motivations). 35 , 150 – 152 This gap could be addressed by qualitative work to better understand and broaden our understanding of gendered risk factors. Programming that better accounts for gendered risk factors is also needed, as gender-neutral programs may not sufficiently address well-established risk factors associated with traditional masculinity. 24 , 36 Further, given the variation in risk for sexual violence victimization for marginalized students, existing data can be used for sub-group analyses to inform tailoring programs for specific student populations. Few existing interventions are designed for marginalized populations specifically, 18 yet programs may be more effective when tailored to cultural beliefs and norms. 153 , 154

Finally, prior prevention programs generally fall within inner levels of the social ecology—addressing individual and relationship risk factors—with little known about what works at the community level. Individual-level efforts must be accompanied by universal, campus-wide efforts in partnership with the local community to make the college experience safer for all students, including vulnerable populations by addressing institutional and structural factors (e.g., racism, homophobia) that enhance risk. Alcohol policies on and off-campus can help mitigate sexual violence for students. 31 , 155 Engaging alcohol outlets by providing bystander training for bar staff has improved their positive bystander intentions and could benefit from research examining long-term outcomes. 156 Campus efforts to support equity across genders and marginalized identities via increasing representation amongst faculty and leadership can promote an inclusive campus culture. 31 Investing in programs (e.g. SafeZone 157 ) for sexual and gender minority students, and bolstering campus enforcement of sexual violence and anti-discrimination policies to be responsive to the lived experiences of sexual and gender minority students and other marginalized students, could be an important part of community-level interventions. Finally, it is noteworthy that an efficacious building-level program for middle school students that involved hot-spot mapping and intervention, Shifting Boundaries, 158 is currently being adapted for college campuses and could yield promising results.

Summary and Directions for Future Research

There is broad agreement that a comprehensive approach is necessary to prevent campus sexual violence. 31 , 121 , 159 , 160 Such an approach should address multiple goals including: changing societal attitudes regarding the continuum of behaviors that comprise sexual violence, 161 , 162 developing programs, policies, or other prevention approaches to stop perpetration and hold perpetrators accountable, 18 empowering potential victims with knowledge and skills to act on their own behalf to defend their sexual rights, 105 , 107 and facilitating the empowerment of bystanders to disrupt harmful social norms, intervene on others’ behalf, and support survivors. 130 , 132 , 134 Existing efforts have not fully mitigated the problem 163 , 164 with consistent rates of sexual violence for decades 9 , 165 and recent data documenting the high economic cost of sexual violence (over $122,000 lifetime cost per rape victim; $3.1 for all victims 166 ).

To build a comprehensive, efficacious approach, key gaps must be addressed with support for additional research. First, although substantial progress has been made to identify risk factors for sexual violence perpetration and victimization, 18 , 24 data on protective factors are generally lacking. 24 Protective factors include characteristics of the individual, their experiences, their relationships, or their environment that can reduce violence or buffer the effects of risk factors. Research identifying modifiable characteristics of the campus culture, environment, policies, or practices that are associated with lower levels of sexual violence is needed, and would improve the ability of colleges and universities to create environments that buffer risk, supporting the development of comprehensive strategies that address external spheres of the social ecology. Although we focus on college campus interventions, it is important to note that risk and protective factors are often established prior to college, and prevention efforts are needed before college entry across levels (e.g., schools, cultural norms, policy) and for the general community.

Second, few studies have examined risk factors or interventions that cut across levels social ecology, with examination of community-level factors and prevention approaches being particularly needed. For example, a CDC systematic review of sexual violence perpetration risk factors in the general population 24 identified 42 individual-level risk factors, 23 relationship-level factors, and only 2 societal/community-level factors. In a parallel manner, few studies have tested the efficacy of multi-component interventions across levels, with additional research needed. 18 , 167 The social ecological framework can be helpful in conceptualizing the problem of sexual violence and targeting prevention, while also considering that factors can interact across levels. Further, programs for both perpetration and victimization prevention that work across levels of the social ecology can be tailored to cultural considerations that may affect one’s risk of perpetrating sexual violence or being a victim, while also considering the different contexts in which sexual violence occurs (e.g., partnered relationships, among acquaintances, in and out of the party scene). Consistent with White House Task Force recommendations, 168 the CDC’s STOP SV: A Technical Package to Prevent Sexual Violence 31 supported primary prevention programs that address different levels of social ecology by fostering healthy relationship skills, social norms that protect against violence, and protective environments as recommended best practices. STOP SV supports the idea that practitioners and communities may work together to identify the approaches best suited for local contexts.

Next, the majority of research focused on campus sexual violence focuses on the context of a male perpetrator and a female survivor, likely due to its high frequency. Conceptual models of perpetration and victimization may need to be expanded to fully incorporate the range of perpetrator-victim relationships. Additional support is needed for future research to increase representation of diverse students in efforts to inform adaptations of evidence-based programs that will be effective for reducing victimization among higher risk marginalized groups (e.g., sexual and gender minorities, students with disabilities, Black, Hispanic/Latino Students, and American Indian/Alaska Native students) with potential use of hybrid implementation-effectiveness designs 169 to more rapidly impact sexual violence rates. Although research findings are mixed regarding the impact of institutional characteristics on perpetration and victimization rates, 34 understudied institutional features or norms may differentially connote risk, particularly for marginalized individuals and men and women separately, and based on type of college setting, with community and commuter colleges being particularly under-represented.

In summary, we assert that comprehensive prevention from a public health perspective involves a set of coordinated multi-component strategies that address risk and protective factors across the social ecology, that complement and reinforce each other with consistent messaging from multiple sources across multiple contexts, including addressing the diverse student population (e.g., racial/ethnic, sexual and gender minorities, those with disabilities, those at community colleges and/or commuter schools). 18 , 88 Such strategies address risk characteristics of the individual and their relationships — which is typical in campus prevention efforts 19 — but also include community-level programs and structural/societal-level policies (e.g., Campus SaVE Act 128 ) that attempt to modify students’ physical and social environment. To build multi-level strategies aimed at transforming the current campus climate, we need a strong evidence base of risk and protective factors and effective strategies at all levels. In addition to the framework provided by the CDC technical package STOP SV: A Technical Package to Prevent Sexual Violence, 31 the CDC guide Sexual Violence on Campus: Strategies for Prevention 32 and the National Collegiate Athletic Association toolkit 33 may be helpful to colleges and universities working to improve campus sexual violence prevention efforts.

Acknowledgements:

Meredith L. Philyaw-Kotov, Jessica S. Roche, & Bethany Pollock

Funding: This work was supported by a grant to the University of Michigan Injury Prevention Center by the Centers for Disease Control & Prevention Award Number R49-CE-002099. Dr. Coker was supported by funding from the Centers for Disease Control and Prevention under Grant #U01CE15003; and by the Eunice Kennedy Shriver National Institute of Child Health and Human #R01HD075783. Dr. Ngo was also supported by funding from the National Institute for Child Health and Human Development under Grant #R03HD087520; and by the National Institute on Alcohol Abuse and Alcoholism under Grant #K23AA022641. Dr. Senn was supported by the Canadian Institutes of Health Research through a Tier 1 Canada Research Chair in Sexual Violence and Project grants. Dr. Thompson was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH under Grant #R15HD065568 and #R03HD053444-01A1. Drs. Miller and McCauley were supported by the National Institute on Alcohol Abuse and Alcoholism under Grant #R01AA023260.

The content is solely the responsibility of the authors and does not necessarily represent the official views of the funders, the Centers for Disease Control & Prevention, or the Department of Health and Human Services.

Declaration of Interest: The authors have no conflicts of interest to report.