To assess prevalence of victimization and perpetration of relationship violence before and during college, to explore variations by gender, and to examine differences by relationship type.
Anonymously surveyed students in 67 randomly chosen classes.
Three urban college campuses.
Nine hundred ten undergraduate college students aged 17 to 22 years.
Main Outcome Measures
Self-reported victimization and perpetration of physical, emotional, and sexual violence; relationship to the victim or perpetrator.
Most (57.1%) students were female, and 58.7% were white, 16.4% black, and 15.1% Asian. Of 910 participants, 407 (44.7%) experienced partner or nonpartner violence: 383 (42.1%) reported victimization and 156 (17.1%) reported perpetration. All victimization and perpetration rates were highest before college. Emotional violence was most common before college (21.1%); during college, sexual and emotional violence were equally common (12.0% and 11.8%, respectively). Women reported more victimization than men, but male victimization was considerable (27.2%). More men perpetrated sexual violence; more women perpetrated physical violence. More than half (130 of 227) of the violence experienced during college was partner related. Students experiencing partner violence during college were more likely to experience physical and emotional violence and were less likely to experience sexual violence.
Relationship violence is prevalent among college students and frequently occurs before college. Emotional violence was most frequent before college; sexual and emotional violence were equally common during college. Women reported more victimization than men, but male victimization was common. Men perpetrated more sexual violence; women perpetrated more physical violence. Physical violence and emotional violence were most often committed by partners, while sexual violence was less likely to be partner related.
Relationship violence among adolescents is concerning and can happen in a broad range of relationships, including those with friends, acquaintances, strangers, and partners.1,2 The existing relationship violence literature is largely limited to partner violence, which is most prevalent among women 24 years and younger.3,4 Extensive literature explores female partner victimization for women of college age.1,5- 20 These investigations have generally included convenience samples of students in introductory psychology or sociology classes,1,5,8,10,11,15,16,18,19 limiting the generalizability of the findings. These students may have a heightened sense of awareness of violence related to course content, or they may have selected the class because of prior personal experience. National studies13,14,17,21 about relationship violence among college-age students have restricted the assessment of violence exposure to the past 12 to 18 months, potentially underestimating exposure to violence during adolescence.
When perpetration has been studied, it has focused on violence with partners,10- 12,15- 18,20,22- 29 with few studies10,16,17,29 exploring perpetration other than physical. In light of data suggesting that male partner victimization is common,13- 20,26,30 further exploration of gender differences for victimization and perpetration of physical, sexual, and emotional violence is a logical next step. Recent studies exploring gender symmetry have used convenience samples of students in social science classes15,16,18,19,24,25,27,28 and have limited reporting periods.13,15,17,20,27,28
Most studies of adolescent relationship violence have focused solely on partner violence, not the broader range of relationships that develop in college (with friends, acquaintances, and partners) in which violence may occur. Transitioning to college may influence adolescents' vulnerability to involvement in violent relationships, as adolescents in an unknown environment may experience isolation as they leave social supports from home. Research suggests that less parental monitoring and the perception of low parental and social support (such as that experienced in the transition to the college environment) are associated with increased violence involvement.31,32 In addition, a strong desire for peer acceptance may influence relationship behaviors, making teens more vulnerable to violence.23,33- 35 It is unclear whether experiences with certain types of violence (ie, physical, sexual, and emotional) during college are associated with the nature of the relationship between victims and perpetrators.
We are unaware of any existing studies that explore past and current physical, sexual, and emotional victimization and perpetration among adolescent men and women in a broad range of college relationships within the same sample. By including randomly selected classes from 3 diverse college campuses, the present study expands on the literature by providing a more comprehensive examination of rates of violence among male and female undergraduate students. We examined physical, sexual, and emotional victimization and perpetration occurring before and during college and in a range of relationships, including (but not limited to) partner violence. The primary objectives of the study were (1) to examine victimization and perpetration experienced before and during college among undergraduate students aged 17 to 22 years, (2) to compare victimization and perpetration rates for men and women, and (3) to explore differences between the type of violence experienced and the relationship between victims and perpetrators.
Self-administered surveys were used to obtain prevalence estimates of relationship violence among urban college students. Three colleges were chosen to provide a demographically diverse sample of students. These included a nonresidential community college offering associate degrees in many fields, a large Ivy League university, and a private mid-sized Catholic university. Both universities offer undergraduate and professional degrees. Institutional review board approval was granted by each participating site and by The Children's Hospital of Philadelphia.
Using current online course rosters from each college, a random list of day classes was generated. Research staff (R.K.M.) contacted professors for the selected classes by e-mail and asked permission to administer a survey on relationship violence at the end of their class. All students attending class on the day of survey administration were eligible to participate. This analysis explores responses from undergraduate students aged 17 to 22 years.
The anonymous pencil-and-paper survey, consisting of 45 questions, was administered by a research assistant (R.K.M.) during the last 10 minutes of class. A brief introduction instructed students not to recomplete the survey if they had previously participated. Students could leave questions unanswered or return blank surveys if they chose not to participate. When students returned a survey, they were given wallet-sized resource guides listing campus and local area services.
The survey included demographic questions about gender, age, race, and length of time in school. Validated scales exist to measure relationship violence; however, they are time intensive and ask far more questions than could be answered in the time allotted. In addition, they focus on partner violence, not a broader range of relationships. Constructs from existing scales were used to shape our questions,36,37 which are similar to those used by others.1,6,14,21,35Physical violence was defined as pushing, grabbing, slapping, choking, or hitting. Emotional violence was defined as being put down, made to feel bad about oneself, being isolated from friends and family, or acting in a possessive manner. Sexual violence was defined as being pressured, coerced, or forced into having sexual contact.
Students were asked whether they had experienced physical, emotional, and/or sexual violence in a relationship before coming to college and during college and whether they had been a victim and/or perpetrator of each form of relationship violence. The age at first experience was asked for violence occurring before college only. For violence during college, victims were asked the gender of and the relationship to the perpetrator. Relationship options included casual acquaintance, friend, partner, stranger, or other, for which the students wrote in a response. Similar questions were asked of perpetrators about their victims. Because violence can occur in relationships before they become sexual, we intentionally surveyed participants irrespective of sexual activity within their relationships.
The analyses explored students' precollege and college experiences with physical, emotional, and sexual violence. For each form of violence, victimization and perpetration rates per 100 students (percentages) were calculated before college and during college. Participants first victimized during college did not provide the age at first victimization. In this case, we assumed the age at first victimization to be the current age. For example, a 21-year-old woman in the fourth year of college identified becoming a victim during college. She was assigned age 21 years as her age at first victimization, although she may have been victimized as young as age 18 years, her first year in college.
Victimization and perpetration rates for each type of violence were compared separately for women and men using the χ2, Fisher exact, and t tests. Specific analyses compared students who experienced partner violence during college with students who experienced nonpartner violence during college. Data were analyzed using commercially available statistical software (SPSS version 10.0; SPSS Inc, Chicago, Illinois).38P ≤ .05 denoted statistical significance.
A total of 298 classes were selected; we received responses from 128 classes, obtained permission to survey 106 classes, and surveyed 67 classes (Table 1). A broad diversity of classes was selected across disciplines. There were no differences between professors who responded and those who did not (P > .99), those who agreed vs those who refused (P =.21), and those who were surveyed vs those who were not (P =.31) by discipline. However, more professors at 1 school did not respond, which we attribute largely to limited use of e-mail (P < .001).
Students returned 910 completed surveys, representing about 70% of students enrolled in participating classes. Many classes at 1 college had far fewer attendees present for survey administration than were enrolled in the class, substantially decreasing the participation rate. We estimate that 2% to 3% of students who had attended class during survey administration left before receiving a survey or indicated that they had previously participated. Demographic characteristics of participants at each college (Table 2) match campus demographics at each school.
Of 910 participants, 44.7% (n = 407) experienced relationship violence; 27.7% (n = 252) experienced emotional violence, 24.9% (n = 227) sexual violence, and 20.9% (n = 190) physical violence. Of 407 students who experienced relationship violence, 51.8% (n = 211) experienced 1 type of violence, 31.9% (n = 130) experienced 2 types, and 16.2% (n = 66) experienced all 3 types. Fifteen percent (n = 132) of participants identified being both a victim and a perpetrator of at least 1 form of relationship violence. Thirty-five percent (n = 322) of participants reported experiencing relationship violence before coming to college, and 24.9% (n = 227) reported experiencing relationship violence during college.
Year in school was not associated with experiencing violence during college regardless of prior experience with violence (P =.06). However, victims during college who also experienced victimization before college were more likely to be in their third or fourth year of school (P =.005). If the risk of victimization during college was solely attributable to time on campus, we would expect those without relationship violence before college to show the same trend; this is not the case.
Victimization was reported by 42.1% (383 of 910) of the students. Among 910 participants, 26.2% (n = 238) experienced emotional victimization, 22.9% (n = 208) sexual, and 17.1% physical (n = 156). Among 383 victims, 57.2% (n = 219) reported victimization of 1 type of violence only, 28.5% (n = 109) of 2 types of violence, and 14.4% (n = 55) of 3 types of violence.
Figure 1 shows reported prevalence of victimization by age. Incidence can be inferred at each age by subtracting the rate for the previous age. It is unclear how participants reporting victimization before age 12 years interpreted the questions. We assume that this childhood relationship violence includes child abuse, bullying, and fighting among friends or siblings. None of the victimization reported before age 12 years was related to sexual violence. Rates of victimization began to rise at age 13 years, about the time adolescent dating relationships begin. Rates rose sharply between ages 15 and 17 years (during high school) and continued to rise between ages 18 and 22 years (during college).
Prevalence of violence victimization by age. Forty-eight victims before college who did not provide an age at first victimization are excluded. *Assumed child abuse, bullying, sibling violence, and others.
Reported victimization rates were higher before college for each form of relationship violence. Of 383 victims, 46.2% (n = 177) were victims before college only, 31.3% (n = 120) were victims before and during college, and 22.5% (n = 86) were victims during college only. Before college, emotional victimization was reported most often, followed by sexual and physical victimization (Table 3). During college, sexual victimization and emotional victimization were reported most often, followed by physical victimization.
Fifty-three percent (277 of 520) of women and 27.2% (106 of 390) of men reported victimization. Women were more likely than men to report being victims of all forms of relationship violence before and during college (Table 3). Although women were more likely to report victimization, the rate of male victimization in our sample was high, with emotional victimization being highest.
Perpetration was reported by 17.1% (156 of 910) of the students. Among 910 participants, 11.4% (n = 104) perpetrated physical relationship violence, 6.3% (n = 57) emotional relationship violence, and 4.1% (n = 37) sexual relationship violence. Among 156 perpetrators, 76.3% (n = 119) reported perpetrating 1 type of relationship violence only, 20.5% (n = 32) 2 types, and 3.2% (n = 5) 3 types.
Similar to victimization rates, perpetration rates were higher before college than during college for all forms of relationship violence. Among 156 students reporting perpetration, 56.4% (n = 88) reported perpetrating before college, 21.2% (n = 33) reported perpetrating both before and during college, and 22.4% (n = 35) reported perpetrating during college only. Before and during college, physical perpetration was reported most often, followed by emotional perpetration and sexual perpetration (Table 3). Almost 1 in 10 students surveyed said they perpetrated physical relationship violence before college, while half as many reported perpetrating physical relationship violence during college. Emotional perpetration rates were more similar before and during college, and few students admitted perpetrating sexual relationship violence before or during college.
Nineteen percent (99 of 520) of women and 14.6% (57 of 390) of men reported perpetrating relationship violence. Before and during college, men were more likely than women to report perpetrating sexual violence, and women were more likely than men to report perpetrating physical violence (P < .001 for both). There were no differences in emotional perpetration by gender before or during college.
Figure 2 shows the relationship between victims and perpetrators for each type of victimization during college. Victims reported that partners perpetrated most physical and emotional violence, followed by friends or acquaintances and strangers. Sexual victims reported that partners, friends, and acquaintances equally perpetrated violence against them. There were no differences between male and female victims' reports of partner perpetration for any type of violence.
Relationship between victims and perpetrators for each type of violence. The number of responses for the victim's description of the perpetrators for physical was 64, for emotional 107, and for sexual 109. The number of responses for the perpetrators' descriptions of victims for physical was 45, for emotional 28, and for sexual 15.
Most perpetrators identified their partner as the victim for each type of violence. Compared with physical and emotional perpetrators, more sexual perpetrators reported victimizing friends and acquaintances. During college, more women (85.4% [41 of 48]) than men (55.0% [11 of 20]) reported perpetrating against a partner (P < .01), but there were no gender differences when exploring specific types of violence.
Table 4 compares students who experienced partner violence vs those who experienced violence in another (ie, nonpartner) relationship during college. Experiencing partner violence during college was associated with being older, female, a physical victim before and during college, an emotional victim before and during college, a physical perpetrator before and during college, and an emotional perpetrator during college (P < .03). In contrast, experiencing nonpartner violence during college was more likely to be associated with sexual victimization during college (P < .001). Although those experiencing partner violence during college were older and in a higher grade, this does not necessarily indicate that partner violence begins later in college. Because we did not capture the age at victimization for college violence, we cannot determine the timing of partner violence relative to other relationship violence. Furthermore, there are no differences in the type of violence experienced by age or by age at first victimization for those with exposure before college.
Twenty students (10 female and 10 male) in our sample reported having same-gender partners during college. No men reported being victims or perpetrators of violence with a male partner. However, 4 women reported being victims of or perpetrators against their female partners.
In our diverse sample of college-age students, we found that relationship violence is a prevalent issue, affecting almost half of the students. Victimization often began before college, with high rates between age 15 and 17 years. Emotional violence was most common before and during college, and sexual violence, while reported before college, increased during college. Although women were more likely to report being victims of violence, male victimization was common. In our sample, men reported more sexual perpetration, while women reported more physical perpetration. During college, partner violence was common. Those experiencing partner violence were more likely to be older and to have experienced physical or emotional violence.
Our rates of partner-specific victimization are similar to those reported in other samples of partner violence among adolescents and young adults.1,4,12,13 Almost half of the students surveyed in this study reported experience with at least 1 form of relationship violence as a victim, perpetrator, or both. Women 24 years and younger have been identified as being at highest risk for violence, particularly partner violence.3,4,30 Our data suggest that adolescent relationship violence frequently happens in earlier adolescence and that education regarding the development of healthy relationships should occur well before college and possibly before high school.
Much attention has focused on issues surrounding physical and sexual victimization of young women; however, emotional violence was the most common form of victimization among men and women in our sample. Emotional violence could be a precursor to other relationship violence.39,40 Because psychological consequences of violence continue throughout young adulthood,41- 43 further studies are needed that explore emotional violence, particularly its relationship to future violence.
During college, sexual violence and emotional violence were equally common. Separation from existing social supports (such as that which occurs in the transition to college) affords many young adults their first opportunity to begin making independent choices without guidance from parents. Because sexuality progresses as adolescents mature,44 college relationships may be the first sexual experiences for many students. This, combined with logistic differences (private living space and no adult supervision), may place adolescents in a vulnerable situation in which they can be victimized more easily.
Female students in our sample were more likely to report being a victim of all forms of relationship violence assessed; however, mounting data, including ours, indicate that male victimization is common.13- 20,26,30 Because men may underreport victimization because of shame, embarrassment, and fear that their masculinity may be questioned,45- 47 rates reported herein may underestimate male victimization. The noted prevalence of male victimization suggests a need for male-friendly or male-specific resources.
Far fewer students in our study reported perpetration of relationship violence compared with victimization. Female and male students seemed to be perpetrating relationship violence at similar rates. Gender differences emerged when comparing specific types of relationship violence, with women more likely to perpetrate physical violence and with men more likely to perpetrate sexual violence. In addition, women were more likely than men to perpetrate partner violence during college. It is unclear whether female perpetration is reflective of women initiating violence within a relationship, female-female violence within same-gender relationships, self-defense, or retaliation for past abuse. Because women typically inflict fewer injuries than men,26,48- 50 they may have less fear of punishment or stigma and consequently may be more comfortable reporting perpetration than men.11,14,18
While victims of sexual violence during college most often identified their perpetrators as acquaintances or friends, sexual perpetrators most often identified their victims as partners. There are several possible explanations. First, although few perpetrators self-identified, more sexual perpetrators were men, and more sexual victims were women. Gender differences surrounding relationship perception may exist. For example, women typically define dating relationships relative to intimacy, while men identify them more casually.51- 53 Second, coping strategies may allow sexual perpetrators to feel less guilt or shame if their victims are partners or individuals with whom they have had sexual relations.54 Third, a prior study9 found that young women who experienced sexual coercion were more likely to be victimized by older partners. Perhaps male college students were more likely to victimize non–college-aged partners. High rates of sexual perpetration by friends, acquaintances, and partners on college campuses speak directly to the need for education and resources related to sexual violence.
There are several important limitations to this study. First, rates of victimization and perpetration were likely underreported because of stigma and fear of punishment.55- 57 It is unclear whether men and women reported differently. Second, student survey involvement was dependent on professor approval. In 1 college, e-mail was used infrequently, resulting in a high nonresponse rate. Third, we did not survey evening classes. Students enrolled in evening classes only may have different rates of violence involvement. However, some students participated in both day and evening classes, and while not surveyed in their evening class, they may have been surveyed in their day class. Although there is potential for selection bias, we had a good balance of surveys at each college across year and course type, and racial demographics of our sample were similar to overall campus demographics. Fourth, the age at first victimization for students was obtained before college only. For those first victimized during college, we used the current age as the age at first victimization. Therefore, the actual age at first victimization during college may be younger than reported herein. Fifth, our subsample of participants in same-gender relationships was small, limiting our ability to perform detailed analyses; future work is needed in this area.
In conclusion, all forms of relationship violence are prevalent among male and female college students; almost half of the students had experienced relationship violence at some point in their lives, more than one-third had experienced violence before college, and one-quarter had experienced violence during college. High rates of victimization and perpetration among men and women in midadolescence suggest a need to assess the availability of high school and college resources for relationship violence prevention and treatment. Emotional abuse was the most frequently reported form of violence experienced by male and female students before and during college. While emotional abuse frequently is not a focus of violence prevention, it can cause poor outcomes and may predispose victims to other forms of violence. Therefore, educational efforts focusing on healthy relationships should begin during childhood.
Although women reported more victimization than men, male victimization was common. More men perpetrated sexual violence, but more women perpetrated physical violence and partner violence. Future studies should explore the needs of male victims, specifically related to the desire for gender-specific services for relationship and partner violence. While most physical and emotional violence occurred in partner relationships, sexual violence was committed in the context of nonpartner relationships such as those with friends and acquaintances. Further research must explore whether this finding is confounded by gender or age and whether different educational campaigns or resources are required to address violence committed by acquaintances and partners.
Correspondence: Christine M. Forke, MSN, CRNP, Craig-Dalsimer Division of Adolescent Medicine, The Children's Hospital of Philadelphia, Ninth Floor, Main Bldg, Room 9416A, 34th and Civic Center Boulevard, Philadelphia, PA 19104 (firstname.lastname@example.org).
Accepted for Publication: December 18, 2007.
Author Contributions: Mss Forke and Myers had full access to all the data in this study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Forke, Myers, and Catallozzi. Acquisition of data: Myers. Analysis and interpretation of data: Forke, Myers, Catallozzi, and Schwarz. Drafting of the manuscript: Forke, Myers, and Catallozzi. Critical revision of the manuscript for important intellectual content: Forke, Myers, Catallozzi, and Schwarz. Obtained funding: Forke, Myers, and Schwarz. Administrative, technical, and material support: Forke, Myers, Catallozzi, and Schwarz. Study supervision: Forke, Myers, and Catallozzi.
Financial Disclosure: None reported.
Funding/Support: This study was supported by the Claneil Foundation, Valentine Foundation, Craig-Dalsimer Fund, Mary D. Ames Chair for Child Advocacy, and Institute for Safe Families.
Additional Contributions: The following members of the Campus Violence Task Force, Institute for Safe Families, assisted with conceptualizing the study, gathering community support, serving as counselors during data collection, and securing funding: Sandra H. Dempsey, MSS, MLSP; Marcia Witherspoon, MSW, LSW; Janice B. Asher, MD; Eileen R. Giardino, PhD, CRNP; Claire A. Washington, MSN, CRNP; and Bettsy McCoubrey, PhD. The Institute for Safe Families forged community partnerships and provided resource cards for participants. Janice Asher donated study incentives. Melissa E. Glassman, MD, MPH, and Lisa K. Tuchman, MD, reviewed and critiqued the manuscript. We thank the professors who allowed us to survey their students and the students who took the time to participate in this study.
Forke CM, Myers RK, Catallozzi M, Schwarz DF. Relationship Violence Among Female and Male College Undergraduate Students. Arch Pediatr Adolesc Med. 2008;162(7):634-641. doi:10.1001/archpedi.162.7.634