Students as pro-social bystanders : opportunities, past behaviors, and intentions to intervene in sexual assault risk situations

Graduation date: 2015Access restricted to the OSU Community, at author's request, from May 8, 2015 - May 8, 2017Sexual assault is a major public health concern in the U.S, and college students are particularly vulnerable to victimization. A health issue that affects nearly one in four women (Fisher, Cullen, & Turner, 2000; Karjane, Cullen, & Turner, 2005) and that is associated with severe negative health outcomes, including depression substance abuse, suicide ideation, and risky sexual behaviors (CDC, 2012), warrants effective prevention programs. Moving away from traditional prevention efforts, which target females as potential victims in risk reduction programs and males as potential perpetrators in attitudinal-shifting programs, bystander engagement programs have become increasingly more widespread. These programs aim to engage all students on the college campus as potential bystanders who can intervene to prevent a sexual assault or reduce the harm of an assault that has already occurred (Banyard, Moynihan & Plante, 2007). Burn (2009) investigated potential barriers to pro-social bystander intervention using the Situational Model of Bystander Intervention, a model based on the original research of bystander behavior of Latanè and Darley (1970). The model outlines five barriers that influence students' intent to intervene as witnesses to sexual assault: failure to notice the situation, failure to identify the situation as high risk, failure to take intervention responsibility, failure to intervene due to skills deficit, and failure to intervene due to audience inhibition (Burn, 2009). She found that students' perception of barriers negatively correlated with intervention behaviors as bystanders to sexual assault (Burn, 2009). \ud Although bystander engagement programs have shown initial promise in increasing students' intent to intervene, more needs to be known about the opportunities students have to intervene, their past intervention actions, and their intent to intervene in the future across the wide range of situations that encompass sexual assault risk. In addition, to develop effective programs that aim to increase pro-social behavior, understanding the salient influences of students' intent is critical. This study uses the Theory of Planned Behavior (TPB; Ajzen & Fishbein, 1991) to examine the influences of students’ intent to perform 12 different pro-social bystander behaviors. The TPB asserts that individuals' behavior is most proximally influenced by their behavioral intentions, and intentions are influences by their perceived behavioral control to perform the behavior, subjective norms that support performing the behavior, and attitudes toward the behavior (Ajzen & Fishbein, 1991). \ud The four primary aims of this study were: 1) to examine the demographic correlates of students' opportunities, past intervention actions, and reported intent to intervene; 2) to examine any differences in students' intent to intervene based on the level of intervention (pre-, mid-, and post-assault) and type of intervention (with the potential or actual victim compared to the potential or actual perpetrator); 3) to examine the influences of perceived behavioral control, subjective norms, and attitudes on students' intent to intervene as bystanders; and 4) to compare the TPB-based model to the Situational Model of Bystander Intervention (Burn, 2009) in its ability to explain students' intent to intervene as bystanders. \ud In the Fall of 2014, a sample of 815 undergraduate students at Oregon State University completed the Sexual Assault Bystander Behavior Questionnaire (SABB-Q), a tool comprised of items to measure students' opportunities, past behaviors, and future intent, in addition to measures assessing the influences of students' intent in line with the Theory of Planned Behavior and Burn's (2009) Situational Model of Bystander Intervention. Students who participate in Greek communities (fraternities and sororities) reported significantly greater odds of having the opportunity to perform four of the 12 intervention behaviors compared to non-Greek students, while student-athletes reported significantly greater odds of having the opportunity to perform two of the 12 intervention behaviors. Females reported significantly more past pro-social intervention behaviors (x̄ = 0.87) compared to males (x̄ = 0.79; p = 0.007). \ud Regarding intent to intervene in the future, females reported significantly greater intent to intervene compared to males (x̄ = 6.07 vs. 5.68; p = 0.007). Students with friends who have been victims of sexual assault reported greater intent to intervene compared those without friends who have been victims (x̄ = 6.04 vs. 5.89; p = 0.02). Students with a personal history of victimization reported significantly greater intent compared to those without a personal history (x̄ =6.13 vs 5.93; p = 0.03). \ud Students reported significantly greater intent to intervene with the potential or actual victim compared to the potential or actual perpetrator (x̄ = 6.19 vs. 5.74, p < 0.001). Females reported significantly greater intent to intervene with both the potential or actual victims and perpetrators (x̄ = 6.31 and 5.84, respectively) compared to males (x̄ = 5.88 and 5.49, respectively). Both males and females reported the greatest intent to perform post-assault intervention behavior (x̄ = 6.23), followed by pre-assault (x̄ = 6.08) and mid-assault behaviors (x̄ = 5.57). Females reported significantly greater intent to perform nine of the 12 pro-social intervention behaviors compared to males. \ud A multiple regression analysis revealed that perceived behavioral control, subjective norms, and attitudes explained a significant proportion of the variance in intent to intervene (R² = 0.55, F(3, 771) = 315.68, p < 0.000). Perceived behavioral control was highly significant (β = 0.48, p < 0.001), as were subjective norms (β = 0.15, p < 0.001) and attitudes (β = 0.30, p < 0.001). Gender differences were also observed. For females, perceived behavioral control was highly significant (β = 0.49, p < 0.001), as were subjective norms (β = 0.15, p < 0.001) and attitudes (β = 0.29, p < 0.001). For males, perceived behavioral control was highly significant (β = 0.49, p < 0.001), as were attitudes (β = 0.29, p < 0.001). However, males' subjective norms were not significantly related (β = 0.07, p = 0.199) to their intent to intervene. Further analysis revealed a significant interaction between gender and subjective norms (β = -0.28; p = 0.039). The TPB-based model including this moderation effect explained a significant proportion of the variance in students' intent to intervene (R² = 0.57, F(6, 766) = 168.46, p < 0.000).\ud Interveners reported significantly greater perceived behavioral control than non-interveners for seven of the 12 intervention behaviors; more supportive subjective norms than non-interveners for six of the 12 intervention behaviors; more positive attitudes than non-interveners for only one of the 12 intervention behaviors; and greater intent to intervene in the future for six of the 12 intervention behaviors. However, differences in the three TPB variables between interveners and non-interveners were not consistent for the 12 intervention behaviors. \ud Regarding Burn's (2009) Situational Model of Bystander Intervention, a multiple regression analysis revealed two of the five barriers were significantly related to students’ intent to intervene: the failure to take intervention responsibility barrier (β = -0.29, p < 0.001) and the failure to intervene due to audience inhibition barrier (β = -0.22, p < 0.001). The model in whole explained a large proportion of the variance (R2 = 0.25, F(5, 768) = 50.14, p < 0.000). Gender differences were also observed. For females, failure to take intervention responsibility (β = -0.23; p < 0.000) and failure to intervene due to audience inhibition (β = -0.23; p < 0.001) both had a significant, negative influence on their intent to intervene. For males, failure to take intervention responsibility (β = -0.21; p < 0.014) had a significant, negative influence on intent to intervene. Additional analysis revealed no significant interactions between gender and any of the five barriers. \ud The TPB-based model explained a greater proportion of the variance (R2 = 0.55) compared to Situational Model of Bystander Intervention (R² = 0.25) in the multiple regression analysis using all 12 intervention behaviors. All three variables in the TPB-based model were significantly related to students’ intent, whereas only two of the five barriers were significantly related. A final multiple regression analysis was conducted using all three significant TPB variables and the two significant barriers to explain students' intent to intervene. The combined model explained a significant proportion of variance in students' intent (R² = 0.58 F(5, 756) = 206.19, p < 0.000) and significantly improved upon the TPB-based model (Δ R² = 0.03; p < 0.000). \ud The results of this study have several implications for future research and public health practice. First, it is important to ask students about their opportunities to intervene in addition to their actual intervention behaviors because this information helps paint a clearer picture of bystander engagement. This assessment could also help identify high-risk groups: students who have greater opportunities to intervene as bystanders and/or report fewer intervention behaviors compared to their reported opportunities. Second, students may conceptualize intervention behaviors differently depending on the phase of the assault and with whom the intervention behavior requires intervening. Accordingly, programs aimed at encouraging students to intervene should take these differences into consideration. Third, the Theory of Planned Behavior, used to explain and change other health-related behaviors, can effectively be applied to help uncover determinants of pro-social bystander behaviors. Perceived behavioral control, subjective norms, and attitudes appear to be salient influences in students' intent to intervene. Therefore, bystander engagement programs should incorporate activities to heighten students' skills to intervene, change social norms that support bystander intervention, and shift attitudes toward the benefits of intervening. This study demonstrates the importance of using an established, evidenced-based theoretical framework to explain behavioral influences and strengthens the argument for continued use of theory to identify, and potentially change, salient influences in behavioral performance. \ud Students as pro-social bystanders have the potential to make a positive impact on the reduction of sexual assault on the college campus. Although the responsibility for sexual assault rests on those who perpetrate such acts, and primary prevention strategies aimed at those demonstrating a risk for perpetration are imperative, sexual assault is a public health issue that warrants a multi-pronged approach to reduce its incidence and migrate its associated harms. Programs that engage students as pro-social bystanders have the potential to make a positive impact on the reduction of sexual assault incidence in the absence of effective primary prevention strategies. The findings of this study make a contribution to the literature examining influences of students' pro-social bystander intervention to sexual assault situations and provide suggestions for strategies to increase bystander engagement