Abstract
The White House Task Force to Protect Students from Sexual Assault recommends bystander training as part of campus sexual assault prevention efforts. The current study sought to understand salient themes among students’ qualitative responses for why they did not intervene in sexual assault risk situations. In 2014, undergraduate students (N = 9,358) at a large public university completed a web-based survey to assess bystander opportunities and responses for six risk situations. Content coding analysis indicated that students report several unique barriers to intervention. These findings have important implications for bystander training programs, as well as future research on bystander behavior.
Introduction
Campus sexual assault is estimated to affect one in five women by the time they graduate from college or university (Cantor et al., 2015; Krebs et al., 2016). Despite research indicating such high rates of violence dating back several decades (Koss, Gidycz, & Wisniewski, 1987), political movements to address this major public health issue received little momentum until recent years. Coupled with increasing awareness of campus sexual violence, a push toward evidence-based prevention programming, often in the form of bystander intervention programs, emerged. In 2013, the Campus Sexual Violence Elimination (SaVE) Act required that all institutions of higher education provide sexual assault prevention training (Clery Center for Security on Campus, 2016). The following year, the White House Task Force to Protect Students from Sexual Assault was established and issued the Not Alone Report recommending bystander intervention programs to curb campus sexual violence (The White House, 2014).
Given the current push to implement bystander training programs, it is important to understand the barriers students face in their decision to engage in prosocial intervention. Such research is emerging (Bennett, Banyard, & Garnhart, 2014; Burn, 2009; Katz, Pazienza, Olin, & Rich, 2015; McMahon, 2010; Pugh, Ningard, Vander Ven, & Butler, 2016), and the body of literature would benefit from further investigation, particularly using qualitative methods.
Bystander Intervention Training
Latané and Darley articulated the theory underlying bystander intervention programming in 1970. The five-step process bystanders must go through to intervene includes (a) notice the situation, (b) identify the event as one appropriate to act on, (c) take responsibility for intervention, (d) decide to help, and, last, (e) intervene as a prosocial bystander. Using this model, previous research sought to understand barriers to acting as a bystander at each level of the five-step model (Bennett et al., 2014; Burn, 2009; Latané & Darley, 1970). For example, after noticing a risk situation, individuals must perceive risk to warrant their intervention, and a lack thereof results in a failure to take responsibility for intervention. In the context of sexual assault risks, as bystanders perceived more barriers to intervening, such as lacking the skills necessary to intervene after taking responsibility, their intent to intervene decreased (Burn, 2009).
Bystander training programs aim to prevent and address sexual violence at multiple stages and in both low- and high-risk situations (McMahon & Banyard, 2012). The stages in which a prosocial bystander can intervene include (a) preventing sexual assaults from occurring, (b) disrupting sexual assaults as they are happening, (c) actions taken after sexual violence has occurred such as supporting survivors, and (d) challenging cultural norms that promote or support sexual violence (Banyard, Moynihan, & Plante, 2007). Intervention techniques used at these various stages may include the “Four Ds” of direct, distract, delegate, and delay (Banyard, Plante, & Moynihan, 2005; Berkowitz, 2002). Direct interventions include confronting a potential perpetrator or the situation in some way through directly addressing the incident that is happening. Distract techniques involve averting the potential perpetrator’s attention to prevent sexual violence. Delegation includes asking others for help. Finally, delay tactics include those used after an assault has occurred, such as providing resources or support.
Current research evidence demonstrates the positive effects training programs have on bystander attitudes and beliefs surrounding violence, intervention willingness or intentions, and sexual violence perpetration or victimization rates on campus (Coker et al., 2017; Coker et al., 2015; Katz & Moore, 2013). Yet, research on actual bystander intervention behaviors is less common (Brown, Banyard, & Moynihan, 2014; Katz & Moore, 2013), which may be due, in part, to issues specific to the measurement of actual prosocial bystander behaviors. First, students’ intervention behavior must be assessed relative to their opportunities to intervene (Jouriles, Kleinsasser, Rosenfield, & McDonald, 2016; McMahon, Palmer, Banyard, Murphy, & Gidycz, 2017; Moynihan et al., 2015). Second, investigating students’ reasons for not intervening can guide programming efforts aimed at alleviating barriers students may face.
Research on Barriers to Bystander Intervention
Emerging research has demonstrated that students face a number of barriers to intervening in risk situations related to sexual violence (e.g., Amar, Sutherland, & Laughon, 2014; Exner & Cummings, 2011; Katz, Colbert, & Colangelo, 2015; Katz, Pazienza, et al., 2015; Pugh et al., 2016). Substantial evidence points to bystanders’ perceived lack of intervention skills (Bennett et al., 2014; Brown et al., 2014; Burn, 2009) or the influence of normative beliefs about the appropriateness, or peer acceptance, of intervention (Austin, Dardis, Wilson, Gidycz, & Berkowitz, 2016; Berkowitz, 2010; Brown et al., 2014). The bulk of the literature is heavily quantitative, however, which limits students to respond to predetermined descriptions of their behaviors. Employing more qualitative methods may yield a deeper understanding of bystander response (Edwards, Rodenhizer-Stampfli, & Eckstein, 2015). For example, previous qualitative work demonstrates that students often find creative ways to intervene not necessarily presented in close-ended survey options (McMahon, Lowe Hoffman, McMahon, Zucker, & Koenick, 2013). Other qualitative studies have examined bartenders’ perceptions of sexual violence and barriers to prosocial intervention within a community setting (Powers & Leili, 2016) and high school students’ perceptions of barriers to acting as a prosocial bystander in situations of dating and sexual violence (Edwards et al., 2015). Among college students, Bennett and colleagues (2014) used mixed methods to understand barriers students face when intervening to stop sexual violence and found students’ barriers to fit within the scope of the five-step bystander intervention model (Burn, 2009; Latané & Darley, 1970). In this latter study, participants were asked to provide potential barriers to intervention when students “think about helping someone in a situation” (Bennett et al., 2014, p. 485), rather than specifically for opportunities they have had to intervene. Thus, reported barriers included those for situations that are potentially hypothetical and provided by students who may never have had opportunities to intervene as bystanders.
Important to consider in the investigation of barriers to bystander intervention is the role alcohol plays in sexual assault risks and the ability for bystanders to identify such. The use of alcohol in facilitating sexual activity is common for college students (Vander Ven, 2011). Much evidence demonstrates a strong relationship between the consumption of alcohol and sexual assault (Abbey, 2002; Krebs et al., 2016; Krebs, Lindquist, Warner, Fisher, & Martine, 2009). Research continues to find confusion among college students in determining consent in sexual experiences, particularly when alcohol is involved (Borges, Banyard, & Moynihan, 2008; Pugh et al., 2016; Ward, Matthews, Weiner, Hogan, & Popson, 2012). Pugh and her colleagues (2016) qualitatively explored perceived barriers to bystander intervention in relation to alcohol use. They found that victim blaming for drinking or prior sexual activity and victim ambiguity in which either the relationship between the parties involved was unknown or the situation at hand was unclear—reflecting the belief that alcohol and sex are interlinked within a campus setting—might prevent potential bystanders from intervening. Given the intersection of alcohol and sexual assault, it is important to investigate bystanders’ intervention barriers for alcohol-related risks and whether potential victims’ or perpetrators’ consumption of alcohol presents unique barriers for bystanders who could intervene to prevent sexual violence.
The Current Study
For bystander training programming to better prevent and address sexual violence, it is essential to understand what barriers confront bystanders and systematically address such during program development and implementation (Bennett et al., 2014; Burn, 2009). Therefore, it is necessary to engage in more nuanced investigation of why students who have opportunities to intervene decide not to. To address this gap in the existing literature, we sought to examine the reasons why students do not intervene in a variety of risk situations, employing both quantitative and qualitative analysis of student survey responses.
Method
Procedures
Participants for this study were recruited at a large public university in the Northeast in 2014. All enrolled students, both undergraduate and graduate students, were invited to participate in the survey that was part of a larger campus climate assessment on sexual violence conducted during that academic year. An intensive outreach campaign was conducted and included emails, print materials, social media messages, and informational stations in locations frequented by students (e.g., dining halls). The survey was online and anonymous, and all participants were eligible for randomly selected cash prizes as an incentive for participation. All materials and procedures were approved by the University’s Institutional Review Board.
Sample
Participants were all undergraduate students (N = 9,358). Those younger than 17 years (n = 109) and older than 25 years (n = 384) were excluded from the analysis, as well as those who failed to indicate their age (n = 203). See Table 1 for sample demographics.
Sample Demographics.
Bystander Intervention Opportunity and Behavior Measures
We assessed bystander intervention behaviors with six items from the Bystander Behavior Scale (McMahon et al., 2014). Students were first asked, “Since attending the institution, (a) Have you seen a group of students sexually intimidating/bothering someone in a parking lot or similar situation, (b) Have you seen a girl taking a drunk guy back to her room, (c) Have you seen a guy taking a drunk girl back to his room, (d) Have you seen a girl you didn’t know go to her room with a group of guys and heard her yelling for help, (e) Have you heard a friend say they planned to give alcohol to someone to get sex, and (f) Have you heard rumors that a friend forced someone to have sex?” Response options were “Yes”/”No.” When students reported “Yes” for any of the intervention opportunities noted above, one follow-up item asked participants, “What did you do?” Response options included (a) did nothing, it wasn’t my business, (b) did nothing because I wasn’t sure what to do, (c) confronted the situation directly, (d) went and got assistance from someone else, and (e) other (please specify).
Students’ Other Bystander Intervention Barriers
We exported students’ qualitative responses from Stata 14 into an Excel spreadsheet for analysis and used an iterative, inductive content analysis to explore the “other” bystander response findings, with a focus on understanding additional barriers to intervening. Due to the small sample (3.57%, n = 1) of students reporting “other” as their response as bystanders for one situation (“Have you seen a girl you didn’t know go to her room with a group of guys and heard her yelling for help”), we excluded it in the qualitative analysis.
The original quantitative bystander response items included two barriers from Burn’s (2009) decisional model, skills deficit (“Did nothing because I wasn’t sure what to do”) and lack of responsibility (“Did nothing, it wasn’t any of my business”). Thus, the analysis was guided by the understanding that students may report barriers in line with the second step in prosocial bystander response, risk identification (Burn, 2009), as well as the understanding that alternative barriers, outside of the decisional model, may also be reported. First, one coder reviewed participants’ text entries and developed thematic codes corresponding to students’ expressed barriers to bystander intervention. To ensure rigor of the methods used (Patton, 2002), a second member of the research team independently analyzed and coded a subset of the qualitative data. Interrater reliability showed high agreement (87.5%). After review and discussion, the coders reached consensus, confirming seven salient themes that emerged during analysis. We present quotes from participants’ qualitative responses that illustrate the barrier theme and did not revise grammar or spelling. Salient codes, barriers that were commonly reported by students, are presented as intervention barrier themes. In some instances, we quoted a word or phrase commonly reported and did not attribute it to a single student.
Results
Students’ Bystander Intervention Opportunities and Behaviors
Table 2 presents the percentage of students who reported opportunities to intervene and of those who had the opportunity, the response items for each of the six bystander situations. Opportunities to intervene varied between situations. For example, 35.24% (n = 2,760) of students “saw a guy taking a drunk girl back to his room” compared with 3.58% (n = 28) who “saw a girl you didn’t know go to her room with a group of guys and heard her yelling for help.” Students’ barriers for bystander intervention also varied between situations. For example, of those who “saw a group of students sexually intimidating/bothering someone in a parking lot or similar setting,” 8.93% (n = 59) “did nothing, it wasn’t my business,” 39.79% (n = 263) “did nothing because I was unsure what to do,” and 6.2% (n = 41) reported “other” barrier. Of those who “saw a girl taking a drunk guy back to her room,” 56.7% (n = 1,003) “did nothing, it wasn’t my business,” 24.42% (n = 432) “did nothing because I was unsure what to do,” and 10.63% (n = 188) reported “other” barrier. Chi-square tests of independence did not uncover consistent, significant relationships between student demographic variables (i.e., gender identity, racial identity, and year in school) and reporting “other” barriers.
Students’ Bystander Intervention Response to Sexual Assault Risk Situations (n = 8,665).
Note. Percentages may not add up to 100% because there were other text entry responses that were not coded as a barrier to intervention.
Students’ Other Bystander Intervention Barriers
In the qualitative analysis of students’ responses for “other” bystander responses, we sought to understand barriers students may face for each bystander situation beyond the two provided in the survey items. The following presents the results for the themes that emerged for the five bystander situations (see Table 3). To focus specifically on the qualitative data, we chose not to include quantitative analysis of demographic differences in students who reported “other” bystander response, but we did not observe patterns of differing findings based on demographic variables. For each statement provided in the text, we reported the participants’ sex/gender identity and age to illustrate that common barriers were expressed among students of different ages and gender identities.
Themes for Students’ Intervention Barriers and Corresponding Risk Situations.
Theme 1: Did nothing because I did not perceive risk for sexual assault
The first salient theme that emerged was students’ perception that there was no risk for sexual assault. Whether due to the alcohol involved in the situation or the lack of severity perceived in the situation, a failure to identify risk was a central barrier to students’ prosocial intervention. Across several risk situations, students reported no intervention because they perceived no risk for sexual assault, yet the rationale students provided to support this perception varied depending on the situation. We organized the results on students’ risk perception according to the specific risk situation students encountered.
A lack of risk perception for sexual assault, as reported by students, emerged from what we observed as students’ normalization of the situations, particularly those involving alcohol, such as “witnessed a girl taking a drunk guy back to her room” or “witnessed a guy taking a drunk girl back to his room.” A female student, aged 21, reported she “did nothing, because it seemed like a normal event at parties.” A large proportion of students cited this as the rationale for not intervening, with the implicit suggestion that they did not perceive risk for sexual assault (e.g., a female student, aged 18, reported, “I was also drunk and at a party, so I didn’t know if it mattered.”). Students reported inaction because of perceiving the situation as “no biggie” (male student, 20), “nbd [no big deal]” (male student, 19), “what happens [when] horny 20 year olds drink together” (male student, 21), and otherwise normalized, such as “half of [this institution] has drunk sex” (male student, 19) and “happens all the time” (female student, 19). A female student, aged 20, reported, “I regret not doing anything now. At the time I was a freshman and believed it was my job to get used to the hookup culture at [this institution].” Similarly, another student specifically referenced “hookup culture” and lack of perception of risk for assault. Another female student, aged 21, reported she did not intervene because she “didn’t think it was a problem, casual hookup.”
Similar to the perceived normalization of someone taking a drunk person back to their room, students also suggested that both parties being drunk challenged their perception of risk for sexual assault. This was, again, particularly relevant when students “witnessed a girl taking a drunk guy back to her room” or “witnessed a guy taking a drunk girl back to his room,” and that a lack of risk perception focused on the presence of alcohol for both individuals. For instance, a number of students reported they did not intervene in these situations because “they were both drunk,” reported by both male and female students of differing ages, suggesting that students either cannot determine a nonconsensual situation or do not perceive risk when both individuals are intoxicated. We also observed, similar to the situations where a drunk person was being taken back to someone’s room, that students perceived no risk for assault when their friend “planned to give alcohol to get sex” because that statement is commonplace among their peers. For example, a male student, aged 20, reported, “this seems like a common known strategy, not secret, not out of the ordinary,” and similarly, another male student, aged 23, reported, “Isn’t that what all college people do?”
Another rationale for lack of intervention, related to risk identification, was students’ reports that the situations when they “saw a girl taking a drunk guy back to her room” or “saw a guy taking a drunk girl back to his room” were not sexual assault situations, but rather a strategy to remove the drunk guy (or girl) from the party. In some instances, bystanders assumed to know the intent of the girl, such as “did nothing because I thought she was taking care of him” (male student, 21), “I thought she was just taking him to sleep” (female student, 18), and “assumed it was her boyfriend and she was taking care of him” (female student, aged 19). Alternatively, a female student, aged 21, reported no intervention in the situation she witnessed on the basis that “they were friends [and] she was taking care of him,” indicating a stronger sense of knowing the intent of the individuals involved. Although cited less frequently than in the situation involving a drunk guy being taken upstairs by a girl, students reported that the intent of the “guy taking a drunk girl back to his room” was also to take the girl upstairs for safety. A female student, aged 20, reported, “they were friends and he took care of her.” Several other students cited a lack of intervention on the basis that, at least implicitly, they understood the guy to be taking “care” of the drunk girl.
Students’ perception of risk, or lack thereof, also emerged as an intervention barrier when they “heard a friend say he planned to give alcohol to someone to get sex.” A female student, aged 20, reported she “did nothing because I thought the person was interested in my friend. [I] thought it would facilitate them getting to know each other. I realize how wrong that is now.” A male student, aged 21, reported he “did nothing, comment mentioned not in earnest,” whereas another reported he likewise did nothing “because there was no implication of intended coercion or violence.” Others reported that when they have heard a friend say they planned to give alcohol to someone to get sex, they interpreted it as “a joke,” both explicitly and implicitly, with students reporting they “laughed” (male student, 21 and female student, 22). A male student, aged 21, reported he did not intervene because his friend “was joking.” Students said of this statement, “it was a joke” (female student, 20) or “assumed it was a joke” (male student, 24).
Theme 2: Did nothing because of the relationship between the individuals
Students commonly reported the context of the relationship between the individuals as a barrier to intervention, specifically when they “witnessed a girl taking a drunk guy back to her room,” such as “the two individuals seemed to be close friends” (male student, 22). Similarly, the context of the relationship was also a barrier when students “witnessed a guy taking a drunk girl back to his room,” such as “I was informed they were in a romantic relationship” (female student, 20). Students indicated that the drunk guy in the situation they encountered was the girl’s “boyfriend,” or that the individuals were “a couple,” “friends,” or “talking” (“talking” was provided by the student in quotes to indicate the status of the relationship). Although students cited the romantic nature of the relationship between the individuals as a barrier for intervening, students also reported not to intervene based on a lesser relationship between the individuals: that they “knew” each other. For instance, a female student, aged 18, reported she did not intervene, but rather she “observed and made sure they knew each other.” Although many students reported inaction based on the individuals’ known relationship status, others noted they “assumed” the individuals knew one another (i.e., female student, 21: “Did nothing, I assumed they knew each other and they knew what they were doing”).
Theme 3: Did nothing because I assessed the potential victim for risk of assault
The third theme emerged as students expressed assessment for risk by looking for cues from the potential victim. Again, this intervention barrier was particularly salient when students “witnessed a girl taking a drunk guy back to her room.” For example, a female student, aged 22, reported she did not intervene because “he was drunk but not too drunk to know what he was doing.” Similarly, other students assessed risk by observing the intoxicated man in the situation: “he was only slightly tipsy” (female student, 21) and “he was drunk but coherent and aware of what he was choosing to do” (female student, 22). Students expressed similar assessments when they “witness a guy taking a drunk girl back to his room.” Students’ perception of risk, or lack thereof, was again based on the level of intoxication of the intoxicated women in the situation. Students’ responses suggested an assessment of the intoxicated woman’s ability to provide consent by observation of her to be “inebriated, but coherent” (male student, 22), “able to speak coherently and make decisions” (male student, 29), “clearly not incapacitated” (female student, 20), “able to walk” (male student, 20), “only a little tipsy and still able to give consent” (female student, 21), and “not too drunk to not know what was going on” (female student, 22). Similarly, a female student, aged 22, reported she did not intervene because the intoxicated woman in the situation she witnessed “was drunk but coherent and aware of what she was choosing to do.”
Also serving as a barrier to perceiving risk for assault, students reported a lack of risk perception based on the interest gauged among the individuals. For example, when students saw “a girl taking a drunk back guy to her room,” they reported, “he was pretty happy about it” (male student, 20) and “[he] seemed enthusiastic” (female student, 21). Students noted similar assessment of risk when they witnessed “a guy taking a drunk girl back to his room.” Some observed active cues of participation, such as “the girl willfully went with him” (female student, 22), and others reported no intervention because “[it] looked like she wanted to go with him” (male student, 22). Several students commented on their assessment of “enthusiasm” of either the intoxicated person or both parties in the situation they witnessed. Conversely, other students reported no intervention on the basis of an absence of cues, specifically from the potential victim, such as “there seemed to be no signs of unhappiness or discomfort or force” (female student, 20). A number of students reported they did not intervene because “she wasn’t resisting.”
For both situations previously discussed, students explicitly stated “consent” between the individuals involved as the rationale for not intervening; however, students’ determination of consent appeared to be based on subjective measures. Numerous students indicated they interpreted the situation as “consensual” in that it either “looked” (male student, 24) or “appeared” (male student, 24) consensual, or students reported that their assessment of the level of intoxication of either party did not preclude the ability to provide consent.
Theme 4: Did nothing because I could not directly intervene
To some degree, the time that had passed or the physical space in between students and the risk situation, as bystanders, provided a barrier to direct intervention. This was expressed particularly when students “heard rumors that a friend forced someone to have sex.” Many students reported a lack of action upon hearing such a rumor on the basis that the assault, whether they believed it to be true or not, was not recent, such as “this happened before I knew either of them” (Gender non-binary student, 23) or “I was not aware of the incident at the time it happened” (female student, 18). Students noted they heard the rumor “after the fact” (female student, 20) or “too late” (female, 19). In addition, students cited the physical space as a barrier to intervention, particularly when they “witnessed a group of students sexually intimidating/bothering someone in a parking lot or similar setting.” For example, a male student, aged 22, reported no intervention in the situation he witnessed because it was “catcalling from a moving car.” Similarly, a female student, aged 18, reported, “It was a passing catcall and the car sped away, so there wasn’t anything to do.” In these two situations in particular, bystander intervention assumes that the specific time and physical space allow for it. Without allowing a further rationale for intervention, failure could indicate such intervention was not possible.
Theme 5: Did nothing because intervening would be unsafe
The fifth barrier theme that emerged was students’ expression of concern for their safety. Although we only observed this barrier when students “witnessed a group of students sexually intimidating/bothering someone in a parking lot or similar situation,” safety appeared a salient barrier to taking action for students who encountered this situation. For example, a female student, aged 21, reported she did not intervene because “I didn’t want to be a target,” and another reported she “did nothing because it wasn’t safe to do anything.” A male student, aged 21, similarly reported, “I knew if I were to do something it would have resulted in putting myself in physical danger as well as the person already being bothered.” A female student, aged 21, reported not intervening because “it was late at night on a Friday and it was a group of males v. group of females so I stayed out of it.” Another student also noted the perpetrators being in a group: “[I] tried to confront them myself. But I was outnumbered by a group of males and thought they would hurt me if I called the cops” (female student, 19) and “im [sic] small and didn’t [sic] get involved for safety” (female student, 20).
Theme 6: Did nothing because the situation did not escalate
Finally, a sixth intervention barrier theme centered on students’ reports of waiting for the situation to escalate. For example, a male student, aged 20, reported, “[I] observed to make sure the situation didn’t become out of hand.” This barrier was particularly salient for inaction when students “witnessed a group of students sexually intimidating/bothering someone in a parking lot or similar situation.” A female student, aged 20, reported she “stayed close by in case things got worse. It ended quickly.” Although the severity of the sexual intimidation in students’ specific encounters is unknown, the salience of students’ perception that there was little risk when the situation was described in this way suggests that “sexual intimidation” may not warrant intervention. Many students reported they “observed” or “watched” the situation, suggesting that if it did escalate, they would have intervened. A female student, aged 18, reported, “I watched cause [sic] I wasn’t sure if the guys would go through with it.” For students who “heard their friend say they planned to give alcohol to get sex,” a lack of escalation was also reported as an intervention barrier. A female student, aged 21, reported she “watched them for the rest of the night to make sure nothing happened (and nothing did happen).” Similarly, a male student, aged 22, reported not intervening because the statement “was never acted upon.”
Theme 7: Did nothing because it was already being dealt with
The seventh intervention barrier theme for students was that risk was not imminent, particularly when they “heard rumors that a friend forced someone to have sex.” For example, students reported the situation was “already handled” (male student, 21), “dealt with” (male student, 21 and female student, 18), “under investigation” (male student, 20), “already reported to the authorities” (male student, 22), “taken care of” (female student, 18), or otherwise in a place that did not necessitate further intervention. A female student, aged 18, reported, “I heard rumors AFTER [emphasis added by student] said friend had been arrested under suspicion of sexual assault, so all I could do is hear everyone’s side of the story,” whereas another reported, “the rumor was unclear and apparently reported to [this institution] already.” A male student, aged 20, reported, “they had already settled it with an RA; so I did nothing.” Another male student, aged 21, did not intervene because the “situation was already handled by the female.” This last expression of the barrier is similar when students “witnessed a group of students sexually intimidating/bothering someone in a parking lot or similar situation,” where the victim or someone else had already intervened. A female student, aged 21, reported no intervention because “her friend yelled at them and they stropped [sic] catcalling her.” A male student, aged 21, reported he did not intervene because “she handled it by walking away.” Similar to Theme 4 described above, where time and physical space provided a barrier for students’ prosocial intervention, students’ lack of intervention due to these barriers does not necessarily preclude an intention to intervene.
Discussion
Implications for Sexual Assault Prevention
The current study sought to understand why bystanders do not intervene when they witness sexual assault risk situations employing both quantitative and qualitative analysis of student survey responses. Students report several salient barriers for a lack of intervention, aligning with, though more nuanced than, the decisional model of bystander intervention (Burn, 2009). For the situations examined here, quantitative analysis found the majority of students who reported not to have intervened cited that it was not their responsibility (“wasn’t my business”) or due to a skills deficit (“unsure what to do”). These barriers suggest the need for programmatic strategies to increase students’ sense of responsibility for assault prevention and build the necessary skills to intervene effectively. Although a smaller proportion of students reported barriers other than those provided in the measures, the qualitative findings of this study indicate that students face a variety of barriers that similarly point to developing programmatic strategies to overcome them, as well as underscore the benefit of using more nuanced measures for assessing bystander intervention behavior.
Perceiving risk in the situation, once noticed, is paramount for engaging in prosocial helping behavior (Burn, 2009), replicated by Bennett and colleagues (2014). The results of this study demonstrate the salience of a lack of risk perception among students as a barrier to intervening when presented with the opportunity to do so, which was articulated in different ways. Although the lack of risk perception arose across the five situations examined here, it was particularly salient when students saw someone taking an intoxicated person back to their room. Katz, Colbert, and Colangelo (2015) note the ambiguity of this situation specifically, in the context of “party rape” (Armstrong, Hamilton, & Sweeney, 2006). Students’ responses here support that perceiving risk for assault in these situations is challenged by the normalization of these situations in campus culture, such that students may perceive comments made about intoxicating someone to get sex as “a joke” or, alternatively but similarly as problematic, a common strategy for sex.
Indeed, many students reported having the opportunity to intervene in these two situations, which supports the notion that alcohol and sex are embedded within the campus social landscape as asserted in previous scholarship (Abbey, 2002; Armstrong et al., 2006; Vander Ven, 2011). Ultimately, the majority of these situations went without intervention. What is known about the intersection of alcohol and sexual assault empirically perhaps does not translate into education of such risk situations as prime opportunities for assault prevention or is universally challenged by the campus culture. The commonality of such situations may result in the normalization of potential risk that challenges the ability for students to detect reasons to intervene (Pugh et al., 2016), despite encouragement of scholars to emphasize the relationship between alcohol and sexual assault as part of prevention programming (Armstrong et al., 2006; Mohler-Kuo, Dowdall, Koss, & Weschler, 2004; Pugh et al., 2016; Rozee & Koss, 2001; Vander Ven, 2011). In this same vein, it is important to note that the bystanders’ own intoxication may limit their ability to recognize cues that indicate a lack of consent or impede their ability to assess risk.
Given the role alcohol plays in campus sexual assault (Abbey, 2002; Mohler-Kuo et al., 2004), it is unsurprising that students’ risk assessment for assault was based, in part, on the level of intoxication of the potential victim. However, students’ assessment based on physical cues was often misguided, again pointing to a misconception about individuals’ ability to provide consent for sexual activity when under the influence of alcohol. According to the Burn (2009) model for bystander intervention, students must interpret the situation as one necessitating intervention, and here, the findings suggest that students assess the potential victims for those signs. Students’ perception that being “able to walk” is a sign of cognitive capacity to provide consent is problematic in light of the documented impact of alcohol on cognitive capacity perhaps undetectable to students who cited individuals as appearing “coherent” (Hindmarch, Kerr, & Sherwood, 1991; Peterson, Rothfleisch, Zelazo, & Pihl, 1990). Students also cited both individuals’ intoxication as an intervention barrier. This confusion on consent, discussed by Borges and colleagues (2008), suggests that absent a sober individual, particularly someone who is unknown to the potential perpetrator, sexual activity cannot be predatorial, despite the research showing that men who sexually assault were more likely to have consumed alcohol prior, and that sexual assault commonly occurs between individuals who know one another (Abbey, 2002).
Students in this study also suggested that an absence of an individual’s resistance is tantamount to consent, which aligns with previous scholarship finding this belief is held among men who perpetrate sexual assault (Rozee & Koss, 2001). Experts have spoken to the complexity of sexual consent, the perception of which can be influenced by factors such as gender, relationship history, and intoxication of both parties (Borges et al., 2008; Humphreys, 2007; Jozkowski, Peterson, Sanders, Dennis, & Reece, 2014; Ward et al., 2012), and that legal definitions may not match students’ conceptualizations (Beres, 2007). Just as consent for sexual activity is challenging to interpret between the individuals involved, those who witness risk situations are often left to interpret consent based on nonverbal cues, or contextual factors (i.e., relationship history/status), even in the absence of alcohol. Research has found that men and women perceive friendliness differently, and men may misinterpret women’s friendliness as sexual interest (Abbey, 1982, 1987; Jacques-Tiura, Abbey, & Parkhill, 2007). As such, bystanders may also rely on these assumptions when observing people who know each other or are in a relationship (Pugh et al., 2016). Students may erroneously equate friendliness with consent.
Furthermore, the consumption of alcohol may impede a victim’s ability to provide cues to bystanders that help is needed (Banyard, 2015). Programming should emphasize this to encourage that, regardless of how the intoxicated individual presents, the mere consumption of alcohol should prompt intervention, such as checking in with the individual to suggest not proceeding upstairs. Previous research shows that past sexual violence victimization correlates with greater knowledge of consent (Fantasia, Sutherland, & Fontenot, 2015); thus, potentially bystanders with less knowledge about sexual consent are less able to identify potential victimization in situations they witness, particularly when alcohol is involved. Any successful sexual assault prevention programming, be it bystander engagement programming or otherwise, should include strategies that focus not only on how to give and receive sexual consent but also how alcohol prohibits the granting of consent (Borges et al., 2008; Jozkowski et al., 2014).
Compounding the problematic nature of how students conceptualize consent, another barrier salient to lack of intervention was students’ perception or knowledge of the relationship between the individuals, which has shown to be a barrier for bystanders to intervene in other forms of interpersonal violence (Pugh et al., 2016; Shotland & Straw, 1976). That students did not intervene because the individuals knew one another or were in a relationship demonstrates the misconception that students are unlikely to be assaulted by someone they know, especially someone with whom they are in a relationship, consistent with the work of Pugh and colleagues (2016). This barrier is problematic given that most victims, in fact, know their perpetrator (Koss, Dinero, Seibel, & Cox, 1998; Wolitzky-Taylor et al., 2011) and that many college students report experiencing sexual assault, including rape, in the context of a relationship (Hoxmeier, 2016; Wolitzky-Taylor et al., 2011).
Another important aspect of barriers students expressed that has implications for practice is the concern for safety, an important consideration for the encouragement of prosocial intervention (Banyard, 2011). This may be particularly relevant when the perpetrator is actively physically violent and/or there are multiple perpetrators and the bystander could be in harm’s way. That said, students might also interpret prosocial intervention as direct confrontation exclusively, which may increase concern for safety in these situations specifically. Bystander engagement programming should echo and emphasize what scholars consider as indirect intervention strategies, such as getting assistance or delegating the intervention (Banyard et al., 2005; Berkowitz, 2002). Although Bennett and colleagues (2014) aligned students’ concern for safety in hypothetical risk situations with skills deficit in the decisional model (Burn, 2009), it is interesting to note that students here were given “did nothing because I was unsure what to do” as a reason for not intervening and chose not to endorse it. Thus, although from a research standpoint we may consider students who expressed concern for their safety as lacking indirect intervention skills, such as calling campus police for assistance, students appear to conceptualize this barrier as unique from a skills deficit, or perhaps, bystanders may still perceive indirect intervention as a threat to their safety.
Implications for Future Research
Currently, the majority of bystander behavior research uses measures to assess prosocial intervention with “yes/no” response options. The reliance on such measures may yield inaccurate results if students do not perceive risk for sexual assault and/or interpret the question as one of risk. To ask students whether they intervened in a situation may erroneously assume that students interpret risk in the situation, such as in the case where a person is taking an intoxicated individual to their room so that they could sleep, where a “no” response indicates that students knowingly avoided preventing a sexual assault. Furthermore, “yes/no” response options do not allow students to provide a rationale for their lack of intervention, which could point to important barriers outside of the decisional model of bystander behavior, such as post-assault situations that occurred in the past when bystanders may not be able to intervene directly due to time or physical space limitations. Finally, behavioral measures with “yes/no” response options fail to consider that someone else could have intervened and/or the potential victim may have diffused the situation, rendering the bystander unable to intervene. Thus, a “no” response may inaccurately capture an unengaged bystander. The allowance of text entry responses may better capture the true nature of students’ engagement as prosocial bystanders, as well as barriers otherwise not considered. The findings suggest that while we can attempt to measure barriers to bystander behavior based on the Burn (2009) decisional model for intervention, students still may face barriers different or unique to their personal experience of witnessing risk situations that a more nuanced approach to measuring bystander intervention behavior will better capture compared with traditional “yes/no” measures.
Alternatively, bystander measures could state the prosocial behavior to inquire whether students perform it rather than measures that illustrate a situation which students may/may not conceptualize in the same manner. For example, researchers may ask students whether they “check in with your friend who looks intoxicated and is being taken to a room by a guy” or “confront your friend who says he plans to get a girl drunk to have sex” (Banyard, Moynihan, Cares, & Warner, 2014; Hoxmeier, Flay, & Acock, 2016). In these cases, “checking in” provides the prosocial response that students could perform to prevent a sexual assault, and a “no” response indicates a lack of that specific prosocial response.
Limitations
The findings should be considered in the context of several limitations. The survey from which we derived the qualitative data was designed for quantitative research, and we recognize the potential for missing nuances in students’ written responses. There is potential bias introduced in this design, as some students may be less likely to type in their intervention barriers for reasons unknown to the authors. Similarly, the lack of purposive sampling is a limitation (Coyne, 1997), we believe, offset by the large sample size of this study. Kitto and colleagues assert that the objective of qualitative research is “to make conceptual generalisations from the local context of a qualitative study to other settings” (Kitto, Chesters, & Grbich, 2008, p. 1). Thus, while we do not assume that each of the barrier themes found salient for students in this study captures intervention barriers in their entirety, the findings show that students face barriers aligned with the bystander decisional model for bystander intervention (Bennett et al., 2014; Burn, 2009). Given the prevalence of alcohol on college campuses, for example, it is plausible that students at other campuses perceive situations involving alcohol as commonplace or otherwise a facet of campus culture, thus affecting their detection of risk in such situations. As well, we recognize that bystanders themselves may fail to notice risk situations due to their own intoxication, and capturing students’ intervention opportunities with self-report is a limitation difficult to circumvent given the nature of researching sexual assault risk situations.
The themes that emerged from this study do not appear so specific to this campus culture as to challenge the consideration of accommodating such barriers in programming or future research in this area, emphasized by Kitto et al. (2015) as fundamental for qualitative research. The intent of this work was not to suggest that all students experience the same barriers. Rather, the intent was to examine bystander behavior beyond a dichotomous measure of intervening and not intervening, and, in doing so, the discovery of students’ report of lack of intervention outside of those barriers or rationale provided suggested that limited measures of bystander behavior may not capture students’ experience. We only examined six situations here and recognize the potential for students to experience barriers for those situations unexamined. This is the first study, however, where participants in a quantitative survey of bystander behavior were able to provide qualitative information about the barriers they faced as bystanders to sexual assault situations, the results of which may have a meaningful impact on future bystander training and research of a similar nature.
Conclusion
As more campuses attempt to promote bystander engagement among their students, it is increasingly important to understand the barriers that exist for intervening in risk situations. Although many students reported barriers in line with those uncovered by previous research, allowing students to describe additional barriers provided meaningful insight into the range of experiences students may face as bystanders. Bystander training, as well as health communication campaigns that reinforce prosocial bystander behavior, can integrate discussions of risk perception, and namely the role alcohol plays in campus sexual assault, within the greater discourse of campus sexual violence. Research in this area should continue to refine measures to assess bystander intervention behavior and may consider qualitative response options to better capture the nature of students’ bystander behavior, as well as better understand students’ intervention barriers.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.