Bystander intervention has emerged as a popular and widespread strategy for addressing campus sexual violence prevention and frames sexual violence as a community issue for which everyone has responsibility (Banyard 2015). Bystanders have the opportunity to help prevent violence by disrupting risky situations, intervening during a situation in which violence is occurring or engaging in tertiary prevention by providing assistance to those that disclose (McMahon and Banyard 2012). Research demonstrates multiple positive effects of bystander intervention programs on a number of outcomes including student attitudes, efficacy, and behaviors (Katz and Moore 2013; Jouriles et al. 2018) as well as victimization and perpetration (Coker et al. 2017).

Despite the promise of bystander intervention programs, we know little about the bystander intervention experiences of minoritized students, such as students of color, LGBTQ-spectrum students, and the intersections thereof (Harris 2017; Linder 2018). Further, this gap has led to neglecting the potential of bystander intervention to address broader forms of violence, harm, and oppression unique to the experiences of students minoritized by race, sexuality, and gender. Sexual violence prevention efforts more broadly have been critiqued for not addressing the role of identity, power, and historical oppression (Harris and Linder 2017). A shift to better “center” the experiences of those excluded from positions of power is especially pressing, given findings that students minoritized by race/ethnicity, sexual orientation, and/or gender identity are significantly more likely to be subject to sexual violence than their dominant identity (white, cisgender, heterosexual) counterparts while in college (Porter and McQuiller Williams 2011). Further, recent research indicates that positive prevention program effects may not apply to students minoritized by sexuality in high schools (Coker et al. 2020), though findings with college students are more promising (Ollen et al. 2017).

In addition, research has yet to specifically explore what factors facilitate or inhibit prosocial bystander intervention from the perspective of minoritized students. It is important to identify these factors to tailor prevention efforts and ensure that current approaches to bystander intervention encompass the experiences of students from diverse backgrounds. Given the stigma and discrimination faced by many minoritized students, it is particularly essential to establish the level of institutional responsibility required to help create environments that foster bystander intervention. As such, the purpose of this paper is to present the findings from an exploratory concept mapping study regarding bystander intervention that focused exclusively on the perspectives of racialized and/or LGBTQ-spectrum students.

Factors Influencing Bystander Intervention and the Role of Identity

Research has focused largely on identifying individual and peer level factors that contribute to prosocial intervention (Banyard 2015). Individual factors include having certain attitudes (e.g., sense of responsibility), skills, and efficacy, and peer level factors include social norms that support helping behavior (Rothman et al. 2019). Experiencing multiple forms of oppression and discrimination may influence students’ perspectives on the unique barriers to intervention that they face. Additionally, there remains a lack of research on the role of institutions in fostering environments in which individuals feel supported to act as helpful bystanders (McMahon 2015), which may be especially salient for minoritized students.

Within research on bystander intervention and campus sexual violence, there is a dearth of attention to student identities (Linder 2018). Literature related to race finds mixed effects; some studies document greater bystander action and readiness to help Black and Latinx students compared to others (Brown et al. 2014; Christensen and Harris 2019), while other studies found that Black and Hispanic students reported more missed opportunities (Hoxmeier et al. 2017) and less likelihood to intervene (Brewster and Tucker 2016). Others found no main effects of race and ethnicity on bystander outcomes but did find interaction effects between gender and race (Burns et al. 2019; Diamond-Welch et al. 2016). Vignette studies have demonstrated less willingness to help minoritized victims (e.g., Katz et al. 2017) and less empathy for victims of color (Franklin and Garza 2018). These mixed findings call for further research.

Intersectionality, Power Consciousness, and Campus Prevention

Increasingly, researchers in public health and prevention science recognize that an intersectional approach is needed, one that acknowledges an individual’s various positionalities (e.g., racial, ethnic, sexual orientation, gender identity, economic, religious), how they work together and impact experiences of sexual violence and its prevention (Crenshaw 1991) and campus sexual violence specifically (Christensen and Harris 2019; Harris and Linder 2017). Linder (2018) draws on intersectionality while foregrounding the role systems of oppression (e.g., racism, sexism, homophobia, transphobia, classism) play in over-exposing minoritized groups to multiple forms of violence. Linder proceeds to develop a power-conscious framework to address the deeply embedded roots of violence (including sexual violence), accounting for the centrality of power and privilege and calling for solidarity in undermining oppressions that allow violence to perpetuate.

A bystander approach informed by these frameworks urges a centering of minoritized identities and how they are impacted by systems of oppression that may lead to different experiences as bystanders and different needs to support acting to prevent violence. This includes asking students about their perceived roles as bystanders, how they view their ability to engage in prosocial action and barriers to action, and what types of support they need to successfully intervene. In addition, it is important to know to what extent students feel they have influence over the factors that facilitate intervention, versus factors that are the responsibility of the institution. As such, the guiding questions for this exploratory study are (1) “What supports do students of color and/or LGBTQ-identified students say they need to overcome their unique set of barriers, to positively intervene in situations related to sexual violence and other forms of harm or oppression?” and (2) “How are these types of support prioritized by students of color and/or LGBTQ-identified students, and how much influence do students believe they have over implementing these supports?”

Methods

This study employed the canonical concept mapping process outlined by Trochim (1989) to learn more from minoritized students regarding their perspectives on bystander intervention related to sexual violence, dating violence, and other forms of harm. Concept mapping is a mixed-methods approach that uses a structured process to gather input from participants through a series of phases that result in a visual depiction of their conceptualization of a particular idea (Trochim 1989). The process includes gathering qualitative statements from participants about a particular topic and then having participants group the statements by perceived similarity. Participants then rank the statements by importance and other factors, which leads to the generation of visual maps demonstrating participant perceptions and priorities. Given its stakeholder-driven process, concept mapping provided an appropriate fit for the study’s intent of foregrounding minoritized student identities, experiences, and priorities within knowledge surrounding bystander intervention and its resulting initiatives (Kane and Trochim 2007).

Specifically, the current study focused on students who were identified as individuals of color and/or on the LGBTQ-spectrum, recruited from multiple campuses that are part of a large, public university in the mid-Atlantic. Campus 1 has a significantly larger student population (more than 40,000 students) than campuses 2 (more than 13,000 students) and 3 (more than 7,000 students). In line with the campus sizes, campus 1 comprises 55% of overall participation, with campus 2 comprising 23%, and campus 3 comprising 22%. Based on concept mapping methodology, three phases were conducted: brainstorming, sorting and rating, and finally mapping and interpretation of the results (Trochim 1989). A total of 101 students participated in the entirety of the study. All materials and procedures for the project received approval by the university’s Institutional Review Board (IRB).

Phase 1: Brainstorming

Brainstorming Sample and Recruitment

The study was open to any currently enrolled undergraduate or graduate student over the age of 18 who was identified as a person of color, identified on the LGBTQ-spectrum, or held overlapping identities. To recruit participants from these hard-to-reach populations, researchers implemented “purposive social network sampling” (Pfeffer 2018), asking that key centers on each campus who work with students of color, LGBTQ-spectrum students, or the intersections thereof partner with us in disseminating the opportunity to participate in the study. Prospective participants completed online eligibility forms that ensured they met inclusion criteria. Prior to the initial round of brainstorming sessions, researchers made all materials available to the violence prevention office on each campus (three total) for feedback prior to submitting for IRB approval. This helped establish face validity of materials. Partners were asked to ensure that materials were complete, inclusive, and clear. The team received minor feedback about the language used to describe certain services and offices on campus, as well as suggestions for recruitment strategies and accessible locations. Partners were advised to adapt the screening form to allow participants to select overlapping racial/sexual/gender identities rather than having to choose between identity groups.

Demographic questions were formulated to account for intersectional identities, allowing participants to select multiple identifiers, to self-identify, or to not disclose (see Table 1). Optional identifiers included orientations and genders that exceed binaries, such as “asexual” or “genderqueer,” as suggested by Consortium of Higher Education LGBT Resource Professionals (n.d.). Terms such as “multiracial” accounted for overlapping racial or ethnic identities (University of Arizona n.d.). Participants often selected multiple gender identities, sexual orientations, races, and ethnicities. Thus in Table 1, the number of responses often exceeds the total number of participants in a given activity. Following the conclusion of the brainstorming activity, with IRB approval, we included two additional racial/ethnic categories (South Asian or Southeast Asian, and Caribbean or Afro-Caribbean) to online sorting and rating to further accommodate the range of participant identities.

Table 1 Participant demographic characteristics for brainstorming, sorting, and rating
Full size table

Brainstorming Procedures

Hour-long brainstorming sessions were conducted on each of the three campuses, with one geared toward persons of color, and one geared toward LGBTQ-spectrum individuals, for a total of six brainstorming sessions. Participants with intersectional identities (e.g., queer of color) were invited to participate in either or both sessions, with the assurance of being compensated for two sessions if they chose this route. This option was offered in recognition of the fact that identities always contain their overlapping positionalities. Two eligible students opted to attend both of their campus sessions. A total of 39 students participated in the brainstorming phase. Participants received a $30 gift card in compensation for their time and input.

Brainstorming Measures

Traditionally in concept mapping, brainstorming revolves around one “focus prompt,” or single complete-the-sentence statement that gets at the intent of the project (Trochim and McLinden 2017). All brainstorming participants were asked to respond to the following focus prompt: “One thing that would support students in becoming active bystanders who prevent violence, harm, and oppression on this campus is…”. Before asking participants to respond, researchers handed out a sheet that defined the terms violence, harm, sexual violence, dating violence, oppression, bystander, and bystander intervention, based on bystander intervention curricula developed by the Utah Department of Health, Violence and Injury Prevention Program (2018). Researchers reviewed the definitions and then invited participants to respond to the focus prompt. Depending on the technology available, participants then responded to the focus prompt by writing their statements on index cards that were read aloud and displayed or had their statements added to a visible list within online software.

Brainstorming Analysis

After the brainstorming sessions, researchers compared transcriptions of participant responses to session audio to ensure collected statements were accurate. Participants collectively brainstormed a total of 165 ideas that then went through a meticulous idea synthesis process by applying the “keyword-codeword framework” outlined by Kane and Rosas (2018). Two researchers identified keywords within statements, assigned statements broader codes, then combined similar statements, separated multi-focus statements, ensured statements were understandable and ratable, and included statements that responded to the focus prompt (Kane and Rosas 2018). While some ideas unique to each campus were presented (such as raising awareness of a campus-specific antibias group), most of the thematic content was found to overlap across the three campuses, suggesting that “saturation of the topic” was achieved (Kane and Rosas 2018). This resulted in a final list of 102 statements across the campuses, on par with the recommendation to produce around 100 statements (Kane and Trochim 2007). These synthesized statements were imported to concept mapping software, CS Global MAX, where they were randomized prior to phase 2.

Phase 2: Sorting and Rating

Sorting and rating activities comprise the statement structuring phase that helps generate overarching themes. Sorting serves to collect individual and thus collective perceptions of the interrelationship between generated statements (Kane and Trochim 2007). This combined data becomes the basis for future point and cluster maps. Rating allows researchers to determine how participants prioritize ideas and, in the case of this study, how they understand the degree of influence they have over implementing ideas. All sorting and rating took place online via CS Global MAX software.

Sorting and Rating Sample and Recruitment

Eligibility and recruitment procedures were the same as for the brainstorming phase. Students who participated in the brainstorming phase were invited to participate, and additional recruitment was conducted by working with campus partners to send emails to listerves. All prospective participants filled out the screening form to determine eligibility. A total of 91 students participated in the sorting and rating phase, with a total of 78 participants providing usable data. See Table 1 for demographics.

Sorting and Rating Procedures

Participants were emailed a link to the sorting activity, which they were instructed to complete first, and rating activity, which they could thereafter choose to complete. Participants had a 3-week period of time to complete sorting and rating.

Sorting and Rating Measures

During the sorting activity, participants were given instructions based on Kane and Rosas (2018). First they were asked to categorize the 102 statements derived from the brainstorming sessions into between 5 and 20 “piles” or groups, based on how similar in meaning they perceived them to be. They were instructed to begin by reviewing all statements and then to click and drag statements into piles in a way that made sense to them while also attributing a descriptive name to each pile (e.g., “Outreach”). Sorting participants were instructed to categorize ideas based on perceived similarity, not based on priority; to attribute descriptive labels, not labels such as “miscellaneous”; and to place all statements in a pile, even if a statement was in a pile of its own.

During the rating activity, students were asked to rate the 102 statements on a 4-point Likert scale for both 1 “how important you feel each idea is for reducing violence, harm, and oppression on your campus” and 2 “the level of influence you feel students have/could have over the implementation of these ideas on your campus”, ranging from 1 (relatively unimportant/little to no influence) to 4 (extremely important/strong influence). Whereas the “importance” measure is commonly used in concept mapping studies (Kane and Trochim 2007), the “influence” measure was developed uniquely for this project to determine whether students felt they had the ability to impact an idea’s implementation. Rating participants were instructed to begin by reviewing all statements and then attribute a value of 1–4 to each statement in comparison to all other statements, to use the full scale range as appropriate, and to avoid rating all statements with the same number.

Sorting and Rating Analysis

Upon closing the sorting and rating activities, researchers went through a meticulous quality review of the data guided by Kane and Rosas (2018). First, activities were removed for participants who had initiated the process but did not provide any data. Next, consultation between researchers and CS Global MAX experts led to the exclusion of activities that did not meet minimum data standards. This included activities that sorted less than 75% of statements, those sorted with large unnamed categories, those activities that attributed the same rating throughout (e.g., rated every statement a “4”), and incomplete ratings (Littlefield, J., personal communication, May 2, 2019–May 6, 2019). This resulted in the exclusion of seven sorts and two ratings. The final analytic sample included 75 total sorts and 65 total rates from a total of 78 participants. All participants who provided data, even if excluded from use, received compensation of $30 for sorting and an additional $30 for rating.

Phase 3: Mapping and Interpretation

Following the generation of a specified series of maps, researchers conducted preliminary interpretation and then presented findings to stakeholders for feedback. In converting qualitative data to quantitative data, a project similarity matrix is internally generated by CS Global MAX based on the combination of all participant sorting activities. The software then applies multidimensional scaling to the similarity matrix, which generates the point map, a two-dimensional representation of the spatial relationships between each statement plotted as a point/coordinate. At this phase, the software calculates stress value, representing the degree to which the distribution of point map coordinates accurately reflects similarity matrix data. The stress value of this study was 0.29, with typical stress ranging from 0.20 to 0.36 and a lower figure representing better fit (Kane and Rosas 2018). The software then applies hierarchical cluster analysis deploying Ward’s algorithm to consider every possible pairing of clusters at each step while prioritizing the integrity of data (Everitt 1980). Researchers explored a number of cluster solutions within CS Global MAX via the cluster replay map feature, which compares on “loop” the possible cluster solutions, noting as clusters become successively merged. Researchers found a six-cluster solution to be most coherent with the original spatial themes that emerged.

With visual maps generated and preliminary outcomes produced, researchers organized a stakeholder interpretation session with the leadership of campus community centers who had been close collaborators, as well as researchers adjacent to the study. This provided a reliability check and an opportunity to inquire if preliminary findings resonated with what experts were seeing on-the-ground, to offer exploration of data that may have gone overlooked, and to discuss next steps such as actionable deliverables. A total of nine stakeholders and five researchers attended the session. Researchers provided a brief presentation to review the study purpose, methods, and preliminary results and then asked a series of questions to seek input, including whether the findings reflected what stakeholders observed in their interactions with students and if there were additional areas of data they wanted explored. Participants confirmed that the findings mirrored the types of conversations and observations they had with students, and they agreed that the clusters made conceptual sense. Minor suggestions were made, such as needing to refine demographic categories for those of Asian descent.

Results

Identification of Supports for Students to be Helpful Bystanders

To respond to the research question regarding supports needed for students to be effective bystanders, statements within the point map were systematically grouped into regional themes to form a point cluster map. The final cluster solution was selected by finding the arrangement that most maintained the ideological coherence of each cluster and minimized the separation of like ideas while disallowing the fusion of clusters containing disparate ideas (Everitt 1980). This led researchers to agree upon a six-cluster solution as seen in Fig. 1, with labels selected for each cluster based on familiarity with regional thematic content and the category labels proposed by participants during the sorting activity (Kane and Rosas 2018).

Fig. 1
figure 1

Point cluster map as conceptually grouped by participants. The point cluster map depicts the original point map or brainstormed statements plotted as points in two-dimensional space, with the final 6-cluster solution overlaid (Kane and Trochim 2007). Statements within each cluster (or “shape”) generally fall under the theme of the descriptive label attributed. A small number of exceptions may occur with “bridging” statements, wherein a statement is conceptually related to multiple regions of a map and its position is thus determined by being pulled toward multiple statements simultaneously (Kane and Rosas 2018)

Full size image

The point cluster map depicts the six primary conceptual themes emerging from the study as brainstormed and sorted by participants: (1) creating and improving student services, (2) intensifying bystander intervention initiatives, (3) programming rooted in identity and social justice, (4) institutional respect for identity, (5) institutional assurance of safety and justice, and (6) shifting campus norms via community collaboration (see Table 2 to review a sample of verbatim statements representative of each cluster; see online supplemental material for a comprehensive list of clusters and associated statements).

Table 2 Sample brainstorming statements by cluster
Full size table

Participants likewise rated each statement based on perceived importance and level of influence, thereby generating average ratings for each statement and for each cluster overall. All statements were rated above the midpoint value of 2.0 in terms of importance and influence. On the 4-point Likert scale, 21% of items were rated under 3.0 for importance (i.e., the majority were rated highly important), whereas 63% of items were rated under 3.0 for level of student influence (i.e., the majority were rated as having moderate influence).

Importance of and Influence over Identified Supports

Visual maps were generated to respond to the research questions regarding what students of color and/or LGBTQ-spectrum students prioritize by importance and feel they do/do not have influence over. An effective way to compare cluster ratings by importance and influence is by way of a “relative pattern match” (Fig. 2), which compares different rating values in the form of a ladder graph, in this study displaying “importance” and “influence” values side-by-side. When a pattern match is generated as “relative,” it applies the actual highest and lowest average cluster rating values to the left and right vertical sides of the pattern match, “anchoring” both data sets to these values and assigning them equal position, thus relativizing how data is displayed (Kane and Rosas 2018). As demonstrated in Fig. 2, there is a mid-range negative correlation between the ratings of importance and influence (Pearson’s r = − 0.47) implying that the clusters participants considered most important were conversely the ones they felt they had the least influence over (Beins and McCarthy 2018).

Fig. 2
figure 2

Relative pattern match depicting participant rating comparison. The pattern match takes on the form of a “ladder graph” as it compares average cluster ratings side-by-side. The left vertical side displays average cluster ratings by perceived importance (high of 3.34; low of 3.04), while the right vertical side displays average cluster ratings by perceived degree of influence (high of 3.02; low of 2.82). The data in this pattern match has been “relativized,” meaning that data sets are anchored to and displayed relative to the actual highest and lowest average cluster rating values

Full size image

There were significant t test differences between participants’ perceptions of average importance and influence ratings for clusters one through five (1, creating and improving student services; 2, intensifying bystander intervention initiatives; 3, programming rooted in identity and social justice; 4, institutional respect for identity; and 5, institutional assurance of safety and justice) although no significant difference for cluster six, shifting campus norms via community collaboration which is also the cluster over which participants felt that they and their peers have the greatest degree of influence (see Table 3).

Table 3 Comparisons of importance and influence ratings for all clusters
Full size table

Discussion

This study focused exclusively on the perspectives of students of color and/or LGBTQ-spectrum students on bystander intervention as a form of prevention for violence, oppression, and harm. Overall, findings indicate a need for bystander intervention efforts to widen their focus by employing a social justice lens. Current bystander intervention programs tend to focus exclusively on addressing sexual and dating violence without integrating other forms of oppression. Yet, in addition to dating and sexual violence, study participants identified various forms of racism, homophobia, transphobia, and microaggressions as needing bystander intervention. Participants also highlighted the ways that their identities are intertwined with their ability to be active bystanders on college campuses. For example, some students reported that intervening may present additional risks or seeking formal assistance may not be perceived as a helpful option. Indeed, the results suggest that intersectional approaches to prevention are needed that acknowledge and address their lived experiences of violence, harm, and oppression as a part of sexual violence prevention—not as separate initiatives, which is typically the practice.

Within the study, participants identified the need for response services as the most important in terms of facilitating bystander intervention. This study was conducted at a resource-rich campus with numerous victim services offices and programs, so this warrants further unpacking. It is unclear whether students are not aware of services, do not feel safe or comfortable seeking assistance, or may not believe the services are relevant to their own experiences. All of these are critical issues to explore to better understand how to make sure all students perceive campus resources as welcoming, power conscious, and intersectional in their approaches. However, this cannot be done with an “ahistorical” approach which neglects to recognize that students who were identified as belonging to a marginalized group may experience a warranted lack of trust in more formal systems or reporting (Hong 2017) and unique barriers based on fear of being disbelieved, mistreated, or discriminated against based on identity (Ollen et al. 2017). While campuses may work to improve their resources and address issues unique to students minoritized by race, sexuality, or gender identity, there is also the need to build authentic relationships and trust.

In calling for increased services, participants in this study acknowledged the need for services tailored toward those committing acts of violence, in line with a social justice paradigm focused on the actions of those who perpetrate violence, harm, and oppression that simultaneously examines the systems that allow these acts to occur (Hong 2017). Applied to college campuses, this means better understanding perpetration on campus, establishing systems to hold individuals accountable for perpetrating harm against others, and offering services to potentially change their behavior. This includes gathering data not just about victimization but also perpetration experiences and building student trust in fair, just adjudication processes. One model gaining increased attention is restorative justice, a victim-centered process in which the person who perpetrated the act accepts responsibility, with the goal of repairing the harm caused to the victim/survivor and those around them (Koss et al. 2014).

As demonstrated by the results, students called for “programming rooted in identity and social justice.” Models are needed for bystander intervention programs that are cross-cutting, addressing multiple forms of oppression, harm, and violence (Hamby and Grych 2012). Work is needed to determine if a single bystander program can effectively address multiple issues. This means that campus efforts to address sexual violence and other forms of harm should not be siloed, with disconnected offices and programs facilitating the work in isolation. Rather, a social justice paradigm calls for multiple entities across campus to work together to build capacity and address the roles of identity, power, and privilege across programs, policies, and procedures (Hong 2017).

Students in this study also called for “intensifying bystander initiatives,” which includes expanding programs to faculty, staff, and administrators. Currently, bystander intervention education programs are typically delivered exclusively to students, yet the results from the study indicate a need for faculty and staff to be equipped with skills to effectively intervene to address microaggressions in the classroom as well as other forms of harm and oppression they may witness. This is an underdeveloped area within the field of bystander intervention, and models are needed for effective program design, implementation, and evaluation.

Study participants also identified the role of the institution as salient in the current study. They identified a number of institutional factors, such as “institutional assurance of safety and justice” and “institutional respect for identity” as critical to being effective bystanders. Typically, research on bystander intervention has focused on individual level correlates (Banyard 2011). Increasingly, there is recognition that the context, climate, and environment also matter in facilitating bystander intervention (McMahon 2015). Collectively, these findings support the critical role of institutional responsibility across the entire campus (in classrooms, among faculty and staff, and across the institution). This again speaks to the need for campuses to not only demonstrate commitment to addressing sexual violence but to be actively engaged in a power-conscious framework that recognizes the intersection among all types of harm. In addition to mitigating systemic issues such as racism, genderism, and heterosexism, institutions must articulate their visions for creating environments that are safe, inclusive, diverse, and welcoming.

Participants identified the need to address campus norms that support those who intervene as lowest in importance yet highest in influence. This suggests that while they do not necessarily view norms change as a priority, they do believe they have a role to play in this area. Prevention efforts can potentially acknowledge this by better explaining the value of influencing norms and the important role that students have to play. For example, those with dominant identities can demonstrate solidarity with their minoritized peers by working to develop critical consciousness regarding how those peers experience the world and how they may encounter systemic and interpersonal harms because of their identity (Linder 2018). Dickter et al. (2011) found that when nontargets confront racist or heterosexist language, they are perceived positively and are more effective at reducing the bias of others; whereas when targeted individuals do so, they tend to be perceived negatively and face greater social consequences. Such demonstrations of solidarity can lessen the burden minoritized individuals face in having to educate those with dominant identities. Institutional practices, policies, procedures, and culture must be scrutinized to understand how they not only disallow harm but actively refute it.

The mean values for both importance and influence ratings were above the scale midpoint. This suggests that overall, students believe that a multipronged support approach is needed, with many components having a high priority. In addition, students generally believe that they can help these efforts, which provides a solid foundation for further engagement by the institution. Despite the relatively high mean ratings for these categories, clusters one through five were found to have significant difference between ratings of importance and influence, implying students feel they have less influence over the most important factors impacting bystander intervention. On one hand, this emphasizes the need for institutional responsibility, as, for example, students should not have to ensure that appropriate services are in place. On the other hand, this finding indicates that students may feel a lack of power and agency when it comes to creating social change. This suggests a need to identify where students feel they can have influence and build greater capacity. For example, participants rated the item “Building community and solidarity amongst peers in classroom settings so they can call out students or professors as necessary” highest for level of influence, signifying that student attitudes, skills, and feelings of efficacy can be fortified to collectively intervene in these situations. In contrast, the items rated lowest for level of influence included those aimed at improving the cultural competency and intervention skills of faculty and staff, signifying that the institution, rather than students, ought to be responsible for them.

The finding for the cluster “programming rooted in identity and social justice” stood out as unique because it was rated highly for both importance and influence. This is a key area in which students from minoritized backgrounds can be effectively engaged. Statements within this cluster call for violence prevention programming “tailored” toward minoritized identities, facilitation by individuals from a variety of backgrounds, and training that reflects “real-life situations” to which students could contribute knowledge of minoritized realities and action in the form of peer education. Thus, engaging students in norms change, campus community building, and peer-focused bystander initiatives and implementing identity-informed, social-justice-oriented prevention programming are a few promising opportunities.

Limitations and Future Research

The results of this exploratory study should be contextualized within a number of limitations. Due to relatively small sample size, the results are not necessarily generalizable to the student body of each campus. The study took place at campuses associated with one large public institution. While the themes appeared to cut across campuses, there may be variability that was not captured. Replication at other institutions is needed, as well as considering institutional-level characteristics such as rural versus urban settings, 2- versus 4-year institutions, commuter versus residential, to list a few examples. A number of items could have been assigned to multiple clusters, so further work is needed with additional samples to assess if this cluster solution is the best configuration. While a number of definitions were provided to participants to ensure consistency, the focus prompt asking about “supports for bystander intervention” did not define “support,” which may be helpful for future work. In addition, researchers developed the “influence” measure; further work is needed to determine if it is reliable.

The current study focused only on whether students identified as persons of color and/or on the LGBTQ-spectrum. There are clearly many other variables that contribute to intersectional identities such as ethnic background, ability status, immigrant and international status, and more. Further research is needed to explore the experiences of these students as well. This study also lacked a true comparison group. Additional work could explore whether the clusters identified by minoritized students are similar or different from students with dominant identities, as well as whether their ratings of importance and influence differ in significant ways. Making meaningful comparisons between the minoritized groups was not possible because both still consisted of minoritized students; thus comparison groups with those who occupy dominant identity positions are needed. There are also numerous gaps that remain in our understanding of the experiences of students from minoritized backgrounds related to participation in sexual violence prevention more generally and bystander intervention specifically. For example, are there additional factors that inhibit or facilitate bystander intervention for minoritized students across the whole social ecology (at individual, group, community, and societal levels)? How might racism, genderism, heterosexism, and other forms of oppression present obstacles to intervening or allow those from dominant identity groups to feel less responsible for intervening on behalf of minoritized individuals? In addition, further work is needed to seek out examples of intersectional approaches to prevention and to evaluate their effectiveness. The current study provides a starting point to develop these important questions for researchers and practitioners to pursue.