Assessing Students' Intention to Intervene in a Bystander Situation

The core premise of the bystander approach to campus violence prevention emphasizes the prosocial role that community members can play in thwarting sexual violence and focuses on encouraging and training students to act in ways that interrupt potentially violent situations (Dovidio, Piliavin, Schroeder, & Penner, 2006). How students perceive themselves as active bystanders—those most likely to disrupt situations with violence potential—is critical for creating effective campus policies designed to help stop violence in its many forms.

To date, however, little research has been done to measure college students' intention to intervene in potentially violent situations, nuanced to the collegiate contexts students actually face (McMahon, 2010). Although some scales have been developed to gauge past bystander behavior, attitudes, and efficacy (see McMahon, Banyard, & McMahon, 2015), they do not measure students' likelihood to intervene in specific situations and the social contexts that motivate such behavior. The purpose of this article is to introduce a scenario-based measure designed to assess students' intention to intervene in a bystander situation.

THEORETICAL FRAMEWORK

To empirically consider college students' intentions to intervene in a bystander situation, we grounded our inquiry in both the social ecological model developed by Dahlberg and Krug (2002) and Latané and Darley's (1970) decision model of helping. These models guided our decisions to use scenario-based questions to generate responses related to students' perceptions of themselves as bystanders and their willingness to intervene across different contexts.

The social ecological model "explores the relationship between individual and contextual factors and considers violence as the product of multiple levels of influence on behavior" (Dahlberg & Krug, 2002, p. 12) by specifying that intrapersonal, interpersonal, organizational, community, and policy factors all contribute to health behaviors. Our measure draws from the social ecological model in its design explicitly for use with undergraduate student populations, with consideration for their culture and experiences; for example—and as a means of nuancing the measure to the collegiate contexts students may actually face—the scenarios considered level of aggression related to potentially high-risk drinking behaviors commonly found on college campuses.

Latané and Darley's (1970) decision model of helping provides a framework for explaining the complex ways in which bystanders decide whether or how to intervene; for example, the relationship of bystanders to those involved in the situation may influence intervention behaviors, as some scholars believe that responsibility to intervene increases if the bystander has a relationship with the victim (Brody & Vangelisti, 2016; Levine, Cassidy, Brazier, & Reicher, 2002), while others have found no relationship between knowing the victim and bystander behavior (Banyard, 2008). The decision model of helping also informed the contexts of two scenarios as relationships to the victim varied across scenarios. The scale in this study provides students with two high-risk, yet differently ambiguous, scenarios and asks them the likelihood they would engage in specific intervention behaviors based on various relationships presented.

LITERATURE REVIEW: MEASUREMENT OF STUDENT BYSTANDER INTERVENTION

As bystander educational programs become ubiquitous on college and university campuses, scholars have developed several quantitative measures to assess the efficacy of these initiatives. Some instruments, such as the Bystander Willingness to Help Scale (BWTHS) and the Bystander Attitudes Scale (BAS; Banyard, Plante, & Moynihan, 2005) focus on students' attitudes related to sexual violence and willingness to engage in bystander intervention behaviors and have been widely used to measure the efficacy of university bystander intervention education programs.

Despite the strengths of these scales—their reliability, contribution to literature, and breadth of information gathered—some areas for improvement exist. For example, the scenarios asked in the BWTHS and BAS (which are the same for both scales) provide potential bystander behaviors and do not leave room for respondents to communicate other possible intervention behaviors. Although the vignettes section developed by Banyard et al. (2005) addresses this weakness somewhat, respondents are only asked to rate the order in which they would intervene, not the likelihood they would take an action. These scenario items also request that respondents evaluate the situation from only one relational perspective or do not reveal the relationship...