Case Study 32.2: Bystander Intervention — What the Evidence Shows

The Promise of Bystander Models

Traditional sexual violence and harassment prevention programs focused on potential victims (self-defense, risk reduction) or potential perpetrators (attitude change toward women). Bystander intervention reframes the problem: instead of targeting the two people in a harassment interaction, it targets the much larger community of witnesses — people who observe concerning behavior and have the potential to disrupt it before harm occurs.

The appeal of the bystander model is both practical and ethical. Practically, there are more bystanders than perpetrators — and training a large number of people to make small interventions could have larger aggregate effects than trying to change the behavior of a relatively small number of people who are already inclined toward harassment. Ethically, the bystander model distributes responsibility for preventing harm more broadly, challenging the "it's not my problem" norm that enables harassment.

Key Programs and Their Evidence

Green Dot is among the most rigorously evaluated bystander programs. Developed at the University of Kentucky, Green Dot uses a social norms approach combined with skills training. A randomized controlled trial in Kentucky high schools (Coker et al., 2011, 2014) found that students in Green Dot schools showed significantly lower rates of sexual violence perpetration, victimization, and bystander non-response compared to control schools — effects that held up over two years.

The Bringing in the Bystander program (University of New Hampshire) targets college populations and has been evaluated in randomized trials showing significant increases in bystander efficacy (the belief that one can effectively intervene) and reported intervention behavior, as well as some evidence of attitude change toward rape myth acceptance.

RealConsent is an online bystander program showing significant attitude and behavior change in male college students in a randomized evaluation.

What Makes Bystander Programs Work?

Research on moderators of program effectiveness identifies several factors:

Social norms correction: Many people overestimate how accepting their peer group is of harassment. Providing accurate information about actual norms ("most people here don't accept this") can shift the perceived social cost of intervening.

Skills training, not just attitude change: Programs that practice specific behaviors — what to say, how to physically insert yourself into a situation, how to check in with a target afterward — show stronger behavioral effects than programs that only address attitudes.

Campus culture integration: Programs embedded in broader campus culture change efforts (leadership training, athletic program integration, fraternity/sorority engagement) show stronger effects than standalone trainings.

Addressing inhibitors directly: Effective programs specifically address pluralistic ignorance ("I thought I was the only one uncomfortable"), diffusion of responsibility ("someone else will do it"), and bystander fear ("it's not safe for me to intervene").

The Barriers

Despite the evidence, bystander intervention has real limits. Research also documents the inhibitors:

Perceived relationship between victim and perpetrator: Bystanders are significantly less likely to intervene when they believe the people involved are in a relationship — even when the behavior would clearly constitute harassment between strangers.

Alcohol and context: High-alcohol environments (parties, bars) are precisely where harassment is most likely, and also where bystander intervention is most socially risky and cognitively difficult.

Gender of the bystander: Studies find that men and women intervene differently and at different rates across different situations. Men are somewhat more likely to use direct confrontation; women are more likely to check in with the target.

Safety: Especially for people from marginalized groups, intervening in a harassment situation involves real personal risk assessment. Bystander training that ignores this reality does a disservice to potential bystanders.

Digital Bystander Opportunities

The bystander model has migrated to online spaces. Dating apps including Hinge and Tinder have implemented reporting features that function as bystander opportunities — users who receive harassing messages can report them; users who see others being harassed in community features can flag the behavior.

Research on digital bystander behavior (Obermaier et al., 2016) finds that online intervention is more likely when: bystanders perceive the behavior as clearly wrong, the bystander has some connection to the victim, and the platform makes reporting easy. Frictionless reporting tools significantly increase rates of use — a finding with clear design implications.

Discussion Questions

1. The Green Dot program shows effects on sexual violence perpetration, not just attitudes. Why is this distinction important for evaluating prevention programs?

2. The research finds that bystanders are less likely to intervene when they believe the people involved are in a relationship. What assumptions underlie this reluctance? How accurate are those assumptions?

3. The "4 Ds" (Direct, Distract, Delegate, Document) give bystanders specific options rather than a single required response. Why is this range important? When would each D be most appropriate?

4. If you were designing a dating app's bystander intervention features, what would you build? Think about both in-app reporting for harassment of oneself, and community-level reporting for witnessed harassment.