What we ask twice
Take one of our five statements: “I’m uncomfortable when a man here makes sexual comments about a woman.” First answer yes, no or not sure. Then estimate the percentage of men in your own group who will answer yes.
Suppose 16 of 20 say yes and the group’s average guess is 45%. The two bars show 80% agreement and a 45% guess: a 35-percentage-point gap. These are example figures. The session uses the group’s own figures only when at least 15 complete responses come in.
Read the three experiments by De Souza and Schmader ↗
What people see their leaders tolerate
In its review of academic science, engineering and medicine, the National Academies identified the climate of a department or institution as the strongest predictor of sexual harassment. The EEOC’s task force also found limited evidence that workplace instruction by itself prevents harassment. That’s why we ask where a report goes here and what a leader does when someone speaks up.
National Academies report, 2018 · EEOC task force report, 2016
Start with the people others follow
In a five-year randomized trial at 26 Kentucky high schools, Green Dot combined schoolwide presentations with intensive practice for influential students. In year four, 4.97% of students in intervention schools reported perpetrating sexual violence, compared with 6.28% in comparison schools. Those percentages describe respondents across the schools. Peer leaders were one part of that broader effort. Locker Room borrows that design choice. Its own impact still needs evaluation.
Coker and colleagues, 2017
What your follow-up measures
Each participant picks an action he can take within thirty days. At follow-up he says whether he did it and what happened. The report identifies how many answered, and shares outcomes only if enough people responded. Participants may have had different opportunities to act during those thirty days.
Those answers describe actions people report taking. To establish whether a program reduced harassment would require a different evaluation, with a comparison group and a measure of harassment itself.