Plausibility of experimental interventions

Recent studies showed that intergroup interventions should be tailored for groups with specific ideological profile and social contexts, so that they satisfy the target groups’ needs. Additionally, some individuals might negatively react to interventions in terms of rejecting their content as untrue. In the experimental context, they can be excluded from the analyses, making the interventions’ impact on them unknown. We argue that individuals who are already prejudiced are more likely to assess interventions as implausible. Across five experiments (N = 1896) in post-conflict ex-Yugoslavia, we first explored the predictors of plausibility assessment of two differently framed types of the Gateway group intervention: outgroup- and ingroup-perspective ones; we then tested whether plausibility moderates the interventions’ effect on intergroup bias. Our results confirmed that those who perceive the outgroup as more threatening, and who are less prone to perspective taking, are also more likely to reject the intervention framed from the outgroup perspective. Furthermore, not only was this intervention successful for those who accepted its content, but it also backfired for those who rejected it. This calls for re-examining the practices of how interventions are experimentally tested, as well as the way their effects are generalized across populations.