Crowdsourcing effective educational interventions

Creating effective educational interventions that correct people’s misconceptions is difficult. This has led many researchers to conclude that people do not properly attend tone information in a way that they should. However, even if ascientifically-grounded intervention fails, it is still possible that other interventions would be effective. Yet, it is not practically feasible to systematically explore and test the entire hypothesis space of possible interventions. Here, we examined whether researchers could use online arguments to develop effective educational interventions, in effect, narrowing the intervention hypothesis space. Across two experiments (N = 1,816), we found that arguments crowdsourced from Reddit’s Change My View were as effective or more effective at changing beliefs than interventions developed by academics and published in top-tier scientific journals. These results suggest that researchers can build on successful crowdsourced arguments to develop effective educational interventions likely to co