Unconvenient scientific evidence: Harm over-estimators were more supportive of censoring scientific research; & those more offended by scientific findings reported greater difficulty understanding them (“motivated confusion”)

Harm Hypervigilance in Public Reactions to Scientific Evidence. Cory Clark, Maja Graso, Ilana Redstone, Philip E. Tetlock. March 2022. DOI:10.13140/RG.2.2.35921.20329

Abstract: Two preregistered studies (n = 1,423; one with a U.S. nationally representative sample) tested the harm-hypervigilance hypothesis in citizens’ risk assessments of controversial behavioral science findings. As expected, people consistently overestimated all harmful reactions to scientific findings with a medium-to-large average effect size (and underestimated all helpful ones). Additional analyses found (1) harm over-estimators were more supportive of censoring scientific research; (2) those more offended by scientific findings reported greater difficulty understanding them (“motivated confusion”); (3) social network ideological heterogeneity predicted more accurate (lower) estimates of harmful reactions (especially among ideologically extreme participants) and social network ideological homogeneity predicted more accurate (higher) estimates of helpful reactions; (4) mixed evidence on whether ideological groups overestimated harms that challenged their moral concerns. These findings raise the question: When does harm hypervigilance become net harmful by impeding scientific discovery and delaying evidence-based solutions to societal problems?