Harm Hypervigilance in Public Reactions to Scientific Evidence

Area of Science:

• Behavioral Science
• Risk Assessment
• Scientific Publishing Ethics

Background:

• Journals are increasingly considering potential downstream harms of scientific findings.
• The "harm-hypervigilance hypothesis" suggests a tendency to overestimate negative consequences.

Purpose of the Study:

• To test the harm-hypervigilance hypothesis in the context of controversial behavioral science.
• To examine associations between harm overestimations, censorship support, and political ideology.

Main Methods:

• Two preregistered studies with representative U.S. adult samples (N = 1,865).
• Participants assessed risks associated with six sets of scientific findings, with some incentivized for accuracy.

Main Results:

• Consistent overestimation of others' harmful reactions (medium to large effect sizes).
• Consistent underestimation of helpful reactions, even with accuracy incentives.
• Harm overestimations correlated with support for censoring science.
• Offense at findings linked to comprehension difficulties.
• Moderate association between conservative ideology and harm overestimations.

Conclusions:

• Public perception of behavioral science risks is skewed towards harm.
• Overestimating harm may influence decisions about publishing controversial science.
• Further research is needed on social pressures influencing science and the costs/benefits of publication.