RESUMO
Omission bias is people's tendency to evaluate harm done through omission as less morally wrong and less blameworthy than commission when there is harm. However, findings are inconsistent. We conducted a preregistered meta-analysis, with 21 samples (13 articles, 49 effects) on omission-commission asymmetries in judgments and decisions. We found an overall effect of g = 0.45, 95% confidence interval (CI) = [0.14, 0.77], with stronger effects for morality and blame than for decisions. Publication bias tests produced mixed results with some indication for publication bias, though effects persisted even after most publication bias adjustments. The small sample of studies included limited our ability to draw definite conclusions regarding moderators, with inconclusive findings when applying different models. After compensating for low power, we found indication for moderation by role responsibility, perspective (self vs. others), outcome type, and study design. We hope this meta-analysis will inspire research on this phenomenon and applications to real-life, especially given the raging pandemic. Materials, data, and code are available on https://osf.io/9fcqm/.