A Practical Significance Bias in Laypeople’s Evaluation of Scientific Findings
研究发现，当媒体报道未报告效应大小时，人们更可能认可昂贵的干预措施，即使实际效应很小，这被称为实际显著性偏差；而低计算能力者即使被告知效应大小，也更容易受小效应影响。
People often rely on scientific findings to help them make decisions—however, failing to report effect magnitudes might lead to a potential bias in assuming findings are practically significant. Across two online studies (Prolific; N = 800), we measured U.S. adults’ endorsements of expensive interventions described in media reports that led to effects that were small, large, or of unreported magnitude between groups. Participants who viewed interventions with unreported effect magnitudes were more likely to endorse interventions compared with those who viewed interventions with small effects and were just as likely to endorse interventions as those who viewed interventions with large effects, suggesting a practical significance bias. When effect magnitudes were reported, participants on average adjusted their evaluations accordingly. However, some individuals, such as those with low numeracy skills, were more likely than others to act on small effects, even when explicitly prompted to first consider the meaningfulness of the effect.