Résumé

A robust finding in social psychology is that people judge negative events as less likely to happen to themselves than to the average person, a behavior interpreted as showing that people are “unrealistically optimistic” in their judgments of risk concerning future life events. However, I demonstrate how unbiased responses can result in data patterns commonly interpreted as indicative of optimism for purely statistical reasons. Specifically, I show how extant data from unrealistic optimism studies investigating people’s comparative risk judgments are plagued by the statistical consequences of sampling constraints and the response scales used, in combination with the comparative rarity of truly negative events. Two studies controlling for these statistical confounds fail to provide evidence in support of the conclusion that people’s probability estimates of future life events are unrealistically optimistic.

Référence

Adam Harris (University College London), « The unrealism in unrealistic optimism research: The statistical artefact hypothesis », IAST General Seminar, Toulouse : IAST, 17 novembre 2011, 15h30–17h00, salle MF323.