Stereotype Formation but No Dissociation: Contradicting Statistics Reduce Explicit and Implicit Stereotypes Engendered by Disproportional Crime Reporting

Abstract

This study investigated the formation of criminal stereotypes about novel groups. Over 2 days, 316 participants from Prolific Academic read either equal or unequal proportions of crime reports about fictitious groups, Laapians and Niffians. On the second day, additional crime statistics indicating equal crime rates among both groups or control statistics were presented. Unequal crime reporting resulted in explicit (d = .44, p < .001) and implicit (d = .23, p = .001) stereotypes assessed with an Implicit Association Test, while equal reporting did not. On the second day, only explicit stereotypes increased in the unequal crime reporting condition (d = .25, p = .025). Contrary to predictions of the associative-propositional evaluation model, but congruent with De Houwer’s single-process propositional model, crime statistics contradicting the learned regularities reduced both explicit (d = −.43, p = .010) and implicit stereotypes (d = −.20, p = .003).