Eavesdropping on social life: The accuracy of stranger ratings of daily behavior from thin slices of natural conversations

Abstract

In two studies the authors examined the accuracy of stranger ratings of daily behavior based on thin slices of natural conversations. Methodologically, the studies extend past research by using a behavioral accuracy criterion, benchmarking zero-acquaintance accuracy against target and informant accuracy, and employing a representative design that sampled contexts from targets’ daily situations. Theoretically, the studies investigate how stereotypes influence the accuracy of first impressions depending on their sample-based validity. Across both studies, after listening to five conversational snippets (2.5 min total), the ratings of strangers were as accurate as the targets’ and informants’ ratings. Further, ratings for gender-stereotypic behaviors with a kernel of truth resulted in greater initial accuracy than ratings for gender-stereotypic behaviors with no kernel of truth.