| The accuracy of predictions regarding certain social and clinical outcomes from brief observations or "thin slices" of expressive behavior was examined in 3 studies and a meta-analysis. In the first study ratings of 30-second video clips of high school teachers on certain molar (e.g. warmth, enthusiasm) and micro (e.g. smiles, nods) dimensions of behavior predicted teacher effectiveness, defined by the school principal's ratings of teachers. In the second study ratings of 30-second content-filtered audio clips on molar dimensions of behavior distinguished between mothers of non-organic failure to thrive infants and mothers of normal infants. In the third study ratings of transcripts of 30-second excerpts of interviews with spouses of unipolar depressed patients predicted the relapse of the patients in 9 months following the interview. A meta-analysis on the accuracy of predictions of various social and clinical outcomes based on short observations of expressive behavior (under five minutes), revealed an overall effect size r =.39 (Z = 22.56, p {dollar}<{dollar} 1/10{dollar}sp{lcub}112{rcub}{dollar}) for 38 different results. Studies employing longer periods of observation did not yield greater predictive accuracy: predictions based on observations under half a minute in length did not differ significantly from predictions based on four or five minute observations. The type of behavioral channel (such as the face, speech, body, tone of voice) on which the ratings were based was not related to the accuracy of predictions. Accuracy did not vary significantly between behavior manipulated in the laboratory and more naturally occurring behavior. Last, effect sizes did not differ significantly for predictions in the areas of clinical psychology, social psychology, and the accuracy of detecting deception (considered separately here, although normally part of social psychology). |
