In the pages that follow we review the evidence showing that there is little support for the view that people utilize consensus information in making attributions. This evidence concerns both instances where the actor is another person and instances, drawn primarily from our own research, where the actor is the self. We then show the similarity between the failure of consensus information to affect attributions and the demonstration by Kahneman and Tversky that base-rate information fails to affect predictions. We propose explanations for both failures in terms of the relative impact of abstract information versus concrete information. Finally, we apply the distinction between abstract and concrete information to questions of communication and persuasion.
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education