Naive Attributors' Attributions and Predictions: What Is Informative and When Is an Effect an Effect?

Wells and Harvey (1977) critically evaluated past research that suggested people do not use consensus information when making predictions and causal attributions. Borgida (1978) takes issue with a subset of Wells and Harvey's conclusions. Below, we address the issues raised by Borgida. In particular, it is argued that (a) Borgida incorrectly states the position of Wells and Harvey concerning the issue of ignoring versus underutilizing consensus information; (b) Borgida incompletely summarizes Wells and Harvey's attribution data; (c) Kahneman and Tversky (1973) statistically controlled for target description informativeness, while Nisbett and Borgida (197S) used no control for target description informativeness; (d) Kahneman and Tversky's presentation of data in their lawyerengineer study tended to underrepresen t the majority of their subjects' data; and (e) a replication of Kahneman and Tversky's lawyer-engineer study, using a more appropriate