Perceived Consensus and Predictive Accuracy: The Pros and Cons of Projection

Accurate prediction of other people is an important social activity. The standard assumption is that we rely too heavily on our own attitudes and behaviors, thereby suppressing predictive performance. The phenomenon is often referred to as false consensus or projection . This article examines the effect of high perceived consensus/projection on predictive accuracy and identifies those conditions where perceiving consensus indeed diminishes performance and where it is a reasonable prediction strategy. The task of predicting a target population was considered from a cognitive or informational perspective: Subjects were viewed as having to first identify and then correctly weight various predictive cues, in this case their own attitudes and other specific or stereotypic information about the target. A model was developed to show how subjects should optimally weight their own positions compared with whatever other target-related information was available to them. Subjects predicted the attitudes of one of three target populations: the average, married American consumer; their average graduate school peer; or their spouse. The results indicated that although perceived consensus was quite high, 65% of the subjects could have actually increased their predictive accuracy by weighting their own positions even more. The heuristic value of relying on own positions varied dramatically depending on the target population: Although 63% of subjects predicting the attitudes of consumers projected too much, only 16% and 24% of subjects predicting attitudes of peers and spouses overprojected. Even when subjects perceived more consensus than actually existed, they often could have increased predictive accuracy by relying more heavily on their own attitudes because they seemed to have difficulty in identifying and consistently using other information about the target.