The Truly False Consensus Effect: An Ineradicable and Egocentric Bias in Social Perception

Consensus bias is the overuse of self-related knowledge in estimating the prevalence of attributes in a population. The bias seems statistically appropriate (Dawes, 1989), but according to the egocentrism hypothesis, it merely mimics normative inductive reasoning. In Experiment 1, Ss made population estimates for agreement with each of 40 personality inventory statements. Even Ss who had been educated about the consensus bias, or had received feedback about actual consensus.or both showed the bias. In Experiment 2, Ss attributed bias to another person, but their own consensus estimates were more affected by their own response to the item than by the other person's response. In Experiment 3, there was bias even in the presence of unanimous information from 20 randomly chosen others. In all 3 experiments, Ss continued to show consensus bias despite the availability of other statistical information. In a study on student attitudes, Katz and Allport (1931) noticed that the more students admitted they had cheated on an exam, the more they expected that other students cheated too. Since then, more than a hundred studies have documented a systematic relationship between people's perceptions of their own characteristics and their estimates of the percentage of people in the population who share those characteristics. Early investigators assumed that the cause of this relationship is that people irrationally project their own characteristics onto others. Much research effort was dedicated to the examination of the psychological causes of projection (Holmes, 1968). Ross, Greene, and House (1977) considered projection to be a consensus bias (i.e., the "false-consensus effect") and introduced it to the attribution and decision-making literature. These authors reinforced the idea that consensus bias is irrational. This argument has two parts. First, a person's own response to a judgment item is a single-case sample. To the extent that other social information is available, the self-related single-case sample provides little information and should be ignored in the inference process. Second, if consensus estimates vary with the person's own response, at least some of the estimates must be incorrect. If raters ignored their own responses, there would be no differences between the mean estimates of people with different responses.