The Role of Self in the False Consensus Effect

The false consensus effect disappears if representative information and monetary incentives are given

“…If this were the case, the e¤ect should be larger for morally relevant items than for morally irrelevant items. Sherman et al (1984) …nd evidence for this hypothesis which is doubted by Alicke and Largo (1995).…”

“…2 Each round consisted of three steps. In the …rst step each subject chose between two options of an item (Decision 1).…”

“…The second step began with the random assignment of four (di¤erent) subjects to each individual subject. In half of the rounds 2 That the number of subjects equals the number of rounds is a mere coincidence and of no signi…cance to the experiment. subjects were then informed about these four subjects' Decisions 1 without identifying them.…”

“…Then a false consensus would be considered present if more consensus is estimated than would result from this procedure. Bayesian updating of a uniform prior belief using a sample of size one results in a posterior belief with an expected value of 2 3 for the own decision (see Dawes 1989). For a group of eleven, this means that the expected value of A choices of the posterior belief is 2 3 ¢ 11 = 7: 33 for an own choice of A and 1 3 ¢ 11 = 3: 67 for an own choice of B.…”

“…6 Second, there is no reason to give the expected value as estimate, at least with our incentive system. To maximize the expected payo¤, subjects have to maximize the probability of an exact hit and should hence give the mode of the posterior belief as the 6 In awareness of this possibility Alicke and Largo (1995) designed an experiment to compare the weight of the own decision with information about other subjects. Their design is similar to ours in that they also provide information about four other subjects.…”