RELEVANCE OF INFORMATION AND SOCIAL-INFLUENCE IN THE PSEUDODIAGNOSTICITY BIAS

Two experiments investigated the socio-cognitive mechanisms intervenin g in the ''pseudodiagnosticity bias''. In Study 1 positive vs. negativ e anchoring information concerning one of four characteristics of two (A vs. B) political candidates or two cars were presented. Subjects' t ask was to decide which other single additional piece of information a bout A or B should be obtained in order to be able to choose between t he two alternatives. Results show that diagnosticity is enhanced when the anchoring information is negative; and when the anchoring informat ion bears on a characteristic that is highly relevant for the subjects . Study 2, conducted on the same tasks, investigated the influence of a majority vs, a minority, the effect of positive vs, negative anchori ng information relative to a highly vs. low relevant characteristic, i n a context where the risk of error was either low or high (making sal ient the random character of the choice vs. the risk of error in the t ask). Results show an effect of the relevance of characteristic on whi ch the information was given, as in Study 1, and an interaction effect between choice vs. error task, majority vs. minority source and posit ive vs, negative evaluation. Addition of these two effects reveals tha t subjects are more diagnostic to the extent that a negative evaluatio n is given by the minority source about a highly relevant characterist ic in the task where the risk of error is salient.