The informational basis of social judgements: Under what conditions are inconsistent trait descriptions processed as easily as consistent ones?

Abstract This paper is concerned with the conditions under which an effort k made to reconcile inconsistent arguments in a message. It assumes, first, that reconciliation requires additional processing; and second, that it does not occur automatically but only when the set of inconsistent arguments are informative and must be integrated in the representation of the judgment. Subjects were given trait descriptions that varied in consistency as well as in informativeness. Their task was to use these descriptions in forming a judgement about a hypothetical person. It was found that the impact of consistency on the ease with which the judgement was made, as indicated by its latency, was more pronounced when the descriptions were highly informative than when they were relatively uninformative: It took much longer for an inconsistent set of trait descriptions to give rise to a judgement than a consistent set if the set was informative; if the descriptions were uninformative, the difference in the latency of judgement diminished considerably. These effects support the hypothesis that the amount of processing invested in reconciling inconsistencies among arguments is determined by the informativeness of these arguments. After making their judgement, subjects were given a surprise recognition memory test. The results provided further support for the notion that highly informative arguments are elaborated to a larger extent than uninformative arguments, namely, recognition memory for the former was superior to that for the latter.