JOURNAL OF EXPERIMENTAL SOCIAL PSYCHOLOGY 23, 217-229 (1987) Accuracy of Estimates in the Social Comparison of Abilities’s2 OFRA MAYSELES~ AND ARIE W. KRUGLANSKI Tei-Aviv University Received June 18, 1986 An experiment was conducted to test the proposition that comparison with others similar to oneself in performance affords more accurate self-evaluations than comparison with dissimilar others. A previous study presumably demonstrating this effect (R. Radloff, 1966, Journal of Experimental Social Psychology, Sup- plement, 1, 6-26) was reinterpreted. It was suggested that in that particular experiment, performance dissimilarity of comparison others may have been con- founded with subjects’ disconfirmed expectancies concerning own performance. In the present experiment, similarity of comparison others (manipulated via performance extremity relative to others in the group) was orthogonally manipulated to consistency (versus inconsistency) with an expectancy. As predicted, incon- sistency with expectancy exerted adverse effects on the accuracy of self-evaluations whereas dissimilarity from comparison others did not. Significance of the findings for social comparison processes was discussed. 0 1987 Academic Press, Inc. Festinger’s (1954) theory of social comparison processes counts among the most important formulations in contemporary social psychology. Pri- marily, the theory addresses factors which determine people’s reliance on the opinions or perceived abiiities of others in the course of evaluating own opinions and abilities. A major tenet of the theory has been that similar as opposed to dissimilar others represent the more efficient standard of comparison. Accordingly, the theory predicted that persons would manifest (1) preference for comparisons with similar rather than dissimilar others, (2) greater confidence in opinions, and (3) greater accuracy of evaluations based on comparisons with similar versus dissimilar others. Address correspondence and requests for reprints to Arie W. Kruglanski, Department of Psychology, University of Wisconsin-Madison, 1202 W. Johnson Ave., Madison, Wl 53706. ’ This paper was written while Kruglanski was on sabbatical leave from Tel-Aviv University and at the University of Wisconsin-Madison, and Mayseless held a post-doctoral appointment at the University of California, Berkeley. * The present research is part of a doctoral dissertation conducted by the second author under the first author’s supervision. This research was supported by Foundation Trustees research fund for Ph.D. candidates, Grant No. 4, Project A615 (1981-1983). 217 0022-1031187 $3.00 Copyright Q 1987 by Academic Press. Inc. All rights of reproduction in any form reserved.