Based on a Selective Accessibility (SA) model of comparison consequences, i
t is suggested that the self-evaluative effects of social comparisons depen
d on the nature of the hypothesis that is tested as a starting-point of the
comparison process. If judges test the hypothesis that they are similar to
the standard, then standard-consistent self-knowledge is rendered accessib
le so that self-evaluations are assimilated towards the standard. If judges
test the hypothesis that they are dissimilar from the standard, however st
andard-inconsistent self-knowledge is made accessible so that self-evaluati
ons are contrasted away from the standard. These predictions are tested by,
inducing participants to test for similarity versus dissimilarity to the s
tandard via a procedural priming manipulation. Consistent with the SA model
, assimilation occurs if participants are procedurally primed to focus on s
imilarities to the standard, whereas contrast results if they, are primed t
o focus on dissimilarities. These findings suggest that similarity versus d
issimilarity testing is a crucial determinant of assimilation versus contra
st. It is proposed that distinguishing between these two alternative hypoth
eses may, provide an integrative framework for an understanding of the self
-evaluative consequences of social comparisons. Copyright (C) 2001 John Wil
ey, & Sons, Ltd.