Relations between implicit measures of prejudice:what are we measuring?

Some recent findings suggest that different implicit measures of prejudice assess the same underlying construct, but other work suggests that they may not. In this experiment, White participants completed a version of a priming measure of racial attitudes that either encouraged categorization of the face primes in terms of race or did not encourage such categorization, and then completed the Implicit Association Test. Correspondence between the two measures was found only when categorization by race was required on the priming measure. Moreover, participants appeared more prejudiced when they were led to construe individuals in terms of race than when they were not so encouraged. The discussion focuses on the potential for dissociations between evaluations of a category and evaluations of members of the category.