Detecting Attitude Change with the Implicit Association Test

Cahier de recherche du Groupe HEC Paris, n° 888The Implicit Association Test and its variants have become pervasive measures of attitudes in a variety of domains and contexts. In two experiments, we provide evidence that a recent variant, the Personalized IAT developed by Olson and Fazio (2004) may more accurately detect changes in personal attitudes than the conventional Traditional IAT devised by Greenwald, McGhee, and Schwartz (1998). Our findings suggest that the Personalized IAT may be more sensitive to detecting attitude changes than the Traditional IAT because it is less affected by extrapersonal associations (i. e. salient associations not contributing to personal evaluations of the object). More generally, this research suggests that for attitude domains characterized by potentially strong extrapersonal associations, using the Personalized and Traditional IATs may provide researchers with complementary insights about knowledge structures