Items' cross-category associations as a confounding factor in the Implicit Association Test

The introduction of the Implicit Association Test (IAT; Greenwald, McGhee, & Schwartz, 1998) has stimulated numerous research activities. The IAT is supposed to measure the degree of association between concepts. Instances have to be assigned to these concepts by pressing appropriate keys as quickly as possible. The reaction time difference between certain conditions, termed the IAT effect, is used as an indicator of the degree of the concepts' association. We tested the hypothesis that the degree of association between one concept (or category) and the instances of the other presented concept also influences reaction times. In our experiment, the instances in the target categories, male and female names, were kept constant. The adjectives in the evaluative categories were manipulated: Either the pleasant adjectives were female-associated and the unpleasant adjectives were male-associated, or vice versa. These stereotypic associations were indeed found to exert a substantial influence on the size of the IAT effect. This finding casts doubt on the assumption that the IAT effect may be interpreted as a pure measure of the degree of association between concepts.