Processing fluency as a predictor of salience asymmetries in the Implicit Association Test

Abstract

The Implicit Association Test (IAT) is the most popular indirect measure of attitudes in social psychology. Rothermund and Wentura (2001, 2004) suggested that artifacts such as salience asymmetries are a source of compatibility effects in the IAT, and, therefore, the IAT does not necessarily measure attitude. They claim that salience asymmetries correspond with visual search asymmetries, such that the stimulus categories that are more quickly detected in a visual search task are also compatible in the IAT. We propose that processing fluency is a more reliable indicator of salience asymmetries in the IAT than are visual search asymmetries. To test this hypothesis, we set processing fluency in opposition to visual search asymmetry to see which variable better predicted IAT effects. In one pair of categories, the category that was more quickly detected in visual search was also more fluently processed in a binary classification task. In a second pair of categories, the category that was more quickly detected in visual search was the less fluently processed category. Across four experiments, we demonstrated that compatibility effects in the IAT corresponded with differences in processing fluency between categories, rather than with visual search asymmetries.

Keywords:

• Implicit Association Test
• Salience asymmetry
• Figure–ground asymmetry
• Processing fluency
• Visual search asymmetry

Acknowledgments

We thank Sachiko Kinoshita, Klaus Rothermund, and an anonymous reviewer for their helpful comments on a previous draft of this paper. We are also grateful to Klaus Rothermund for his advice in preparing Experiments 3a and 3b.

Notes

¹ The IAT experiments used a 300-ms criterion consistent with standard IAT experiments (e.g., Greenwald et al., 1998). This minimum outlier value is higher than that adopted in the binary classification tasks (200 ms), which is in keeping with the relative difficulty of the two tasks.