Auditory implicit association tests

Abstract

Three studies demonstrated that auditory implicit association tests (auditory IATs) using speech and environmental-sound stimuli could be used to measure implicit attitudes and stereotypes. Study 1 showed a preference for songbirds over insects in auditory IATs that used (a) speech stimuli (spoken names of songbirds and insects), and (b) stimuli consisting of sounds made by songbirds and insects, in conjunction with text stimuli consisting of positive and negative words. Study 2 showed that two IATs using only auditory stimuli could measure implicit gender stereotypes. The first used male and female voices in conjunction with high-power and low-power words spoken by a non-gendered computerized voice to measure sex differences in gender stereotypes. The second used the same high- and low-power words in conjunction with male and female names spoken by the non-gendered voice. Study 3 showed implicit attitudes favoring European-Americans over African-Americans using an auditory IAT in which participants classified speech segments based on the accent of the speaker, thus measuring personally and/or socially undesirable implicit racial attitudes that were not detected through explicit self-report measures. The studies demonstrate that the auditory IAT is an extension of the IAT that can measure implicit associations in a variety of situations in which IATs consisting entirely of visual or text stimuli are not applicable. Exploration of the relationships between the various IATs suggested that their results reflect automatic cognitive associations that are influenced by factors within the IAT procedure, such as the stimuli chosen to represent the target categories, and external contextual factors such as associations made salient by preceding experiences. For example, Study 3 found that a picture IAT using neutral faces measured significantly stronger implicit attitudes favoring European-Americans over African-Americans than were measured by a sound IAT using friendly greetings. However, the difference was only observed when the picture IAT was completed first. When the sound IAT was first, the associations with friendly targets apparently remained salient, and implicit attitudes did not differ. Such findings suggest possibilities for better understanding automatic associations and the way they are measured by the IAT.