Effects of response distortion on the validity of Implicit Association Tests of personality

Self report measures of personality traits have been shown to be susceptible to response distortion or "faking." It has been suggested that Implicit Association Test (IAT) trait measures be used for personnel selection because participants are less able to distort their responses on these tests. Participants (N = 400 university students) were randomly assigned into 1 of 2 groups. The first group (honest condition; n = 200) completed a series of self report (explicit) and IAT (implicit) trait measures honestly. The second group (applicant condition; n = 200) was given a job description for a sales position and told to complete the tests as though they were part of an employment testing process for this job.;Correlational analyses and Confirmatory factor analytic (CFA) model contrasts indicated IAT trait measures showed weak convergence with self report measures. However, IAT trait measures showed evidence of discriminant validity that was as good or better than self report measures, and showed stronger relationships to supervisor ratings of job performance and academic performance compared to self report measures.;IAT trait measures also showed a method confound related to positivity of adjective valence (implicit self esteem) and cognitive ability. In spite of the solid performance of the IAT, the construct validity of IAT trait measures needs to be evaluated further before they are used for selection purposes.