Measuring individual differences in implicit self-concept: initial validation of the self-apperception test

Implicit self-concept as a prominent form of implicit social cognition (Greenwald and Banaji, 1995) has been eluding psychometric quantification, in the absence of indirect measurement procedures that are sensitive to implicit self-representations. The capacity of a new Self-Apperception Test (SAT; Aidman, 1997) to measure implicit self-attitudes is examined. The instrument employs Fedotovaʹs (Fedotova, 1987) version of Liggettʹs (Liggett, 1959) projective facial images in a visual semantic differential format. In four studies, the SAT showed acceptable levels of internal consistency and retest stability, and promising validity characteristics. Its Implicit Self-Liking Index (ISL), despite weak correlations with Coopersmithʹs (Coopersmith, 1981) and Rosenbergʹs (Rosenberg, 1965) self-esteem scales, significantly correlated with Tafarodi and Swannʹs (Tafarodi and Swann, 1995) self-liking scale, a theoretically constructed measure of self-worth (a subdimension of global esteem). The ISL and its twin index of Implicit Self-Confidence, correlated with theoretically derived markers of implicit self-esteem _ initial letter preferences (Nuttin, 1987) and self-positivity projection (Greenwald and Banaji, 1995) _ while neither of the questionnaires did. Results confirmed a distinct ontological status of the implicit measures and indicated directions for their further validation. 1999 Elsevier Science Ltd. All rights reserved.