The Karp Inkblot Response Questionnaire: an evaluation of social desirability responding

Abstract

A reputed advantage of projective psychological measures such as the Rorschach Test over self-report psychological tests has been projectives' comparative freedom from social desirability response (SDR) bias. SDR bias is generally defined as the tendency for subjects to respond to personality test items in a manner that consistently presents the self in a favorable light (Wiggins, 1964). Research has shown that self-report instruments such as the Minnesota Multiphasic Personality Inventory (MMPI) are susceptible to such bias (Meehl & Hathaway, 1946). However, to date projectives’ comparative freedom from SDR bias has only been a reputed advantage, lacking empirical support Additionally, critics of projective measures in general and of the Rorschach in particular, have repeatedly pointed out varying scoring and interpretation approaches among clinicians as lacking in uniformity or consistency. The e purpose of the present study was to assess the susceptibility to SDR bias of a new objectively-scored measure intended to be administered in conjunction with the Rorschach. The Karp Inkblot Response Questionnaire (KIRQ) requires a subject to complete a questionnaire about his/her first responses to each of the ten Rorschach plates. The KIRQ was designed to draw upon the strengths of both projective and self- report tasks, hopefully incorporating projectives’ reputed freedom from SDR bias with self- report measures’ reliable and uniform data collection techniques. 103 subjects (18 males and 85 females, mean age = 21.56 years) participated in the study. Subjects completed the MMPI-2, Marlowe-Crowne Inventory, Rorschach Inkblot Test, and the KIRQ. Three hypotheses were tested: (a) Rorschach and KIRQ scales would show no significant relationship with SDR measures, (b) MMPI-2 scales would show significant relationships with SDR measures, and, (c) support would be found for existing KIRQ validity data. Hypotheses (a) and (b) were tested using stepwise multiple regression, while hypothesis (c) was tested using correlations between individual measures. Overall Rorschach and KIRQ scales showed little significant relationship to measures of SDR bias, namely the MMPI-2 L scale, MMPI-2 K scale, and Marlowe Crowne Inventory. Comparatively, MMPI-2 scales showed significant relationships with SDR measures more frequently. Validity findings were more variable, providing data corroborating existing validity figures as well as data contradicting existing findings. Rndings were interpreted using Paulhus’ (1984) model of SDR which breaks social desirability responding into self-deception and impression management Limitations of this study and recommendations for future research are discussed.