The utility of the Personality Clinical Form Indicators of Response Distortion: Receiver Operating Characteristic Analysis

Abstract

A Receiver Operating Characteristic Analysis (ROC Analysis) was conducted to assess the efficiency of six validity scales included in the Personality Clinical Form (PCF) to detect responses distortion. Undergraduate students were randomly assigned to simulate malingering, simulate defensiveness or complete PCF under standard instructions (no faking). Fake-good participants scored significantly higher than standard participants on all underreporting scales. The difference observed was even higher when the comparison was made between the fake-good and the fake-bad participants. Likewise, a reverse trend was observed for the overreporting scales. Participants in the fake-bad condition scored the highest, and the participants in the fake-good condition scored the lowest on all overreporting scales. Large effect sizes were found in most cases. The responses resulted from the malingering condition were also compared with those obtained from psychiatric inpatients. The responses resulted from the defensiveness group were also compared with responses obtained from employees in a high-stake assessment condition. The area under the ROC curve (AUC) provided an index of discriminative power. The validity scales discriminate better between the normal and the fake conditions than between malingerers and psychiatric inpatients, but most AUC values were within good or excellent range. Cut-off scores and their corresponding sensitivity and specificity were presented for each validity scale based on this explorative endeavour