Response Manipulation on the Multiple Affect Adjective Check List-Revised

Abstract

The scoring pattern of the Multiple Affect Adjective Check List-Revised (MAACL-R) under both "simulate good" and "simulate bad" conditions and under two instructional sets, either "simulate" (Sample 1) or "simulate with caution" (Sample 2), was studied in an experiment in which the MAACL-R state form was administered to two groups of college students, each consisting of 160 participants. Each participant was tested twice with both a baseline "actual" trial and either a "simulate bad" or "simulate good" trial with the "actual" and experimental "simulate" trials counterbalanced. Participants in Sample 2, but not Sample 1, received instructions to be "cautious so as not to be detected." As expected, the scoring pattern of the MAACL-R was susceptible to response manipulation in the predicted directions. However, instructions to "simulate with caution" did not produce significantly different scores than did instructions to "simulate" alone. In the second study, Receiver Operating Characteristics (ROC) was used to determine the utility of each MAACL-R scale to detect response simulation. For both "simulate bad" and "simulate good," the PA and PASS scales showed the highest correct classification rates. However, detection of simulating bad was more accurate than determination of simulating good.