Method for detecting random answers on personality questionnaires

Abstract

Developed and evaluated several strategies for detection of randomly answered protocols for the 16 pf using 250 undergraduates. Cattell's sabotage index predicted poorly, indeed in the wrong direction, between normal and randomly answered sheets (t = 2.10, p = .05). Random and carelessness indexes were derived and predicted very well, providing ts of 15.02 and 12.67 (significant at well beyond the .001 level), respectively, between random and normal sheets. Both indexes, while correctly selecting 86 and 88% of the random answer sheets, respectively, incorrectly picked out only about 6% of the normal group. Since the design of the study included a form of cross-validation, it is suggested that these indexes might be of great practical use. (PsycInfo Database Record (c) 2025 APA, all rights reserved)