How easy is it to effectively fool a personality test?
This study provides a partial test of the model of faking proposed by McFarland and Ryan (2000) by examining the degree to which self-monitoring, knowledge of the constructs being measured, job familiarity, and openness to ideas account for variance in ability to fake on a personality measure. Undergraduates (N = 342) completed a modified version of the NEO Personality Inventory-Revised under both honest and faking instructions. In addition, some participants were asked to “fake good” and others were asked to fake toward the requirements of a specific job (i.e., accountant). Consistent with prior research, the fake good manipulation was found to increase scores on 8 of the 9 personality variables. In contrast, the fake accountant manipulation resulted in a personality profile consistent with a priori hypotheses in which some personality scores were significantly increased whereas others were significantly decreased. Results also revealed that the individual difference variables explained a significant portion of the variance in the ability to fake and that the two faking conditions produced different relations with these individual difference variables. As hypothesized, the participants’ openness to ideas was a significant predictor of the ability to fake like an accountant (β = .22) but not the ability to fake good. In contrast, the participants’ knowledge of the constructs being measured was a significant predictor of the ability to fake good (β = .18) but not the ability to fake like an accountant. We conclude by elaborating on the importance of using a hypothesis testing approach in the assessment of the relations between personality variables and job performance.
Source: “Individual Differences in the Ability to Fake on Personality Measures” from Human Performance, Volume 22, Issue 1 January 2009 , pages 86 – 103