A meta-analytic investigation of the susceptibility of integrity tests to faking and coaching

Although it has been consistently found that test takers can effectively fake good on self-report noncognitive measures when instructed to do so, not all measures are equally susceptible. The present review meta-analytically synthesized studies that have investigated the extent to which individuals can inflate their integrity test scores when coached or instructed to fake good. Both overt and personality-based integrity tests were investigated. Results indicated that the overt test was especially susceptible to both fake good (d = 0.90) and coaching (d = 1.32) instructions. Personality-based measures appeared to be more resistant to both faking good (d = 0.38) and coaching (d = 0.36). Implications of these results for integrity testing are discussed.