Faking on a self-report personality inventory: Indiscriminate, discriminate, or hyper-discriminate responding?

Abstract This research evaluated an easily-computed novel method, the inter-item standard deviation (ISD), for detecting faking on a self-report personality inventory. Six independent samples of instructed fakers (total N = 1360) answered NEO-FFI items under varying dissimulation directions, modes of testing, and response time limitations. Across samples while demonstrating some variability possibly related to sample-specific factors, the ISD demonstrated a medium effect size for identifying faking good and a medium effect size for identifying faking bad. Further, the direction of these effects indicated that faking is a complex process resulting in heterogeneous responding rather than a simple strategy that produces overly consistent, homogeneous answering.