| Abstract: | Using a sample of individuals undergoing medico-legal evaluations (690 men, 519 women), the present study extended past research on potential gender biases for scores of the Symptom Validity (FBS) scale of the Minnesota Multiphasic Personality Inventory-2 by examining score- and item-level differences between men and women and determining the extent to which FBS scores were able to correctly identify men and women who were divided into credible responders (n = 837) and noncredible responders (n = 372) on the basis of performance on symptom validity tests. Results indicated that women had slightly higher raw FBS scores than men (d = .29), and significant differences between men and women in item endorsement were demonstrated for 14 FBS items. Step-down hierarchical logistic regression procedures indicated predictive bias (χ2Δ = 23.72, p < .001). Follow-up analyses indicated intercept bias (χ2Δ = 23.51, p < .001) but not slope bias (χ2Δ = 0.22, p = .64). However, using the test publisher's recommended FBS cutoff scores (Ben-Porath, Graham, & Tellegen, 2009), classification accuracies were similar for women and men (T > 80, h = -.02; T > 100, h = -.22, respectively). On the basis of these results, we conclude there is no evidence of clinically meaningful bias in predictions of symptom validity test failure made using FBS scores for men and women. (PsycINFO Database Record (c) 2012 APA, all rights reserved). |
