A-206 “Mittenberg-Type formula for Identifying Feigned ADHD Using WAIS-IV”
Abstract Objective This study utilized logistic regression to determine whether performance patterns on the WAIS-IV subtests could differentiate between genuine-effort and simulated ADHD groups. Method Participants were 355 college students (55.3%% female; 65.7% Caucasian, 23.9% African American, 4.2% Asian; age range 17–51 years, Mage = 20.93 years, SD = 4.63; 76.4% no psychological diagnosis) who completed the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV). Mean FSIQ for control group was 102.35 and 87.48 for the simulator group. Eighty participants (22.5%) simulated ADHD symptoms, while 276 participants (77.5%) provided genuine effort. Results A logistic regression analysis was performed with known group as the dependent variable and WAIS-IV subtest scaled scores as predictor variables. The final model of five predictor variables (Similarities, Information, Digit Span, Symbol Search, and Coding) significantly predicted group status (χ2 = 140.91, df = 5, N = 355, p < .001). The model accounted for 33% to 50% of the variance (Cox and Snell R2 = .33; Nagelkerke R2 = .50) in performance classification with overall 86% of individuals correctly predicted to their known group. At a cut value of .65 the sensitivity was 69.6% and the specificity was 90.6%. A cut value of.55 the sensitivity was 53.2% and the specificity was at 94.6%. Conclusions These findings are consistent with prior research suggesting discriminant function analysis derived from clinical measures (e.g., WAIS-IV) can be useful as embedded validity measures in distinguishing suspicious and genuine performance. Sensitivity, specificity, and predictive power of the WAIS-IV subtests were acceptable at 5% and 10% false positive rates.