40 Plus or Minus 10, a New Magical Number: Reply to Russell

Abstract

Russell (2009 this issue) has criticized our recently published investigation (Larrabee, Millis, & Meyers, 2008) comparing the diagnostic discrimination of an ability-focused neuropsychological battery (AFB) to that of the Halstead Reitan Battery (HRB). He contended that our symptom validity test (SVT) screening excluding 43% of brain dysfunction and 15% of control patients using computations based on Digit Span inappropriately excluded patients with brain damage, due to the correlation of Digit Span with the Average Index Score (AIS). Our exclusion of 43% of brain dysfunction participants matches the frequency of invalid neuropsychological data of 40–50% or more reported by numerous studies for a wide range of settings with external incentive. Moreover, our study was not an investigation of malingering; rather, we screened our data to insure that only valid data remained, for the most meaningful comparison of the AFB to the HRB. Russell's argument that Digit Span is correlated with brain damage confounds the criterion, AIS (a composite cognitive score), with the predictor, Digit Span (another cognitive score), rather than employing a truly independent neurologic criterion. The fact that Digit Span is notoriously insensitive to brain dysfunction underscores the robustness of our findings, for if we inappropriately excluded brain-damaged patients for low Digit Span, as Russell claimed, this resulted in our sample reflecting more subtle degree of brain dysfunction, and the superiority of the AFB over the HRB was demonstrated under the most challenging of discriminative conditions.

Keywords:

• Base rate
• Invalid data
• Diagnostic discrimination