Utility of two new embedded measures of performance validity for the Wisconsin Card Sorting Test-64

Area of Science:

• Neuropsychology
• Clinical Psychology
• Forensic Psychology

Background:

• Performance validity testing is crucial in clinical assessments, particularly for individuals with traumatic brain injury (TBI).
• Existing embedded validity indices require further validation in diverse clinical populations.
• The Wisconsin Card Sorting Test-64 (WCST-64) is a widely used measure sensitive to executive function deficits common in TBI.

Purpose of the Study:

• To evaluate the diagnostic utility of two novel embedded validity indices for the WCST-64.
• To determine if these indices meet established criteria for specificity (≥.90), sensitivity (≥.40), and positive likelihood ratio (≥2).
• To assess index performance in a clinical sample of individuals with TBI across a spectrum of injury severity.

Main Methods:

• Logistic regression analysis was employed to assess the discriminative ability of each index.
• 173 individuals with TBI, evaluated 1-36 months post-injury, were classified into valid (n=146) or invalid (n=27) performance groups based on established criteria.
• The performance of the Henry (2024) and Kosky et al. (2022) indices was compared against these classifications.

Main Results:

• Both indices demonstrated acceptable likelihood ratios.
• The Kosky et al. index exhibited suboptimal sensitivity and specificity.
• The Henry index achieved acceptable sensitivity (.41) and specificity (.88) but frequently resulted in false positives when indicating invalid performance.

Conclusions:

• The Kosky et al. index did not meet the predefined performance criteria.
• The Henry index, while more robust, identified over half of the "invalid" cases as false positives, limiting its utility for confirming invalidity.
• Findings suggest the Henry index is more effective for ruling out invalid performance rather than ruling it in, potentially due to base rate differences.