Exploration of an Embedded Symptom Validity Overreporting Cutoff for the Perceived Stress Scale in a Diverse Adult Outpatient Neuropsychological Sample

Introduction Perceived stress is routinely assessed in clinical evaluations. Ensuring symptom validity is critical when interpreting symptom measures and embedded symptom validity cutoffs have recently been explored for several widely used measures of psychological distress. This study explored potential embedded validity cutoff scores in the 10-item Perceived Stress Scale (PSS) for detecting psychological symptom overreporting. Methods The final sample included 503 patients referred for neuropsychological evaluation who were independently classified into valid, possibly overreporting, or definitely overreporting groups based on the Minnesota Multiphasic Personality Inventory-2-Restructured Form (MMPI-2-RF) or Minnesota Multiphasic Personality Inventory-3 (MMPI-3) overreporting scales. Receiver operator characteristic curve analyses determined the optimal PSS cutoff for detecting noncredible responding using both the total raw score and race-corrected z-scores. Results Classification accuracy ranged from unacceptable to excellent for PSS raw scores and z-scores, with greater classification accuracy corresponding to greater likelihood of overreporting. Among models with acceptable or excellent classification accuracy, a cutoff of ≥28 produced 37%–57% sensitivity and 88% specificity for PSS raw scores. For PSS z-scores, a cutoff of ≥2.15 yielded 46% sensitivity and 89% specificity for detecting ≥3 possible overreporting elevations and ≥2.08 yielded 61% sensitivity and 88% specificity for detecting ≥1 definite overreporting elevation. Conclusion Present findings provide preliminary evidence for an embedded overreporting cutoff for the PSS. While these results are promising, there remains a need for further replication in other clinical populations before it can be routinely adopted in general evaluations.