Cross-Validating the Atypical Response Scale of the TSI-2 in a Sample of Motor Vehicle Collision Survivors

• Published in: Psychological injury and law Vol. 16; no. 4; pp. 351 - 370
• Main Authors: Nussbaum, Shayna H., Ales, Francesca, Giromini, Luciano, Watson, Mark, Erdodi, Laszlo
• Format: Journal Article
• Language: English
• Published: New York Springer US 01-12-2023; Springer Nature B.V
• Subjects: Accuracy; Behavioral Science and Psychology; Classification; Clinical assessment; Clinical Psychology; Cognitive ability; Credibility; Domestic violence; Generalizability; Law and Psychology; Measures; Medicolegal aspects; Neuropsychology; Personality; Personality tests; Post traumatic stress disorder; Psychology; Psychopathology; Self report; Trauma; Validity scales; Symptom validity, Performance validity; TSI-2; ATR
• ISSN: 1938-971X; 1938-9728
• DOI: 10.1007/s12207-023-09487-z

This study was designed to evaluate the utility of the Atypical Responses (ATR) scale of the Trauma Symptom Inventory – Second Edition (TSI-2) as a symptom validity test (SVT) in a medicolegal sample. Archival data were collected from a consecutive case sequence of 99 patients referred for neuropsychological evaluation following a motor vehicle collision. The ATR’s classification accuracy was computed against criterion measures consisting of composite indices based on SVTs and performance validity tests (PVTs). An ATR cutoff of ≥ 9 emerged as the optimal cutoff, producing a good combination of sensitivity (.35-.53) and specificity (.92-.95) to the criterion SVT, correctly classifying 71–79% of the sample. Predictably, classification accuracy was lower against PVTs as criterion measures (.26-.37 sensitivity at .90-.93 specificity, correctly classifying 66–69% of the sample). The originally proposed ATR cutoff (≥ 15) was prohibitively conservative, resulting in a 90–95% false negative rate. In contrast, although the more liberal alternative (≥ 8) fell short of the specificity standard (.89), it was associated with notably higher sensitivity (.43-.68) and the highest overall classification accuracy (71–82% of the sample). Non-credible symptom report was a stronger confound on the posttraumatic stress scale of the TSI-2 than that of the Personality Assessment Inventory. The ATR demonstrated its clinical utility in identifying non-credible symptom report (and to a lesser extent, invalid performance) in a medicolegal setting, with ≥ 9 emerging as the optimal cutoff. The ATR demonstrated its potential to serve as a quick (potentially stand-alone) screener for the overall credibility of neuropsychological deficits. More research is needed in patients with different clinical characteristics assessed in different settings to establish the generalizability of the findings.