“Sandbagging” Baseline Concussion Testing on ImPACT is More Difficult than It Appears

Archives of Clinical Neuropsychology -


Glatts, C. and P. Schatz.



Objective: Claims and beliefs that athletes intentionally malinger (or “sandbag”) on baseline assessments have not been systematically or prospectively evaluated. We sought to identify the utility of ImPACT, Medical Symptom Validity Test (MSVT), and Balance Error Scoring System (BESS) in detecting naı¨ve versus “coached” malingering and identify patterns in performance on these measures. Methods: Sixty participants were assigned to one of three groups, using revised scripts (Mittenberg, 1993) to guide performance: Optimal effort, Naı¨ve malingerers, Coached malingerers, and completed ImPACT, MSVT, and BESS. Performance below 90% was considered malingering for MSVT; ImPACT automatically “flagged” cases as “invalid.” Results: All participants in the optimal effort group provided valid test results. However, the MSVT identified more participants in the Naı¨ve (90%) and Coached (90%) groups than ImPACT (70% and 65%, respectively). For those participants not “flagged” by ImPACT, mean performance remained 1–3 SD below normative data. MANOVA revealed a significant effect of malingering group on all ImPACT Composite and Symptom scores (p , .001) and BESS scores (p , .034). Post hoc tests revealed that naı¨ve malingerers were unable to distinguish themselves from coached malingerers on Verbal and Visual Memory and Symptom Scores. Coached malingerers performed similar to controls on Impulse Control and were not significantly different from malingerers or controls on Reaction Time or the BESS. Conclusions: Intentional “sandbagging,” naı¨ve or coached, can be readily detected on ImPACT. Relative to naı¨ve malingerers, exposure to “coaching” resulted in some improvement performance, but few of these improvements were statistically significant.

Links to full article: