Examining reaction time variables to detect invalid performance on the TOMM 2: comparing mTBI to simulators using two procedures.

Two studies are presented involving simulator designs to investigate the utility of reaction time (RT) indices from the first two trials of the computerized TOMM 2. A sample of largely mild traumatic brain injuries (TBI) was compared to simulators whose correct responding was below the traditional passing score of 45/50 but not significantly below chance. In Study 1, 47 healthy participants were presented a scenario in which they were given the opportunity to malinger for financial compensation; a procedure that more closely parallels real-world settings where individuals are incentivized to malinger but are not explicitly instructed to do so. Reaction time indices for those who chose to simulate were compared to 30 individuals who experienced a diagnosed TBI. Both the Coefficient of Variation (CoV) of RT and the standard deviation of RT for Trial 2 differentiated simulators from those with TBIs. In Study 2, 69 individuals were explicitly instructed to feign in a more traditional simulator design and compared on the TOMM 2 reaction time indices. The findings from Study 1 were replicated, showing both standard deviation and CoV of RT in Trial 2 differentiated between TBIs and simulators, and the effect size for the CoV of RT was larger. Classification statistics were maximized using a base rate of 40%. The present findings provide preliminary evidence for the utility of the TOMM 2 reaction time measures in detecting subnormal performance, especially when respondents score below the typical passing score of 45, but do not score significantly below chance on the performance validity test.