Open Access Dissertation
Date of Award
Lisa J. Rapport
Performance validity test (PVT) inaccuracies can be explained by both test and extra-test (e.g., research design components) factors. Eye-tracking is a promising technology to enhance assessment of performance validity. Prior research has established that ocular behaviors are reliable biomarkers of (un)conscious cognitive processes. Experimental research on deception has shown that ocular behaviors reliably distinguish feigned concealment of information from honest responding. The primary objective of this study was to examine the incremental utility of incorporating eye-tracking into a clinical PVT to distinguish adults with verified TBI from adults coached to feign cognitive impairment. A secondary objective was to determine the effect of financial incentive and preparation time on individuals’ ability to simulate cognitive impairment.
Participants were adults with moderate to severe TBI (TBI; n = 49), healthy adults coached to simulate TBI in a single session (SIM-SC; n = 46) and those given time to prepare and financial incentive (SIM-IP; n = 49), and healthy adult comparisons providing full effort (HC; n = 67). A computerized version of the Warrington Recognition Memory Test of Words (RMT) was completed in the context of a full neuropsychological battery.
Eye-tracking indices generally did not differ among the groups during presentation of stimulus items but did differ during decision-making (forced-choice) trials. Compared to TBI and HC, SIM-IP had significantly more transitions, fixations, and time spent looking at correct and incorrect response options. Logistic regressions and ROC curve analyses showed that accuracy was the best predictor of SIM-IP vs. HC. For SIM-IP vs. TBI, eye-tracking variables exceeded accuracy in distinguishing the groups. Eye-tracking added incremental predictive value to accuracy in discriminating SIM-IP from HC and TBI.
SIM-IP performed worse than SIM-SC on tasks of processing speed and executive functioning; the two simulation groups performed equivalently on all other cognitive tests. SIM-IP performed better than SIM-SC across forced-choice PVTs with small effects; however, groups did not differ with respect to rates of successful simulation. Overall, the combined effect of preparation time and financial incentive led to a pattern of neuropsychological test performance that was more consistent with patterns observed in TBI. Despite this finding, individuals receiving incentive and preparation time were not more likely to accomplish the challenging goal of successful simulation, which requires performing impaired on tests of cognitive ability while at the same time avoiding PVT detection.
Eye-tracking indicated that persons feigning TBI showed multiple signs of greater cognitive effort than persons with verified TBI and healthy comparisons. Effectiveness of RMT accuracy and eye-tracking depended on the groups compared. In the comparison of greatest interest (SIM vs. TBI) eye-tracking best predicted group status and led to “excellent” discrimination when combined with accuracy. Successful simulation was achieved by 32 – 40% of the simulators, depending on the criterion for “impairment;” remarkably, eye-tracking indices improved detection of feigning even among this group. Eye-tracking may be an important complement to traditional accuracy scores on PVTs.
Kanser, Robert John, "The Eyes Never Lie: Detecting Simulated Traumatic Brain Injury With Eye-Tracking" (2019). Wayne State University Dissertations. 2205.