Preliminary validation of the computerized N-Tri-A Tri-Choice naming and response bias test

The study describes the validation of a computerized adaptation of the novel Tri-Choice Naming and Response Bias Measure (N-Tri) developed to detect untruthful responding while being less susceptible to coaching than existing measures. We hypothesized that the N-Tri would have comparable sensitivity and specificity to traditional tests but would have improved accuracy for detecting coached simulators. Four-hundred volunteers were randomly assigned to one of three groups: uncoached simulators' group (n = 118), coached simulators' group (n = 136), or control group (n = 146). Both simulator groups were asked to feign concussion symptoms, but the coached group received a test-taking strategy and a description of concussion symptoms. The participants were administered the computerized version of the new measure in conjunction with computerized adaptations of two well-validated response bias tests commonly used to detect cognitive malingering, the Reliable Digit Span (RDS) and Portland Digit Recognition Test (PDRT). Our data show the new measure correlated highly with other established measures. However, the classification accuracy did not significantly increase when compared to the traditional tests. Our findings support that the N-Tri performs at a comparable level to existing forced choice measures of response bias. Nevertheless, the N-Tri could potentially improve the detection of response bias as existing tests become more recognizable by the public.

Automated summary

The study validates a computerized adaptation of the Tri-Choice Naming and Response Bias Measure (N-Tri), which aims to detect untruthful responding and minimize susceptibility to coaching. Results show that while the N-Tri correlates highly with established measures, its classification accuracy does not significantly improve compared to traditional tests. The findings suggest that the N-Tri performs similarly to existing forced choice measures of response bias.