Differences in performance on the test of variables of attention between credible vs. noncredible individuals being screened for attention deficit hyperactivity disorder.

Measuring performance validity in Attention Deficit Hyperactivity Disorder (ADHD) assessments is essential, with multiple studies identifying how easily young adults can feign symptoms on self-report measures. Few methods, however, exist to identify such feigning when it occurs. While some clinicians include computerized tests of attention (e.g., Test of Variables of Attention [TOVA]) when assessing for possible ADHD, it is unclear how symptom exaggerators perform, and whether the TOVA Symptom Exaggeration Index (SEI) adequately identifies performance-based exaggeration when it occurs. Using archival data from a university-based ADHD screening clinic we investigated the performance of 245 late adolescents/emerging adults. Three groups were created: (1) Good effort but not ADHD ( = 183); (2) Good effort and diagnosed ADHD ( = 13); and (3) suspect effort ( = 49), based on final diagnosis and performance on an existing validity measure. Results showed clearly that those with suspect effort performed more poorly than the other two groups on all but second-half commission errors on the TOVA. Similar to Nicholls et al., the suspect effort group showed significantly subaverage (i.e., greater than two standard deviations below the mean) scores in Omission errors; in this replication, however, this was true for both the first and second half of the test. Response time variability was similarly exaggerated, with the suspect effort group again returning extreme scores in both halves of the test. Suspect effort students were indistinguishable from those with genuine ADHD when looking solely at self-reported symptoms; however, embedded symptom validity measures on an ADHD rating scale discriminated well between groups. Overall, results support the use of the TOVA as an embedded performance validity measure in the assessment of late adolescents/emerging adults and support previous findings that symptom report alone cannot distinguish credible from noncredible ADHD presentation.