This study investigated the effects of faking bad (FB) on the Test of Variables of Attention (TOVA) using subjects randomly placed into two groups. Subjects in Group 1 took the TOVA under normal conditions (NC) first; they were then requested to subtly fake bad. Group 2 subjects took the TOVA under the same fake bad instructions first, then took the test under normal conditions the second time. An analysis of the effects of test order yielded non-significant differences for basic TOVA variables across all four quarters, both halves and the total score. An analysis for group mean differences between the NC and the FB instructions yielded significant differences across the basic TOVA variables across the four quarters, two halves and total score. The FB group had excessive amounts of omission and commission errors, a greater response time mean (i.e., slower to respond) and had greater variance around their mean response time. The study affirms that the professional using the TOVA needs to carefully eliminate a fake bad test-taking bias when subjects produce excessive test results.

Author notes

Portions of this paper were presented as a poster paper at the 19th Annual Meeting of the National Academy of Neuropsychology, San Antonio, TX, November 1999. The TOVA Research Foundation supplied use of the test for this research project.