Novel ImPACT validity indices in collegiate student-athletes with and without histories of ADHD or academic difficulties.
Clin Neuropsychol. 2018 Dec 05;:1-12
Authors: Manderino LM, Zachman AM, Gunstad J
OBJECTIVE: Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT) assesses neurocognitive functioning in sports-related concussion. Previous work demonstrates ImPACT’s validity indices detect poor effort at disproportionately higher rates in athletes with histories of attention deficit hyperactivity disorder (ADHD) or academic difficulties. The present study examines whether previously suggested ‘dormant’ validity indices, Word Memory Correct Distractors (WMCD), and Design Memory Correct Distractors (DMCD), reduce disproportionate invalidity in collegiate athletes with such histories.
METHOD: Six seasons of ImPACT protocols were examined (n = 1727). Athletes were grouped by self-reported histories of ADHD, academic difficulties, or comorbid ADHD and academic difficulties. Chi-square analyses compared invalidity rates using existing validity indices and both standard and conservative cutoffs for WMCD and DMCD.
RESULTS: Using standard cutoffs for dormant indices (WMCD < 22, DMCD < 16) produced significant differences in rates of athletes identified as having an invalid protocol, with the comorbid group exhibiting the highest invalidity rate (63.2%) and the no history group producing the lowest (42.0%), χ2 (3) = 11.57, p < .01, Cramer’s V = 0.08. This difference remained when utilizing conservative cutoffs (WMCD < 18, DMCD < 10), with the comorbid group again producing the highest (26.3%) and the no history group producing the lowest (10.4%), χ2 (3) = 15.64, p < .005, Cramer’s V = 0.10.
CONCLUSIONS: Student-athletes with self-reported histories of ADHD and academic difficulties are more likely to produce invalid protocols, even with dormant indices. These findings emphasize the difficulty in assessing validity in special populations of athletes and encourage further work in this area.
PMID: 30516427 [PubMed – as supplied by publisher]