Test-retest, retest and retest: Growth curve models of repeat testing with ImPACT

• Published in: Journal of clinical and experimental neuropsychology 2016-06, Vol.38 (8), p.869-874
• Main Authors: Maerlender, A.C., Masterson, C.J., James, T.D., Beckwith, J., Brolinson, P.G., Crisco, J., Duma, S., Flashman, L., Greenwald, R., Rowson, S., Wilcox, B.
• Format: Article
• Language: English
• ISSN: 1380-3395; 1744-411X
• DOI: 10.1080/13803395.2016.1168781

Computerized neuropsychological testing has become an important tool in the identification and management of sports-related concussions; however, the psychometric effect of repeat testing has not been studied extensively beyond test-retest statistics. The current study analyzed data from Division I collegiate athletes who completed ImPACT baseline assessments at four sequential time points that varied over the course of their athletic careers. Administrations were part of a larger NIH study. Growth curve modeling showed that the two memory composite scores increased significantly with successive administrations: change in Verbal Memory was best represented with a quadratic model, while a linear model best fit Visual Memory. Visual Motor Speed and Reaction Time composites showed no significant linear or quadratic growth. The results demonstrate the effect of repeated test administrations for memory composite scores, while speed composites were not significantly impacted by repeat testing. Acceptable test-retest reliability was demonstrated for all four composites as well.