Equivalence of Alternate Forms of a Computerized Neuropsychological Test

Journal of Athletic Training. -


Resch, J. E., S. N. Macciocchi and M. S. Ferrara.



Context: Computerized neuropsychological (CNP) testing is widely utilized for sport-related concussion management. The most widely used CNP is the ImPACT. There have been several studies that have evaluated the reliability and sensitivity of the ImPACT but virtually no studies that have evaluated equivalence of ImPACT’s alternate forms. Form equivalence is important to determine clinical utility of CNP testing for management of concussion. Objective: To determine alternate form equivalence of the ImPACT. Design: Repeated Measures. Setting: Research laboratory. Patients or Other Participants: Participants consisted of 108 (N=108), healthy, college participants aged 20.6 + 1.52 years, 171.6 + 9.71 cm tall, and mean mass of 66.9 + 11.92 kg. Interventions: Participants were divided into 6 equal groups and were administered alternate forms of ImPACT using clinically relevant time points (day 1, day 45, and day 50) mimicking a typical concussion recovery pattern. All participants completed a baseline evaluation using form 1 followed by a serial combination of the remaining alternate forms at day 45 and day 50. Participants were also administered Green’s Word Memory Test (WMT) as a measure of effort. Main Outcome Measures: ImPACT calculates five composite scores: composite verbal and visual memory, visual motor speed, and reaction time. Inferential Confidence Intervals (ICI)s were calculated to compare ImPACT’s alternate forms and their respective composite scores. Pearson moment correlations and independent t-tests were also utilized to further analyze alternate form equivalence. Repeated Measures ANOVA was utilized to determine changes in effort across time for Green’s WMT. All analyses were performed with α = .05. Results: ICIs revealed non-equivalence of composite verbal memory between form 1 and forms 2, 3, and 4. Non-equivalence was also observed between form 1 and form 3 for composite visual memory and visual motor speed. Forms 2 and 4 were non-equivalent for composite verbal memory. Forms 3 and 4 were nonequivalent for composite reaction time. Pearson r’s ranged from -.263 and .712 for non-equivalent forms. All participants scored greater than 85% percent on immediate and delayed recall, consistency, and paired associates of the Green’s WMT suggesting good effort was provided at all time points. Conclusions: Alternate forms of neuropsychological tests are utilized to minimize practice effects over repeated assessments. Alternate forms of ImPACT were not equivalent for one or more composite scores. Clinicians should consider psychometric properties when interpreting results of neuropsychological testing.

Links to full article: