Psychometric Data for the NFL Neuropsychological Test Battery

Abstract

As part of a comprehensive league-wide study of concussion, the National Football League's Committee on Mild Traumatic Brain Injury sponsored a neuropsychological testing program from 1996 through 2001. Nearly 1,000 athletes participated voluntarily in the study. Traditional paper-and-pencil neuropsychological tests were used for baseline assessment. Neuropsychological tests used in the study included the Hopkins Verbal Learning Test-Revised, the Brief Visuospatial Memory Test-Revised, Trail-Making Tests (Parts A and B), and the Controlled Oral Word Association Test. We present a factor analysis of these baseline data on 513 athletes who completed all of the neuropsychological tests and present normative psychometric data on the multiple baseline tests employed.

Key words:

• assessment/diagnosis
• sports
• tests

ACKNOWLEDGEMENTS

Neither author has any financial interest in the neuropsychological tests used in this study.

Notes

*Sample means are significantly different, p < .05.

HVLT-R Total = Hopkins Verbal Learning Test-Revised Total = number words learned across three trials (maximum = 36).

HVLT-R Delayed = Hopkins Verbal Learning Test-Revised Delayed Recall (maximum = 12).

BVMT-R Total = Brief Visuospatial Memory Test-Revised = total number of points across three learning trials (maximum = 36).

BVMT-R Delayed = Brief Visuospatial Memory Test-Revised Delayed Recall (maximum = 12).

Trails A = time to completion in seconds.

Trails B = time to completion in seconds.

COWAT = Controlled Oral Word Association Test (total number of words generated across three trials).

Note. Negative t-score is due to the extreme skew of the asymmetric distribution.

Note. Negative t-scores are due to the extreme skew of the asymmetric distribution.

Note. Factor structure coefficients in boldface type denote the factor with which each item has its highest correlation.

Extraction method: principal axes factoring.

Rotation method: oblimin with Kaiser normalization.

¹The correlation between Factor 3 to the remaining factors is negative because the scaling of the factor is keyed to TMT-A and TMT-B variables, which are time-to-completion variables. A lower score is a diagnostically desirable outcome and a high score is an undesirable outcome. Hence, the correlations between the TMT variables to the factors and to the other variables in the set of 11 are negative.