LAUSR

ImPACT is a popular tool used to assess neurocognitive function after concussion. Age- and gender-based percentile ranks are provided as part of the ImPACT clinical report, but only raw composite scores are provided for research. Percentile ranks may provide an alternate look at gender differences on neurocognitive testing. We sought to compare the use of percentile ranks versus composite scores when comparing neurocognitive performance between male and females. Pre-season baseline data from 3,238 collegiate athletes were exported from a larger multi-site database, 1,714 male (53%) and 1,524 female (47%), ages 18–22 (Mean = 19.3, SD = 1.2). ImPACT composite scores were extracted (Verbal Memory, Visual Memory, Visual Motor Speed, Reaction Time) and Percentile Ranks were calculated using age- and gender-based reference tables in the ImPACT 2017–2012 Technical Manual. One-way analyses of variance (ANOVA) were conducted with gender as the IV and the four composite scores and their associated percentile ranks as the DVs. Bonferroni correction for 8 comparisons set the alpha level to p < .006. ANOVAs revealed a significant effect of gender on ImPACT Verbal Memory [F(1,3236) = 16.32, p < .001], Visual Memory [F(1,3236) = 66.06, p < .001], Visual Motor Speed [F(1,3236) = 15.61,p < .001] but not Reaction Time [F(1,3236) = 2.59, p = .11]. When using percentile ranks, ANOVAs revealed a significant effect of gender on only Visual Motor Speed [F(1,3236) = 31.07, p < .001], but not Verbal Memory [F(1,3236) = 2.43, p = .12], Visual Memory [F(1,3236) = 0.48, p = .49], or Reaction Time [F(1,3236) = 0.61, p = .44]. Use of ImPACT Composite Scores in concussion research may exaggerate or inappropriately promote gender differences on computerized neurocognitive testing. Given that age- and gender-based normative data are available, use of percentile ranks is recommended.