학술논문

Reliability and Minimal Detectable Change of the Standardized Assessment of Reaction Time.
Document Type
Article
Source
Journal of Athletic Training (Allen Press); Jun2023, Vol. 58 Issue 6, p579-587, 9p, 3 Charts, 1 Graph
Subject
Research methodology evaluation
Research methodology
Comparative studies
Longitudinal method
Sports participation
Statistical power analysis
Analysis of variance
Statistical reliability
Confidence intervals
Sports
Psychometrics
Pearson correlation (Statistics)
Brain concussion
Questionnaires
Intraclass correlation
Reaction time
Statistical sampling
Language
ISSN
10626050
Abstract
Postconcussion reaction time deficits are common, but existing assessments lack sport-related applicability. We developed the Standardized Assessment of Reaction Time (StART) tool to emulate the simultaneous cognitive and motor function demands in sport, but its reliability is unestablished. To determine the intrarater, interrater, and test-retest reliability of StART and to examine the dual-task effect, time effect, and relationships between StART and computerized and laboratory-based functional reaction time assessments. Prospective cohort study. Clinical laboratory. Twenty healthy, physically active individuals (age = 20.3 ± 1.8 years, females = 12, no concussion history = 75%). Participants completed the StART, computerized reaction time (Stroop task via CNS Vital Signs), and laboratory-based jump landing and cutting reaction time under single-task and dual-task (subtracting by 6s or 7s) cognitive conditions at 2 testing sessions a median of 7 days apart. We used intraclass correlation coefficients (ICCs), repeated-measure analysis of variance, and Pearson r correlations to address our aims. Overall, good to strong interrater (ICC [2,k] range = 0.83–0.97), intrarater (ICC [3,k] range = 0.91–0.98), and test-retest (ICC [3,k] range = 0.69–0.89) reliability were observed. A significant reaction time assessment-by-cognitive condition interaction was present (P =.018, ηp2 = 0.14), with StART having the largest dual-task effect. Main time effects for dual-task conditions were seen across all reaction time assessments (mean difference = −25 milliseconds, P =.026, ηp2 = 0.08) with improved performance at the second testing session. No StART outcomes correlated with computerized reaction time (P >.05), although some correlated with single-task (r range = 0.42–0.65) and dual-task (r range = 0.19–0.50) laboratory cutting reaction time. The StART demonstrated overall reliable performance relative to other reaction time measures. Reliability coupled with a strong dual-task effect indicates that StART is a valid measure for examining functional reaction time and may have future utility for sport-related concussion return-to-play decision-making. [ABSTRACT FROM AUTHOR]