Development of a Virtual-Reality Eye Movements-Based System to Assess Basketball Players’ Decision Making

Abstract

Background and Purpose: This study developed a reliable and ecologically valid virtual reality eye movements-based assessment system to evaluate basketball players’ decision-making abilities.

Research Design: The system incorporated expert ratings, inter-group differences, analysis of covariance, and test-retest reliability assessments to validate its effectiveness and reliability.

Study Sample: A VR system with 100 task scenarios was used to assess decision-making performance and visual behavior. 30 high-level and 30 low-level players participated in two phases.

Data Collection and Analysis: In Phase 1, a panel of basketball experts (N = 3) rated the decision-making scenarios. Kendall’s coefficient of concordance (W) was used to analyze expert ratings, confirming content validity. Mann-Whitney U and independent samples t-tests were employed to assess decision quality and decision time differences between high- and low-level groups. Gender was included as a covariate in ANCOVA to control for gender effects. Gaze patterns were analyzed to examine differences in visual behavior. Phase 2 involved test-retest reliability analysis using Pearson’s correlation coefficient (r).

Results: High-level players performed significantly better in decision-making (p < 0.001) and had broader gaze distributions, while low-level players focused less on critical information. Test-retest correlations for decision scores (r = 0.846) and reaction times (r = 0.802) were significant (p < 0.001). Therefore, the high test-retest correlation reflects the strong reliability of the assessment system.

Conclusions: This study is the first to validate the reliability and validity of a VR eye movements -based assessment tool for evaluating basketball players’ decision-making abilities. By integrating eye-tracking technology with VR, this tool enables more accurate and reliable evaluations of basketball players’ decision-making skills in future research.