Immersive virtual games: winners for deep cognitive assessment
Journal:
arXiv
Published Date:
Feb 14, 2025
Abstract
Studies of human cognition often rely on brief, controlled tasks emphasizing
group-level effects but poorly capturing individual variability. A suite of
minigames on the novel PixelDOPA platform was designed to overcome these
limitations by embedding classic cognitive tasks in a 3D virtual environment
with continuous behavior logging. Four minigames explore constructs overlapping
NIH Toolbox tasks: processing speed, rule shifting, inhibitory control, and
working memory. In a clinical sample of 60 participants outside a controlled
lab setting, large correlations (r=0.42-0.93) were found between PixelDOPA
tasks and NIH Toolbox counterparts, despite differences in stimuli and task
structures. Process-informed metrics (e.g., gaze-based response times) improved
task convergence and data quality. Test-retest analyses showed high reliability
(ICC=0.52-0.83) for all minigames. Beyond endpoint metrics, movement and gaze
trajectories revealed stable, idiosyncratic gameplay strategy profiles, with
unsupervised clustering differentiating participants by navigational and
viewing behaviors. These trajectory-based features showed lower within-person
variability than between-person variability, facilitating participant
identification across sessions. Game-based tasks can therefore retain
psychometric rigor of standard cognitive assessments while providing insights
into dynamic individual-specific behaviors. By using an engaging, customizable
game engine, comprehensive behavioral tracking can boost power to detect
individual differences without sacrificing group-level inference. This
possibility reveals a path toward cognitive measures that are both robust and
ecologically valid, even in less-than-ideal data collection settings.
