Differences in characteristics of two aspects of procedural learning in action video game players

Action video game experience has been widely associated with differences in cognitive functions. However, it remains unclear whether such experience is differentially associated with distinct components of procedural learning, including statistical learning (sensitivity to probabilistic regularities) and higher-order sequence learning (extraction of serial order structures). This study aimed to examine descriptive patterns of these components using the Alternating Serial Reaction Time (ASRT) task.

A total of 142 undergraduate students were recruited and categorized into action video game (AVG; n = 72) and non-video game (NVG; n = 70) group based on their self-reported gaming history. The ASRT task was used to evaluate procedural learning. Reaction times to three types of stimulus triplets were recorded across four epochs to assess general skill learning, statistical learning, and sequence learning. Mixed-design analysis of variance (ANOVAs) were conducted to examine learning effects and group differences.

AVG group exhibited significantly faster reaction times across all epochs, reflecting a descriptive pattern of faster general response speed. NVG group demonstrated a significant improvement in sensitivity to statistical regularities over time, whereas AVG group showed consistently lower levels of statistical learning. Conversely, AVG group showed reduced inverse learning effects in high-order sequence learning, suggesting better extraction of sequential structures.

AVG experience was associated with faster motor responses and better sequence extraction ability, whereas NVG participants showed greater sensitivity to probabilistic regularities. These descriptive patterns are consistent with a multifactorial view of procedural learning and highlight potential interactions between statistical and sequence learning components.