Role-based task modes modulate L1 and L2 subtitle reading in video gameplay: An eye-tracking study.

Abstract

Understanding reading in video games can help harness their potential for reading training and learning. The current eye-tracking study examined subtitle reading in two role-based task modes (Player vs. Viewer) among first-language (L1) and second-language (L2) readers. Ninety-two participants read subtitles while alternating between controlling the game (Player) and watching noninteractive game sequences (Viewer) within the same gaming session. To assess the robustness of the findings, simpler and more complex statistical models were compared after being applied to the Holm-Bonferroni method. Across both model sets, the Viewer mode consistently showed more fixations, a greater proportion of reading time, longer mean fixation durations, lower subtitle-skipping rates, and shorter progressive saccades than the Player mode. The results suggested systematic differences in subtitle reading between the two role-based task modes. Differences in language background were consistently observed in saccadic behavior, with L2 readers producing shorter progressive and regressive saccades than L1 readers. Yet the interaction effects in the simpler models did not remain significant in the complex models, indicating substantial individual variation might affect the estimated effects on subtitle reading. In sum, subtitle reading during gameplay was modestly and systematically influenced by role-based task modes and language background, although evidence for their interaction remains inconclusive. These findings may inform the presentation of textual information in educational and training contexts.