Abstract
Purpose:
Action video game playing has been shown to modify several parameters of cognitive processing. One key question is how changes that are often measured with abstracted paradigms relate to visual behavior outside the laboratory. We investigated the effect of action game training on visual search in abstract, naturalistic (counting the number of people in city street photos) and game-related search tasks, to examine how changes extend beyond game-relevant tasks.
Methods:
40 non-gamers (20 female, 17-35 years old) were trained for 10 hours on an action game (first person shooter) or a computer-based card game (control group). Eye movements were recorded pre- and post-training in 5 different visual search tasks: conjunctive search for a singleton, counting people in city photos, counting soldiers in action game photos and finding a gabor patch embedded in the city and action game photos. Repeated measures ANOVAs were conducted for each task with the game type (action or card) as between and session (pre or post) as within-group independent variables, and saccade amplitude, fixation duration, distribution of fixations and search time as dependent variables.
Results:
There was no main effect of session, game type and no interaction for saccade amplitude or fixation duration in any of the tasks; however, there was a main effect of game type (F1, 73 = 4.86, p=0.046) in the vertical distribution of fixations, but only in the soldier counting task. Action gamers showed a significant reduction in their vertical distribution of fixations after training, but there was no main effect of session (p=0.16) and no interaction (p=0.1). Repeated measures ANOVA on search time showed a significant main effect of session (F1, 74 =4.43, p=0.03) and game type (F1, 74 = 6.152, p=0.01) of finding the gabor in game and city photos respectively.
Conclusions:
After 10 hours of action gaming, participants learned to search game images differently, with the vertical distribution of fixations approaching that in the natural scenes. However, there was no effect of training on searching the naturalistic scenes. Faster detection of gabor patches in game and city scenes can be attributed to a test-retest effect rather than training effect, as it happened in both groups. A crucial direction for future research is to better understand how skills enhanced in gaming can also be improved in natural settings.