An electrophysiological study of the effects of visuospatial training on mental rotation, object identification, and enumeration performance

Video game naive participants were trained on a first-person shooter action video game (AVG) for 30 hours or received no training, then were tested on three visuospatial tasks while electroencephalography was recorded. Accuracy was significantly better in AVG-trained than non-trained participants for naming objects rotated in-depth and enumerating 2-4 targets (i.e., subitizing). Major group differences were evident at anterior sites during an N3 complex (∼300--500 ms) indexing the process of matching perceived information to stored structural knowledge (i.e., object model selection). Compared to non-trained participants, electrophysiological analyses indicated that AVG-trained participants utilized object-specific and rotation imagery processes efficiently during mental rotation (Experiment 1), view-dependent model verification processes more consistently enhancing object constancy (Experiments 2a and 2b), and early beneficial pattern matching processes facilitating subitization (Experiment 3). Overall, goal-specific percept-to-memory matching was better in the AVG-trained group likely contributing to their superior accuracy.