Abstract
Playing action video game can substantially improve visual performance. Here we consider the mechanisms that support such improvements. Using the equivalent external noise technique (Lu & Dosher, 1998), we found improved thresholds across a wide range of external noise levels (downward shifts of TvC curves) in Action Video-Game Players (VGPs) as compared to Non-Action Video-Game Players (NVGPs). Both the perceptual template model and a probabilistic neural model (Bejjanki et al, 2011) explain the improved TvC functions in terms of better task-relevant perceptual templates utilized by the VGP group. How does action video game experience improve the quality of the perceptual templates? One possibility is that action video game experience increases general visual sensitivity. If that is the case, the VGP advantage would be present on the very first trial of any visual task, and would not require any training with the task. Another possibility is that action video game experience may result in an increased ability to learn task-relevant statistics. This "learning to learn" view predicts reasonably equivalent performance between groups early on in training of a new visual task, with the VGP advantage appearing and then increasing through experience with the task. To test these hypotheses, VGPs and NVGPs performed a peripheral Gabor orientation identification task in high Gaussian noise. Each subject completed 8 blocks in an initial training phase and then 8 additional blocks in a transfer phase (where both the reference angle and location of the Gabor were changed from the training phase). As predicted by the "learning to learn" account, VGP and NVGP performance was comparable before training.However, as the training proceeded, VGP improved their performance much faster than NVGP, suggesting more efficient learning. Together, these results suggest that action video game play results in enhanced performance by fostering "learning to learn".
Meeting abstract presented at VSS 2012