Increasing Speed of Processing With Action Video Games

Abstract

In many everyday situations, speed is of the essence. However, fast decisions typically mean more mistakes. To this day, it remains unknown whether reaction times can be reduced with appropriate training, within one individual, across a range of tasks, and without compromising accuracy. Here we review evidence that the very act of playing action video games significantly reduces reaction times without sacrificing accuracy. Critically, this increase in speed is observed across various tasks beyond game situations. Video gaming may therefore provide an efficient training regimen to induce a general speeding of perceptual reaction times without decreases in accuracy of performance.

Figures

Fig. 1

A Brinley plot showing the reaction time (RT) of non-video-game players (NVGPs) on the X-axis versus that of expert video-game players (VGPs) on the Y-axis, for 89 different experimental conditions from nine different types of tasks. For each experimental condition, the RTs of VGPs and NVGPs were retrieved and plotted as one separate data point. A simple linear function (y = mx) was used to describe the relationship between VGP and NVGP RTs (dashed line). VGPs responded 11% faster than NVGPs across a wide range of RTs. Importantly, similar accuracy was observed across groups, ruling out an explanation in terms of simple speed-accuracy tradeoff (VGP accuracy = 0.99 × NVGP accuracy, R2 = 0.92). The studies are (a) Greenfield, deWinstanley, P., Kilpatrick, H., & Kaye, D. (1994), (b) Castel, Pratt, J., & Drummond, E. (2005), (c) Bialystok (2006), (d) Dye, Green, & Bavelier (2009), (e) Green and Bavelier (2003), (f & g) Bavelier & Bailey (unpublished data).

Fig. 2

Brinley plots comparing pretest and posttest reaction times (RTs) for action-game trainees (A) and control-game trainees (B) for four tasks: motion discrimination (based upon Palmer, Huk, & Shadlen, 2005), task switching (based upon Monsell, Sumner, & Waters, 2003), visual search for letters (based upon Castel et al., 2005), and visual search for Gabor patches (based upon Cameron, Tai, Eckstein, & Carrasco, 2004). For both action- and control-game trainees (7 males and 7 females in the action group and 7 females and 4 males in the control group), training consisted of playing randomly assigned videogames for 50 total hours over a period of 8 to 9 weeks. Members of the control group played the game The Sims™2 (Electronic Arts Inc.); members of the experimental group played the game Unreal® Tournament 2004 (Epic Games) followed by the game Call of Duty® 2 (ActiVision). The action-trained group demonstrated a 13% decrease in their RTs, whereas the control-trained group exhibited only a 6% decrease (from Green, 2008). Importantly, changes in accuracy for both groups were negligible, with the action game group showing a 0.3% decrease in accuracy (posttest accuracy = 0.997 × pretest accuracy, R2 = 0.96) and the control group a 0.6% decrease (posttest accuracy = 0.994 × pretest accuracy, R2 = 0.95) ruling out any explanation of the RT changes in terms of speed accuracy tradeoff.

Fig. 3

The Test of Variables of Attention (A), used to assess differences in impulsivity and sustained attention between non-video-game players (NVGPs) and expert video-game players (VGPs), and results for both reaction time (B) and accuracy (C) measures. VGPs were faster at responding than NVGPs on both the impulsivity and sustained attention measures, but the groups did not differ on the accuracy measure, suggesting that the faster responses of VGPs were not due to impulsive responses to the stimuli and that they did not have greater problems sustaining their attention. (n.s. = nonsignificant.)