The video is relevant to the article’s examination of cognitive flexibility, which is defined in wikipedia “as the mental ability to adjust thinking or attention in response to changing goals and/or environmental stimuli”, and real-time strategy games. Brian Glass (Queen Mary, University of London) and colleagues have an article published in PLoS ONE. Through Jamie Madigan, Brian Glass contacted me about posting his study in my blog and I asked him to write the blog post, which is great so I can have some extra time to write on other studies.
Training in action video games can increase the speed of perceptual processing. However, it is unknown whether video-game training can lead to broad-based changes in higher-level competencies such as cognitive flexibility, a core and neurally distributed component of cognition. To determine whether video gaming can enhance cognitive flexibility and, if so, why these changes occur, the current study compares two versions of a real-time strategy (RTS) game. Using a meta-analytic Bayes factor approach, we found that the gaming condition that emphasized maintenance and rapid switching between multiple information and action sources led to a large increase in cognitive flexibility as measured by a wide array of non-video gaming tasks. Theoretically, the results suggest that the distributed brain networks supporting cognitive flexibility can be tuned by engrossing video game experience that stresses maintenance and rapid manipulation of multiple information sources. Practically, these results suggest avenues for increasing cognitive function.
Their study was also reported in ScienceDaily and CNET. The article can be freely read at the journal website.The authors were aware that while observational studies of video game players versus non-video game players were an important early step in video game psychology research, the future lies with experimental manipulations. In other words, training regimens which involve within-game manipulations. Instead of just sending non-gamers off with a game, and then measuring them before and after, the researchers took an extra step of customizing the game for the research purposes.
While action games, such as FPS, have been shown in observational and training studies to be associated with increases in information processing speed and perceptual abilities, the authors knew that prior results have been mixed as to whether other forms of gaming can lead to higher level cognitive changes. Boot, Blakely, and Simons (2011) point this out, as well as the fact that methodological inconsistencies and inadequacies have led to serious drawbacks in the field. Specifically, they list the following problems:
- Overt recruiting (possible differential demand characteristics).
- Unspecified recruiting method.
- Potential third-variable / directionality problems(cross-sectional design).
- No test of perceived similarity of tasks and gaming experience.
- Possible differential placebo effects.
The study at hand mostly dodges these potentially problematic design issues. 1) Screening was done, as has to be done, in order to obtain novice gamers (less than 2 hours per week), especially when the large majority in the subject pool play more often than 2 hours per week. However, the advertisement that led to the screening was for a “long term study” of decision making, and not for a gaming study. Also, this is more of a problem for observational studies, where experienced gamers are sought. There was no overt reason for responders to under-report video gaming. 2) The study does report the recruiting method. 3) This study was a pre-test versus post-test training manipulation, so there was no cross-sectional or strictly observational component. 4) This is an interesting question, but if anyone has played StarCraft or The Sims, they will know that the experience is much different than tasks like ANT or Stroop. Check out this amazing video to see how challenging Real-Time Strategy gaming really is — and remember, the study used totally novice gamers with little or no gaming experience. 5) It’s hard to imagine how this could be possible. Although there were re-test effects, which was expected due to the fact that participants engaged in pre- and post-testing, the critical contrast was between one form of game and another form, so both groups completed the act of playing in a video game.
A major departure from other video game psychology studies was the customization of the StarCraft game play. The major manipulation was that two different complexity levels of StarCraft were tested — one with twice the map size and two starting bases (SC-2), and one with only one starting base, and half the map real estate (SC-1). Special software installed on the participant’s laptops controlled various aspects of the game. First, it controlled the amount of time that could be played. It would not let a new game start if 2 hours had already been played that day. Second, it switched between 2 different map styles — this kept the game interesting for the participant. Third, if the player won the previous game, the next game would be slightly harder (the opponent would start with more resources). If the player lost, the next game would be slightly easier. In this way, ability level was titrated to keep players at a win rate near 50%. In fact, the empirical win rates were 42.6%, SD = 8.8% for SC-1 and 43.0%, SD = 8.7% for SC-2. So, it’s not that those in SC-1 simply had it easier. Also, this kept the participants in lock-step as far as how demanding their StarCraft experience turned out to be.
The psychological testing took place before and after the 6 to 8 weeks of gaming. The battery of tests included a basket of cognitive flexibility related tasks (Attentional Network Test, Stroop, Ospan, Task Switching, and a novel Multi-Location Memory task), as well as tasks most likely unrelated to flexibility (visual search, information filtering, BART, and digit span).
The findings were such that those who played SC, especially the more complex SC-2, had enhancements specifically on the flexibility tasks when compared to those who played The Sims. This was tested using a meta-analytical approach. The approach was non-parametric, in that traditional null hypothesis testing was not used. Rather, a Bayes factor approach tested the null hypothesis against the alternative directly, resulting in an evidence ratio. In this experiment, the alternative hypothesis that SC had a positive effect of cognitive flexibility had a 6:1 evidence ratio, meaning it was 6 times likelier than the null hypothesis that SC led to no change. This is an interesting way to report the results, and more informative than a p-value.
The authors state that they could potentially build on these findings in a couple of ways. First of all, it’s important to know exactly what gaming mechanisms and what gaming behavior result in the cognitive change. To begin to address this, behavior was recorded during StarCraft game play using a customized DLL made from the open source BWAPI project. This allowed the authors to test the number of potential gaming features those in SC-1 and SC-2 were attending to, and they concluded that those in SC-2 were attending to more gaming features. The full data set is publically available at the first author’s website, http://www.brianglass.net. It is over 200GB and is available for download via FTP. The website also has other downloads and details about the game.
It could also be possible to design clinical interventions based on this game — there is some evidence that attentional training can be beneficial for children with ADHD, but the training regimens aren’t much fun. Paradigms like video gaming could offer more engaging training procedures for this sort of clinical population.
Glass, B. D., Maddox, W. T., & Love, B. C. (2013). Real-Time strategy game training: Emergence of a cognitive flexibility trait. PLoS ONE, 8 (8), e70350+. DOI: 10.1371/journal.pone.0070350