Video Game Influences

This newsletter is prompted by two new studies on the impact of video games.

These studies are intriguing – they both seem to show that gaming improves cognitive functioning. They also illustrate how good statistics can't save bad logic.

Background

Let's start with a simple question: Is playing video games good for you? Well, it depends. For instance, it depends on what you mean by "good". Numerous studies show that gaming improves attention and spatial reasoning, but a common theme in this literature is that the benefits don't transfer much. In other words, playing video games makes you good at...playing video games. Only a few studies show benefits that extend to other domains.

(A gamer might point out that "other domains" are becoming increasingly game-like. Three examples that come to mind are instructional technology, health aids like Fitbit, and financial services apps like Robinhood. In other words, as time goes on, playing video games might yield increasingly transferrable skills.)

Whether gaming is good for you also depends on who you are. Gaming can promote coordination and balance among people with medical conditions ranging from cerebral palsy to stroke. Elderly people who are taught to play video games for the first time show improvements in multitasking skill. Children with autism spectrum disorder, dyslexia, or ADHD can benefit from games tailored to their particular challenges. (Since 2020, the FDA actually permits the game EndeavorRX to be prescribed as an ADHD treatment, and both pre- and post-approval trials suggest that it's effective.)

As that last example indicates, whether gaming is good for you also depends on the type of game you play. Surgeons can acquire manual dexterity from gaming, but only games that promote manual skill. Gaming can help children with autism spectrum disorder engage comfortably in learning and socialization, but only certain games help, and only if precautions are taken against over-engagement in the games.

There's also evidence that violent video games spur more aggressive behavior, as well as desensitizing players to violence, which is to say that no matter who you are, some games are harmful. (The way news media treat this topic reminds me of the way they used to treat the link between human activity and climate change. Most experts agree on a link between violent video games and aggressive behavior, but a small, extremely vocal minority disagree – and get enough media attention that the findings are popularly considered "mixed".)

Finally, whether gaming is good for you depends on how much you play.  Too much play has been linked to physical problems like carpal tunnel syndrome, gamer's thumb (inflammation of the tendons that move the thumb), eye strain, and obesity, as well as psychological problems such as negative mood and social withdrawal. The Diagnostic Statistical Manual (DSM-5) now lists Internet Gaming Disorder among "conditions for further study", meaning that in a few years, IGD may be officially considered a mental illness. Meanwhile, one study estimates that between 0.3% and 1% of adults experience this disorder.

The two studies I'll be discussing collapse most of the distinctions I described here. Each study claims to show that most or all kinds of games benefit the cognitive functioning of just about all players.

Study 1: Video game play and decision-making

In the September 2022 issue of Neuroimage: Reports, now available online, two researchers from Georgia State University's Department of Physics and Astronomy present a study on the cognitive effects of video game play. The title of the study nicely summarizes the main findings: "Video game players have improved decision-making abilities and enhanced brain activities."

This study compared two groups of undergraduates. Video game players (VGPs) consisted of 28 students who reported playing video games at least 5 hours per week over the past two years. Non-video game players (non-VGPs) consisted of 19 students who reported less than 5 hours per week of play. No other information was provided about the groups, except that the non-VGPs averaged less than an hour per week of play, which suggests a reasonably clear distinction between the groups.

I'm not a gamer, but it's critical for interpreting the findings that the VGP played games that require constant attention to rapidly shifting detail. (Specifically, they were involved in first-person shooter, real-time strategy, multiplayer online battle arena, and/or battle royale genres. It's interesting that three of the four genres involve violence...)

Participants performed multiple versions of the same task. Each participant sat a computer screen and examined two sets of 600 rapidly moving dots. The two sets of dots, which were identical except for color, overlapped while moving in different directions relative to each other. The instructions were to press a button, as quickly as possible, to indicate whether one set of dots (e.g., the red ones) were moving to the left or the right. Response times were less than a second.

I'm not sure how clearly I've described the task, but only a rough sense of it is needed. Participants were asked to quickly judge the direction of some moving dots while ignoring some other moving dots that may have been distracting. (Here's a video of a similar task involving fewer dots and a higher rate of speed.)

The results were straightforward: The VGP group was faster and more accurate, on average, than the non-VGP group on all versions of the task. Overall, the average difference was 190 milliseconds (just under a fifth of a second). The two groups also differed on a variety of fMRI measures; for example, the VGP group showed greater neural activity in several areas of the brain involved in visuospatial processing.

I found no particular problems with the statistical approach. However, there are two enormous, honking flaws in the way the statistics were interpreted.

Study 1 flaws

1. The study purports to look at the impact of video game play on "decision-making". That's the phrase used in the title, the abstract, and everywhere else in the article. But, as you can see from the task, this a very, very specific sort of decision-making. It's not: Which college should I attend? It's not even: Should I get more coffee now? It's just identifying the directions of moving dots.

Ordinarily, I would call this a "limitation" rather than a "flaw". That is, I would say that given the task the researchers used, their results might not generalize to a lot of things that we call decision-making. I use the term "flaw" instead, because the task tests the very skills that get practiced every moment one plays the kinds of video games that the VGPs play. This is an extreme case of the transfer problem I mentioned at the outset.

Why is this a flaw? Well, imagine studying two groups of adults. Group A has been taking tennis lessons for several years, while Group B is sedentary and rarely engages in any sport. You now ask each person to stand in one place, watch tennis balls being launched at them, then judge the trajectory of each ball as quickly as possible. Their instructions are to say "forehand" if the ball is moving toward their forehand side, and "backhand" if it's moving toward their backhand side.

Not surprisingly, you'd find that Group A is slightly quicker, on average, at performing this task. Would you conclude that tennis lessons improve "decision-making skills"?

Of course not. You would only conclude that tennis lessons improve tennis skills. Or, more specifically, that tennis lessons improve the ability to discern the trajectory of small round objects as they're approaching you. We didn't need a study to tell us that.

In short, the first major flaw of the study is that it didn't test the effects of video game play on anything but video game skills.

One might argue that I'm wrong about lack of transfer. If you're faster at discerning the direction of moving objects, you've acquired a skill with very broad application, since pilots, hunters, athletes, and others make use of this kind of skill. There's even Dr. Elizabeth Bik, who's famous for spotting forged or inappropriate duplications of microscopy images in biomedical research, an activity that requires, in a very literal sense, visual inspection of lots and lots of dots.

I'm not opposed to this interpretation, but it's not what the study shows. There's only one task. Participants didn't have the opportunity to demonstrate transfer to anything other than computer-generated displays of moving dots.

2. Throughout the article, the researchers use the language of causality, arguing that video game play "affects" or "causes improvement in" decision-making. In other words, if you take up gaming, you'll end up being a better decision-maker (at least as far as the movement of dots go).

However, young people who are better at this kind of decision-making may have been more likely to choose video games as a leisure activity in the first place. In other words, if you're the kind of person who can look at a computer screen and quickly discern patterns of moving dots, you might be the kind of person who enjoys video games.

In short, it's possible that gaming didn't improve the decision-making ability of the VGPs. They may have already been better to begin with.

Study 2: Screen time and intelligence

In May, Dr. Bruno Sauce and a team of researchers from the Netherlands, Germany, and Sweden published a study in Scientific Reports looking at associations between screen time and intelligence. Along with video game play, they also considered two other types of screen time: socialization (via social media) and watching (TV or online videos).

Participants consisted of 9,855 American 9 and 10 year olds from a huge, multi-site longitudinal project called The Adolescent Brain and Cognitive Development (ABCD) Study. Children described the amount of screen time they devote to each of the three categories (video games, socialization, and watching) on a typical weekday and a typical weekend day using the following rating scale: none, less than 30 minutes, 30 minutes, 1 hour, 2 hour, 3 hours, 4 or more hours. Scores were also available for five intelligence-related tasks (vocabulary comprehension, reading fluency, short-term memory, selective attention, and spatial reasoning.)

The main finding was that after controlling for genetic and socioeconomic variables, video game play and watching were each associated with greater gains in intelligence. Specifically, children at ages 9 and 10 who played more video games or watched more TV/online video showed greater increases in intelligence task scores two years later. No effects were seen for socialization.

Study 2 flaw

This study has many limitations. For example, children may not have been accurate in describing the amount of screen time they pursued, and the researchers didn't measure many dimensions of intelligence. I would call these limitations rather than flaws, because there's no evidence that they systematically biased the results. For example, all sorts of mistakes are possible when describing screen time; these mistakes could balance out and have no impact on the association between screen time and intelligence, or they could cause the strength of that association to be over- or under-estimated.

The main flaw of this study stems from claims about causality. The researchers repeatedly framed their results as "evidence of a beneficial causal effect of video games on cognition." In brief, they claimed that gaming accelerates the growth of intelligence.

The problem with this conclusion is the same as one we saw in the preceding study: We don't know whether some other variable is responsible for the gaming-intelligence association, and we don't know whether smarter kids played more video games in the first place.

The researchers relied on three strategies to address these possibilities. First, they controlled for pre-existing differences in socioeconomic status and cognition. This is a good start, but it doesn't solve the problem, because many other variables could account for the findings. Here's just one example: More tech-savvy parents tend to make more video game options available to their children, and they may also do other things that promote intelligence (like engaging in more intellectually stimulating conversation about video game technology and other topics). If so, it's not gaming that boosts intelligence; rather, parental influence would be responsible for promoting both.

Second, the researchers looked at changes in intelligence rather than absolute scores. This does absolutely nothing. If there's a third variable at play (like the tech-savviness of parents), it doesn't matter whether you look at absolute scores or change scores, because that third variable continues to exert an influence.

Finally, the researchers controlled for genetic differences in cognition. They identified genetic markers previously associated with cognitive performance, and then controlled for differences between children in those markers, so that the effects of gaming would not be attributable to them. Unfortunately, the way the researchers went about this was a hot mess. (a) The genetic markers were drawn from a separate database consisting primarily of genomic data obtained from white Europeans, then applied to the ABCD dataset, which is representative of the U.S. in terms of race and ethnicity. (b) The European dataset only suggests broad associations between genetic markers and cognition. At the level of the individual child, you can’t say that because a child has such-and-such genetic makeup, they'll have a particular level of cognitive functioning. (c) What those genetic markers were associated with in the European dataset is different from what was measured in the current study. The European dataset looked at a combination of cognitive performance, self-reported attainment, and self-reported ability, which is very different from performance on the five intelligence tasks in the current study. In short, controlling for genetics in this study is almost like controlling for interest in chocolate pudding. There's not much point.

In all, there's no clear evidence from this study that gaming boosts intelligence.

Conclusion

Is gaming good for you? As I mentioned earlier, evidence suggests that it can be. The two new studies discussed here don't add to that evidence. Rather, they illustrate some of the most prevalent limitations in the literature. Gaming may have cognitive benefits, but (a) it's not clear whether these benefits extend beyond gaming contexts, and (b) it's not clear whether gaming per se is responsible for those benefits.

That second point is critical. There are a lot of correlational studies I haven't mentioned that link gaming to other benefits, like problem-solving skills, grades, creativity, social functioning, etc., but in study after study, it's just not clear that gaming is responsible. Often, the problem is a failure to control key variables. Kids who don't play video games may often be the kids whose parents can't afford to buy games, maintain a reliable internet connection, or provide intellectually stimulating activities.

In my opinion, the main advantages of gaming come from creating the best possible fit between games and individuals.

As a parent of a small child, that might mean steering your child toward the PBS Kids games page (free and lightly educational) and away from Grand Theft Auto (expensive, violent, and sexist).

As a health care professional, that might mean introducing your cancer patient to one of the Re-Mission games, which has been repeatedly shown to enhance mood and increase treatment compliance, or your diabetes patient to mySugr, an app that gamifies diabetes management.

As an educator, that might mean advocating for more games in the regular classroom, working in separate game-oriented spaces such as the esports room at Apollo Junior High in Richardson Texas, or spreading the word about cell phone apps that promote adult literacy. 

These are just a few of a seemingly inexhaustible supply of possibilities. Grand Theft Auto's not the only game in town…!