video games

The Problem With Studying ‘Deviant’ Video Games

Photo: Rockstar Games

The debate over the connection between playing video games and real-world behavior has subsided a bit since the days of its Columbine-era peak cacophony. But it’s very much an ongoing controversy, and one with important ramifications given the sheer popularity of video games. A new study purporting to show a long-term link between risk-glorifying game play and various deviant behaviors highlights just how tricky a subject this is — and psychologists’ ongoing internecine battle over this issue.

For the study, published in Journal of Personality and Social Psychology, a team led by Dartmouth psychologist Jay Hull tracked a group of teens over four years, gathering data via telephone interviews on whether and how often they played three “mature-rated, risk-glorifying” (MRRG) games — Spider-man 2, Manhunt, or GTA III — the extent to which they engaged in risky and delinquent behaviors, and how they evaluated their parents’ parenting style (among other questions). Overall, the researchers found that playing the games was “associated with increases in alcohol use, cigarette smoking, aggression, delinquency, and risky sex,” as a result of an “increased tolerance of deviance.”

The point of the study is that the breadth of the effects on generally deviant behavior I think [are] kind of impressive,” Hull said. This isn’t primarily about violence, though. “The aggression is not the strongest effect,” he said. “You get much stronger effects for these other kinds of devious activities, particularly drinking.”

Our theoretical account is that what you’re doing is you’re playing a character, you’re essentially being a person with bad values, bottom line,” Hull said. “You’re playing somebody who’s morally bankrupt. And characters matter, and motives matter.” Hull is careful to draw a contrast between past attempts to link violence and video games — which have, he says, mostly been about the idea of players being rewarded for “practicing” violence in a virtual setting and then carrying that behavior into the real world — from his and his colleagues’ own, more character-based approach. In his view, players are, in a sense, partially inhabiting the identities of the characters they’re playing as. “We argue that you are practicing being another kind of person when you are playing these games,” he said.

So it’s less about the actions than the characters, which is why Hull said there might be different effects in games like those in the Medal of Honor series, in which characters are fighting for (ostensibly) noble reasons. “If you think about it from our perspective, if you’re like a sergeant, storming a beach, leading a band of brothers into that, what are your moral values? You’re defending democracy, right? You may be acting in a violent way, but for a good cause,” he said.

He cited his study’s specific findings about the effects of Spider-man 2, which, as he put it, “don’t predict much of anything, especially with respect to these kinds of deviant behaviors. Because I really do think that it matters: Why are you doing what you’re doing in the game?” In other words, Spider-man is a good guy, so he promotes less deviance.

The theory does appear to break down a bit here, though — at least slightly. I pointed out to Hull that he and his team observed effectively the same (small but statistically significant) correlation between playing Spider-man 2 and delinquent behavior as they did between playing GTA III and engaging in delinquent behavior. That doesn’t make sense — Spider-man is very much a righteous character; Claude from GTA III isn’t.

I can’t account for why that’s as similar,” he said. Playing as Spider-man was also correlated, in modest but statistically significant ways, with smoking and having sex without a condom. It’s hard to come up with explanations for this that can be tied to Spidey’s character or morality.

Because again: Hull and his team aren’t connecting their findings to risky or deviant behavior being rewarded in MRRG games — which would at least give them a coherent explanation, because maybe all that web-slinging simply teaches kids to be more reckless, or something — but rather are going with the stronger claim that playing these games, as they put it in the paper, “would appear to result in personality development and attitude change consistent with the values of characters enacted in such games and indicative of a growing tolerance for deviant others and generally deviant behavior.” Do Spider-man’s values foster a culture of drinking and bareback sex? More broadly, if players are so affected by the characters they’re playing, why isn’t there more of a link to aggression among those who play games like Manhunt or GTA III?

Setting aside the question of whether video-game players really extract much from the “values” of the characters they’re controlling, there’s also the tricky question of just how meaningful the effects uncovered by Hull and his team are.

Christopher Ferguson, a professor at Stetson University and critic of many studies linking video games to negative behavior — he’s giving a talk to the American Psychological Association on Friday that’s part of his long-running effort (explained in this article he wrote for American Psychologist) to shift the organization’s stance on this issue — didn’t have any major methodological problems with the study, but argued that Hull and his team simply didn’t show much of a link between playing MRRG games and engaging in risky behavior.

When we get to findings that are this small, they’re what I would call fragile,” said Ferguson. They could be explained by other factors, in other words: They could be artifacts of certain kinds of kids not answering certain kinds of questions truthfully or accurately, for instance (since all the data on the kids’ habits were self-reported), or of other so-called omitted variables.

Hull, for his part, pointed to data his team had collected showing that among those adolescents who switched from having never binge-drank to having binge-drank between each wave of data collection, the vast majority were heavy gamers. Ferguson replied that this measure can’t tell us much because it lacks proper controls. (Note: Due to an error on my part, an earlier version of the paragraph incorrectly stated that Ferguson criticized a table from the study for lacking controls, rather than the data showing differences in rates of binge-drinking adoption. Ferguson’s lack-of-controls criticism was leveled at the latter, not the former.)

It’s easy to get lost in this wonky labyrinth, in this dissection of numbers from just one video-game study among the multitude released every year. But it’s also telling that a study purporting to present clear-cut findings can engender such a spirited back-and-forth from researchers with differing views on the underlying issue being examined.

This ties into the broader argument Ferguson’s has made in his American Psychologist article and elsewhere. By now there have been tons of studies on video games with a range of results that, taken as a whole, don’t generally offer up much of the jaw-dropping variety, at least in terms of suggesting major, long-term effects on player well-being and proclivity for deviancy. But Ferguson is worried that, in part because of lingering echoes from the 1990s cultural panic over games (elements of which last to this day), the psychological community is promoting a lot of research that, while not necessarily bad science, isn’t really showing the meaningful effects you would expect given the sometimes-panicky tone of media coverage of this issue.

The incentive structure that we have in psychology for the most part is this sense that, ‘Hey, as long as you get as long as you get something that’s statistically significant, scare the shit out of people! Try to make it look like video games as bad as possible,’” he said. “And that culture is doing a lot of damage to the credibility of our field.”

A study should stand or fall on its own merits, of course, and Hull’s will receive plenty of scrutiny from researchers in the months to come. But given that we live in an age in which millions of adolescents are simultaneously playing video games while apparently engaging in less risky behavior than cohorts from years past, it’s hard to get all that worried about hordes of teens roaming the streets mimicking the behavior of Niko Bellic.

The Problem With Studies of ‘Deviant’ Games