Science —

Kids who played shoot-em-up games in the ‘90s were probably (mostly) OK

Study looking at negative impacts of video games finds small effects.

Kids who played shoot-em-up games in the ‘90s were probably (mostly) OK

The persistent suggestion that video gaming leads to violent behavior prompts innumerable eye-rolls and Internet rants from gamers. But it’s persistent because it’s surprisingly hard to nail down a solid answer to the question. A lot of the research just raises more questions, so consensus remains elusive, despite claims to the contrary.

A fair number of studies suggest that there is a link, but those can be contrasted with other research that says there isn’t. The problem is that there are so many different factors to take into account, along with a swiftly-changing medium and difficulty in obtaining high-quality data—we'd need an avalanche of research to answer the question definitively.

While it's not an avalanche, a group of researchers, led by biological psychologist and video game violence skeptic Peter Etchells, has published an analysis suggesting that players of violent games might face a very small increase in risk for behavioral problems. They’re the kinds of small results that would be met with disappointment by authors who were hoping to find an effect, but they’re there. And yet, as always, this analysis isn’t the final word.

Juggling the confounds

This field is so tricky to navigate partly because of the wealth of potential confounding factors and difficulty finding causation among the correlations. It’s conceivable that substantial exposure to on-screen violence could cause a change in behavior, but it’s also conceivable that people with behavioral problems would seek out violent media. And many studies have looked at “games” in general, which encompasses a wide range of media.

Even if we manage to focus on just violent games, there’s the problem that games that are violent are also other things as well: challenging, competitive, fast-paced, and so on. If players of these games show higher levels of violence, it’s difficult to say for certain that it’s the violence that's causing the problem. For instance, games that are highly competitive have been found to increase aggression in the short term, even if they're non-violent.

Games are also a modern and evolving phenomenon, which means that the best data is still in production.

Ideally, to see whether a long-term habit has long-term effects, you need longitudinal studies that takes measurements from people over years of their lives, but that data takes decades to collect. Etchells and his team were lucky enough to be able to use an existing longitudinal study, but it came hand-in-hand with the weakness that the games in question were all from the '90s. To see the effects of modern games, we'd need to wait even longer.

No sense of Doom

The data used in this analysis came from people participating in the Avon Longitudinal Study of Parents and Children (ALSPAC), which had started with more than 14,000 pregnant women in 1991 and 1992. Around 2,400 of the children in the study had answered a questionnaire on their gaming habits when they were eight or nine years old, and around 5,000 had completed an interview called the Development and Well-Being Assessment (“DAWBA”) at the age of fifteen. Approximately 1,800 children fell into both categories.

The researchers focused on two outcomes of the DAWBA: risk for depression, and risk for “conduct disorder,” which is a term describing antisocial behaviors in children. Then, they tried to eliminate or control for as many confounding factors as possible. They looked for children who had been rated as high-risk for conduct disorder by their parents when they were seven years old and removed them from the study. The researchers included family history of mental health, maternal education and socio-economic status, religiosity, family structure, gender, bullying victimhood, IQ, and social and emotional problems in their model as well.

What they found was that children who had reported playing shoot-em-up games (a stand-in category for violent games) at the age of eight or nine had a slightly increased chance of conduct disorder, even after controlling for all the confounding factors. The effect was weak, though; just on the border of statistical significance. The initial risk for conduct disorder is small, too, so any additional risk above that is going to be even smaller.

There didn’t seem to be an association between the number of games the children reported owning and an increase in risk for conduct disorder. When examining depression among shoot-em-up players, there was evidence for increased risk before the researchers controlled for all the confounding factors, but not afterwards.

Not exactly smoking

There are some obvious weaknesses in this study. They couldn’t control for the question of whether the children who played shoot-em-ups were already predisposed to problems but didn't yet display them at age seven. They also had to rely on self-reported data from eight-year-old children, who might not have known what a “shoot-em-up” game was when they answered the questionnaire.

Even more problematic is the question of what games eight-year-olds are likely to play. Maybe kids generally only really start with violent games when they’re a little older, which is why this study didn’t find much. A more detailed look at gaming habits in later childhood and adolescence could find a stronger link.

In any case, '90s games are not today’s games, which have changed drastically, and often don’t fit in clear genres that are clearly violent or non-violent. We need to move away from a generalized discussion of “video game use,” the authors argue, and look rather at the content of particular games.

This research sits squarely in a middle ground: it found an effect that ties in with other, more alarming research, but the effect sizes were so small that they don’t really provide much support for the idea that video games are a major concern. “Some have claimed that the magnitude of this effect is larger than the effect of exposure to smoke at work on lung cancer rates,” they write. “Our findings do not support such claims.”

PLOS One, 2015. DOI: 10.1371/journal.pone.0147732  (About DOIs).

Channel Ars Technica