Predicting Youth Aggression

Many parents worry whether purchasing action-oriented video games will harm their kids, making them more aggressive or less empathic. At the extreme, it’s not uncommon to hear politicians attempt to blame high-profile acts of violence, such as mass shootings, on aggressive video games. Although links between games and criminal violence have largely been discredited, it’s not unreasonable to ask: what about bullying, or schoolyard fights, or just being less kind? Is this something we should worry about?

Experimental studies of aggression have always been something of a mixed bag, both in terms of methodological quality and outcome. And they have increasingly come under criticism as unreplicable or using weak aggression measures. I conduct a fair number of experiments myself but agree that finding out whether someone is inclined to give another person more hot sauce in a sandwich as a prank (a common aggression measure used in many experiments) doesn’t tell us much about what we really want to know about fighting or assaults.  Increasingly, scholars are turning to longitudinal studies that track kids over time to see what factors in their early lives are predictive of later outcomes like assaults or decreased helping behaviors.

A Singapore database of just over 3,000 youth followed over two years has gotten a lot of attention. This excellent, representative sample tracks multiple early-life predictors, including game play, family environment, personality, and mental health. The data make it possible to look at multiple outcomes related to aggressive and helping behaviors. In the past, some studies using this dataset suggested that aggressive games might promote aggressive behavior later in adolescence. However, these studies have since been criticized for potential questionable researcher practices. That is to say, the research group involved seemed to calculate the video game variables differently from one study to the next. This would be like calibrating a thermometer differently from one study to the next, something scientists should never do. Doing so increases the risk of false positive results, meaning we can’t be sure if the reported aggressive game effects were real. Psychological science has a wide problem with scholars injecting their opinions into data, either inadvertently or intentionally. That appears to be what might have happened here.

We found that playing aggressive video games early in life was not related to any of those outcomes. Kids who played more aggressive video games showed no more problem behaviors than those who played fewer such games.

In our study, we sought to reexamine this excellent dataset using more rigorous methods. Using a process called preregistration, we publicly posted our analysis plan in advance of looking at the data. That way we could not recalculate our variables or alter our analyses to achieve a particular result. We were able to control for numerous variables to examine the impact of early exposure to aggressive video games. These included preexisting aggression, family environment, mother’s education, sex, intelligence, and self-control.  We looked at seven outcomes related to aggressive behaviors and fantasies, prosocial behaviors, and forgiveness.

We found that playing aggressive video games early in life was not related to any of those outcomes. Kids who played more aggressive video games showed no more problem behaviors than those who played fewer such games. Factors like negative family environment and low impulse control were more relevant to later aggression. Our results suggest that earlier attempts to use this dataset to link aggressive games to kids’ aggression were mistaken.

Ours is just one of many recent studies to conclude aggressive games are not a predictor of youth aggression. It increasingly appears that fears of games are akin to the moral panics of the past over comic books, rock and roll, and the radio.

Photo by Aleks Dorohovich on Unsplash