Video games, racism and violence: the problems behind the science 'proving' they're linked

Earlier this week a study seemed to suggest that video games could entrench racist beliefs on the part of the player.

The study asked the player go through two missions in Saints Row 2 with a black protagonist, and then measured their association between black individuals and certain negative words. Video games, it seemed, may lead to solidifying racism in the minds of those who played games with black heroes.

"The media have the power to perpetuate the stereotype that blacks are violent, and this is certainly seen in video games," Dr. Brad Bushman, the co-author of the study, said. "This violent stereotype may be more prevalent in video games than in any other form of media because being a black character in a video game is almost synonymous with being a violent character."

There are some massive issues with this study, and the conclusions reached based on the provided data.

The science is questionable, and it's important to make that clear

The study relies on the implicit association test, which isn't the best predictor of real-world behavior, negative or positive.

"IAT is fascinating science and I'd encourage everyone to try it out. However it is a proxy measure," Dr. Tyler Black told Polygon. Dr. Black is the clinical head of the BC Children's Hospital on the Child and Adolescent Psychiatric Emergency Unit. He's studied violence and video games, and has spoken on the subject extensively at shows like PAX. [Disclosure: I was a guest speaker on one of his panels about the connection between video games and school shootings.]

"We're not sure of the relationship between IAT and real world racism, stereotypes, etc. It is a fantastic paradigm but it doesn't ‘prove negative racial thought,’" he continued. "In the same way, unless homicide/violence by way of hot sauce becomes a common method of assault, it too is a proxy for violence. How does hot sauce selection relate to real world violence?"

This is the study’s other big issue: It links the use of hot sauce to aggression, stating that the amount of hot sauce someone puts on food meant for another is a measure of aggression, and by extension, violence. The line between IAT and racist behavior or thoughts is much like the line between hot sauce and real-world violence … in that it’s not nearly as clear and direct as these researchers would like us to believe.

No one is worried about the use of hot sauce among children, we're worried about real-world violence. My son is active in both soccer and karate, and those sports in tandem teach both team work and self-reliance. I have no doubt that they also both make him more aggressive in the short term. But is that a bad thing? He's likely respond with more aggression to many of these abstract studies after a game or a sparring match, but that doesn't make him likely to commit violence against another child.

"Confounding variables" also provide us with some insight into what this study does and doesn’t show. A confounding variable is anything that it is statistically related to the effect being studied without being the focus of the study. They can be tricky to control, and can skew the results of the research, especially when not explicitly discussed or controlled.

"This reflects the brains desire to make efficient connections"

"We do not know the IAT results for the group to begin with. We know they were white and not black, and were forced to play black characters," Dr. Black explained. "Is the confounder ‘likeness?’ If the avatar was green and the IAT about negative green stereotypes, would it show an effect?"

The list of confounding effects goes on and on. "If the character were forced [to be] female, stereotypically nonviolent, would the IAT of male participants have shifted toward bias of negative attitudes toward female? How about other races?"

There are so many variables the data becomes hard to take seriously when such strong claims are being made. "Is the IAT simply detecting the novelty of the black Avatar? In five minutes of brainstorming with a colleague, we came up with about four obvious confounders that were not controlled for here," Dr. Black stated.

"I played Mass Effect as a female Commander Shepard. I am confident that this changed my IAT scores for female/aggression or female/weapon immediately after playing," he said. "This reflects the brain's desire to make efficient connections, more than it likely does an evolving stereotype of female violence in my brain."

This is a trend

Dr. Bushman has a long history of trying to tie violence and video games together in the court of public opinion. He’s used bowls of candy to try to prove that video games make children dishonest after playing games. He’s claimed that playing video games cause children to be better with firearms, and more likely to aim for the heads of their targets.

The study in question asked participants to play video games with a "pistol shaped" controller and then rewarded them for head shots. How many people play games that include a controller shaped like a gun? The game being played specifically trained the participants to aim a firearm and then rewarded their behavior for aiming for the head, and this is supposed to prove that all gamers will somehow model that behavior?

He’s claimed that playing video games cause children to be better with firearms, and more likely to aim for the heads of their targets.

"My colleagues and I found that typical college students who played violent video games for 20 minutes at a time for three consecutive days showed increasingly higher levels of aggressive behavior each day they played," Dr. Bushman wrote on CNN. "If that's what happens to typical college students, how might someone like [Aaron] Alexis react to playing for 16 straight hours? What if he does this for months or years?"

These sorts of leading questions leave science completely behind in order to scare the reader into action. Dr. Bushman then makes the link between video games and gun violence explicit.

"Alexis was not the first mass killer to have an obsession with violent video games. Adam Lanza, who killed 26 children in an elementary school in Newtown, Connecticut, was also said to be a fan of first-person shooting games," Dr. Bushman continued. "Other killers have been found to be avid players."

Dr. Bushman has already made up his mind about the link between video games and violence, and looking at his past words and work in this area give us some context for the studies he’s recently today and will likely continue to release in the future. The racial aspect of the latest study is just another excuse to show video games in a bad light.

The data often shows that violent acts committed by children have decreased since video games have become more popular, which is rarely reported in the media.

"Not surprising research coming from these primary authors, who have a long tradition of gaming negative scientific publication," Dr. Black said. "That bias, however, doesn't necessarily affect the paper, but the study design led likely led them to the result."