May 14, 2018

Violence in video games has always been a hot topic in politics. President Trump recently met with the video game industry to discuss the impact of video games, in part, blaming them for the recent school shooting in Parkland, Florida. This is a very heated debate right now, with many in the gaming and publishing industries – and some in the academic world – declaring that playing video games has no real impact. But what does the science show?

A recent press release from the University of York states that there is “no evidence to support [a] link between violent video games and behavior.” As evidence, they cite several studies on violent video games by Zendle, Kudenko, and Cairns (2018). However, a careful reading of their paper reveals that they cannot prove that violent video games do not impact behavior.

Zendle and his colleagues created a first-person shooter game under various conditions, some more realistic than others. They reported two experiments in the paper and both followed a similar protocol. After playing either of the games, participants engaged in a word-completion task (described below) where they were asked to fill in a word based on a letter prompt. The study found that, regardless of realism, playing violent video games had no effect on subsequent aggressive thoughts. Now, on to the issues.

First – and most importantly – this research does not examine behavior. It looks at cognitions (e.g., thoughts). More precisely, it had players complete a word-fragment task to measure what they were thinking about after playing a video game. For example, participants might have been asked to complete the word “K_ _ _,” and researchers examined whether people filled in the missing letters to make it violent (like “KILL”) or non-violent (like “KISS”). However, this measurement of “what people are thinking” is very different than measuring actual behavior, like punching someone. So, first of all, the press release’s claim that playing violent video games has no effect on behavior was not tested in the current research.

The  working assumption was that people will be more likely to think of a violent word if they have just finished playing a violent video game. However, the Zendle study did not confirm this hypothesis, finding that participants who played a realistically-violent video game were no more likely to list a violent word than those who played a less-realistically violent game. Interestingly, this exact methodology has been used in a number of other studies that report that exposure to violent media increases aggressive thoughts (including this meta-analysis), but that playing non-violent games decreases them. Zendle’s study appears to be at odds with others that show a link between playing violent video games and increased aggressive thoughts.

Additionally, neither of Zendle’s experiments had a control condition. In other words, participants played either: 1) A violent video game, or 2) A violent video game. Yes, that’s right: both participants played violent video games where the “player’s task was to kill as many enemies as possible before the time ran out.” To be clear, there was no difference in the level of violence in either condition—the only difference was that one was slightly more realistic than the other in terms of how the bodies fell to the ground after being shot. Therefore, it would be completely foreseeable that there would be no differences between the two conditions in terms of the aggressive thoughts that came to mind because they both just played a violent game. To truly make the leap that playing violent video games had no impact on aggressive cognitions, you would need a control condition that contained no violence. Which interestingly, a study by Sestir and Bartholow included and found that playing violent video games did result in more violent word completions.

Saying all of this, I don’t believe the authors themselves make the outlandish claims described in the press release. Their study is on realism. Hypothetically, realistically-violent video games increase the likelihood of anger and aggression more than non-realistic games. For example, a recent study found that playing violent video games in an immersive environment resulted in more angry feelings than playing in playing in a less-immersive one.

But what about the Zendle study? How did they measure “realism”? In the first experiment, the only difference between the games was the way the bodies “crumpled” after being shot. In one condition, “enemies had death animations which dynamically mimicked how the real world behaves through the use of procedurally-animated ragdolls. In the other condition, enemy deaths were shown via pre-defined animations.” In other words, slain enemies fell to the ground more realistically in one condition than the other. (It is laudable that the authors created their own game and could manipulate its formal features.) Similarly, Zendle’s second experiment programed the enemy soldiers in one condition to act more like real soldiers in terms of defensive tactics. However, the acts of pulling out a gun and shooting an enemy did not differ in the two conditions. Again, it is perfectly predictable that there would be no difference between groups in terms of their aggressive thoughts.

Finally, the paper’s samples were less than ideal. When combined, they had nearly 2,800 participants, an impressive number, unheard of in traditional experiments. However, by recruiting and doing the study completely online, the researchers lost any sense of fidelity or experimental control, which any introductory research methods textbook will tell you is really important. We have no clue what participants were doing when they were (supposedly) playing the video game. They could have been texting, chatting on social media, running out for a snack… really anything besides focusing on the task at hand. In other words, we don’t even know if participants were playing the game, let alone focusing on it, giving us very little faith in the experimental process.

In sum, there are several issues with this study, only a few of which are highlighted in this article. Its original press release declares that there is “No evidence to support link between violent video games and behavior,” and several news articles picked up on this sensational claim. However, a careful reading of the paper provides little evidence to support that view.

Image Credit: Wikimedia Commons user Bayo // GNU General Public License

About the Author

Sarah Coyne is an associate professor at the College of Family, Home, and Social Science at Brigham Young University.

About our Editors