Violent Video Games: Do We Know That They Cause Violent Behaviors?

California has banned the sale of violent video games to minors, and concerned persons have argued that playing violent games causes children and adolescents to become more aggressive—an undesirable outcome as far as most of society is concerned. In a petition before the Supreme Court of the United States, two related groups, the Entertainment Merchants Association and the Entertainment Software Association, have asked for the ban to be removed. They argue that although it would be undesirable for young people to become more aggressive, in fact the evidence does not suggest that violent video games would cause this outcome.
The cynical—and rather natural-- response to the Entertainment Merchants is that their financial interests in the sale of violent video games are outweighing whatever concerns they might have about aggressive behavior and its impact on society. However, eighty-two people (of whom I am one) have signed an “amici curiae” (friends of the court) brief arguing against the ban. This brief can be read at http://www.scribd.com/doc/37676518/EFF-PFF-Supreme-Court-Amicus-Brief-in-SCHWARTZENEGGER-v-EMA-Video-Games-Case. The signatories are all people who have contributed in some way to the discussion of the possible influence of violent video games on youth aggression, and they have presented information, based on understanding of systematic research, that supports the argument of the Entertainment Merchants. (It is, after all, possible that the Entertainment Merchants can be both motivated by financial interests and right in their assessment of the effects of violent video games.)
The amici brief counters the research cited by the California state senator Leland Yee and condemning the ill effects of violent video games. Without saying so directly, the brief suggests that Senator Yee (who has a Ph.D. in child psychology) and others have been cherry-picking among the available research, and have concocted an unpalatable pie with the results. There are hundreds of published studies on violent media and their effects, going back to the ‘50s and the great comic book investigations. As it would be very unwieldy to cite all these, it is not surprising that Californians in favor of the ban have chosen among them.
Given that some choice among studies is necessary, though, how do we choose? The choices should not be a matter of searching for investigations that have the “right” outcome and support whatever belief someone had to begin with. An important theme in scientific investigation is that researchers search hard for evidence that will counter their hypothesis; the rule is to try to prove that the hypothesis is NOT true, rather than to look only for information that confirms the prediction. This means that in selecting from an overabundance of material, if we were to pay attention to conclusions at all (which we really shouldn’t), we would be looking for those that argued against our expectation.
The choices made among studies should be based on the quality of the investigation and the acceptability of the conclusions drawn. Because it is really quite difficult to do good studies of a complex subject like the connections between violent video games and aggressive or violent youth behavior, these choices must be made carefully. It can’t be assumed that any study published in a peer-reviewed journal has met all the highest standards for data collection and interpretation.
One real problem is that most studies of the effects of video games are in fact correlational studies. This means that the researchers have sought young people who already had habits with respect to video games , have asked them about their preferences and their game-playing habits, and have in some way assessed either their actual violent behavior or some related factor like angry moods. Taking these two kinds of information, video game use, and violence or anger, the researchers carried out a statistical analysis to show how much the two factors were related to each other. If they also created a graph to show the relationship, the way the graph looked would indicate how much of a relationship there was. Points (each representing a tested individual) would be scattered at random all over the graph if there was little connection between video game use and aggressiveness. If there was a strong relationship between the two-- for instance, that people who played a lot of violent video games were quite aggressive, and those who played few were not-- the graph would form a tight, straight line. The scattered graph would be said to show a low correlation, and the straight line graph a high correlation.
It’s important and interesting to know whether two measures are correlated. But when we have that information-- well, that’s all we have. Even the highest correlation does not tell us that one factor is the cause of the other. It doesn’t even tell us which one might have been the cause and which one the effect. In the case of correlational research on video games, there are two equally possible interpretations about cause and effect. One is that violent video games cause violent behavior, as Senator Yee and many others have argued. The other is that people who are prone to violent behavior enjoy violent video games and choose to play them often, while continuing their usual level of aggressiveness.
In recent years, researchers have used brain imaging in the form of the fMRI to try to get something better than a correlational approach to the study of violence. Perhaps the time will come when this kind of work can be more conclusive, but at present there are many questions about the results of fMRI studies and their relationship to real-world behavior. Like correlational studies, imaging studies cannot be assumed to give us solid information about causal connections between violent video games and violent behavior.
As the amici brief has argued, we presently have little choice among studies that use good methodology for investigating causes of violent behavior, because such methodology is so poorly developed. As a result, it is presently impossible to conclude that violent video games cause youth violence. If the only argument in favor of banning the games is that scientific work supports the decision, there is no reason for such a ban, because there is no adequate scientific work.
That said, I acknowledge that I dislike all video games myself, and think there are powerful esthetic arguments against them! But my likes and dislikes, or those of others, should not be confused with systematic research evidence.