New study on videogames and aggression yields surprising results

Last month's issue of the respected journal Developmental Psychology details a study on the possible links between aggression and wishful identification with violent game characters. The study, which involved 112 adolescent boys, concludes that "identifying with violent video game characters makes players more aggressive." While this conclusion is not surprising, and will most likely just confirm most peoples' fears about videogames, I found a couple much more fascinating results buried in the article's discussion section. These are:

1. There was no significant correlation between the perceived realism of the game and the player's feeling of immersion.
2. Realism and immersion did not influence the players' aggression levels.

Both of these results seem to fly in the face of the conventional wisdom about videogames. The first statement confirms something that I have actually been trying to articulate for a long time, but have often felt unequipped to do so. The second statement also goes against the idea that the (supposed) negative effects of videogames are heightened by increased realism. These conclusions are worth pondering, and I'm surprised that the authors didn't highlight them more in the article.

Also of interest is the way the researchers measured aggression. After playing a randomly assigned game for 20 minutes, the subject then played a competitive reaction-time game against another boy. (For ethical reasons, the game was rigged to make it seem like they were playing against someone.) Before starting, the subject was told to select the noise level (from 0 to 10) of a sound that would be blasted in the opponent's headphones if he lost, with the added warning that a noise level of 8 or above could cause permanent hearing damage. The point was to get an idea of the subject's willingness to inflict unprovoked harm.

On its face, I find this approach potentially problematic. First, some (though not all) young adolescents cannot fully grasp the notion of something having a permanent effect. If someone is fundamentally incapable of understanding what it means to cause permanent hearing damage, then such a warning would be lost on them. Second, many adolescents are used to having their parents and authority figures try to scare them with warnings about how dangerous everything is. Someone who has been overloaded with deterring statements might be less likely to take what adults say seriously, instead figuring that they are just exaggerating. Finally, given that the subjects don't actually see the results of what they are (apparently) doing to their opponents, the scenario could be too abstract and unrealistic to apply in everyday life.

Despite these reservations, I believe the researchers did a creditable job with this study. While I'm usually skeptical of studies that claim causative relationships (correlation is not causation), in this case I don't think the authors are venturing too far beyond what the data supports. The games that the boys played prior to the aggression task were assigned at random, ruling out the possibility that more aggressive subjects were simply more likely to choose violent games. In addition, the games were divided into four categories using input taken from a separate group of adolescent boys of comparable age to the test subjects. Incidentally, the games used in the study were as follows:

• America's Army, Killzone and Max Payne (violent-realistic)
• Doom 3, Quake, and Metroid Prime (violent-fantasy)
• Pro Evolution Soccer, The Sims 2, and Tony Hawk's Underground (nonviolent-realistic)
• Mario Kart, Mario Sunshine, and Final Fantasy (nonviolent-fantasy)

With the possible exception of Final Fantasy, the above groupings appear reasonable. At the very least, the fact that they were organized based on feedback from adolescent males lends some added credibility to the article.

In the future, I hope to see more research that examines the connection between perceived realism and levels of immersion. Although not the focus of this particular study, the finding that perceived realism does not necessarily influence a player's feeling of immersion has important implications for how game developers approach their work.