Illustration: Chris Sharp

Aburly young man named Toni Cipriani strolls through city streets, intent on murder but low on cash. He needs to buy a gun, and after a few muggings and carjackings, he has plenty of money. Afterward, he runs over some pedestrians, just for kicks. Every victim’s scream is followed by an audible squish.

Kenneth Lachlan, a Boston College assistant professor of communication, watches it all go down. “This is just mind-bendingly antisocial,” he comments. “There’s no reason for him to kill pedestrians at this point. He’s just having fun.” Nevertheless, to Lachlan, it means something. This taped segment of the video game “Grand Theft Auto: Liberty City Stories” features about four acts of violence per minute and is part of Lachlan’s research focus—the psychological impact of interactive-media violence. Every day, droves of Americans, children to adults, become Toni Cipriani for a while, until his ultraviolent world disappears in an electronic blink when it’s time for bed or school or work. Lachlan’s goal, as noted in “Models for Aggressive Behavior: The Attributes of Violent Characters in Popular Video Games,” published in the December 2005 issue of Communication Studies, is to find out which players may have trouble leaving characters like Toni Cipriani behind. Lachlan’s coauthors in the study are Stacy L. Smith of the University of Southern California and Ron Tamborini of Michigan State University.

The springboard for the researchers’ analysis is social cognitive theory, developed in the 1970s by the Canadian psychologist Albert Bandura, which posits that people are more likely to imitate characters who resemble themselves. The theory has been born out in studies of television violence and its viewers, but there has been no comparable analysis of video games—where the violence is interactive and can be more extreme. To learn what types of characters are committing video game violence, and hence which players might be most at risk for negative psychological influences, the researchers studied the first 10 minutes of the 60 best-selling games of 1999 (from “Major League Baseball 2000″ to “Zombie Revenge”). Each act of violence was counted and categorized in terms of graphicness (mild, moderate, extreme), motivation (from protection of life to mental instability), and the perpetrator’s age, gender, and ethnicity. The researchers noted 1,389 violent interactions in all, averaging out to 2.3 violent acts per minute.

They found that, of the aggressors, 71 percent were human, and 79 percent of those humans were male. By ethnicity, the largest group of perpetrators were Caucasians (about 70 percent of the violent humans). Asians were the next most violent group, accounting for around 14 percent of the mayhem caused by humans. Hispanic and black characters perpetrated 8 and 7 percent, respectively. Asians were the most likely to be “extremely” violent (12.5 percent were), but they were also always justified in taking action, whether from self-defense or altruism, compared with whites, whose violence was justified slightly under 76 percent of the time. Violence by the remaining ethnic groups was rarely justified (1.7 percent of the time).

The researchers speculate that the Asian profile of goodness and power owes to the “large number of games made by Japanese software companies.”

Their most surprising finding was that only about 10 percent of the violent perpetrators were children or teenagers. (In fact, children outnumbered teens by three to one.) Considering that young people are both avid video game players and more easily influenced than adults, the overall low incidence of youth violence “relieved our anxiety a little bit,” says Lachlan. On the other hand, he notes, social cognitive theory holds that behavioral models are not just those who resemble us, but also those whom we aspire to be. Violent adult characters, therefore, do “pose some risk” as behavioral models for the young, he says.

Lachlan is working on a new study about how individual game players’ personality traits influence the type and level of violence in the interactive experience. “If your child is [already] beating up other kids in the playground,” he suggests, “you might not want to let him play ‘Grand Theft Auto.’”

Chris Berdik is a writer based in Boston.

Read more by Chris Berdik