Computer Games and Violence: Is There Really a Connection?

This study examines whether violent video games are uniquely associated with violent behaviour among adolescents, drawing on a representative sample of 9,887 Norwegian youth aged 13 to 18 surveyed in 2002 through the Young in Norway study. Its central aim is to test whether the reported link between gaming and violence persists once gender and age are controlled, and whether particular game categories, rather than gaming in general, drive any such association. Eight game categories were used, and violent behaviour was measured through self-reported frequency of fighting, assault, and threats over the prior year. The authors position the work against a literature that has been divided, with some researchers finding modest correlations and others finding none, and with earlier studies limited by small samples that prevented meaningful analysis by age or gender.

The findings indicate that the bivariate correlation between violent games and violent behaviour (around r = .22 to .23) shrinks considerably once gender is included, falling to about .15 for first-person shooters and .18 for action games. In multivariate models that enter all game categories simultaneously, the association with first-person shooters largely disappears after controlling for age and gender, except among 8th graders, while action games remain a significant predictor across all age groups. Racing games showed a unique positive association only in 11th grade, and role-play games a negative one in 12th grade. Taken together, all game variables explained roughly 4 to 9 percent of the variance in violent behaviour, with the largest shares in the oldest groups. The authors conclude that gaming has no strong general effect on violence, that the relationship is specific to certain categories, and that causality cannot be determined, since violent youth may simply be drawn to violent games.