VIDEO GAME VIOLENCE AND BEHAVIORAL AGGRESSION: A META-ANALYTICAL EXAMINATION OF POTENTIAL MODERATORS

Abstract

The current study used fixed-effects and random-effects meta-analyses to test four additional potential moderators of the relation between exposure to media violence and aggression. First, fourteen relevant studies meeting Anderson et al.?s (2010) best practices criteria were coded for characteristics related to the violent and non-violent (control) video games and for study authorship. Next, the program Comprehensive Metaanalysis was used to conduct a fixed-effects meta-analysis in order to examine the variability between the studies within the sample. Then, potential moderating effects were tested via random-effects analyses. Consistent with previous research, an average effect of r+ = .26 was found for the relation between violent video game play and aggression. The realism of the violent and non-violent video games, competitiveness of the non-violent video game, and authorship were tested for moderating effects. Although, violent video game realism was not found to be a significant moderator Q(1) = 1.45, p > .05, the unrealistic subgroup of studies produced a significant average effect, r+ = .33, p < .05, whereas the realistic subgroup did not, r+ = .18, p > .05. Non-Violent video game realism, non-violent video game competitiveness, and Craig Anderson as author did not moderate the relation, Q = .08, p = .776; Q = .01, p = .91; and Q = .30, p = .58, respectively. Future research should evaluate whether violent video game realism moderates the relation between playing violent video games and aggression using a more definitive methodology.