I just spent some time going back through Giumetti and Markey's research. It refined the work of Craig Anderson and stressed the importance of anger as a factor of increased aggression after playing violent video games.
The 2007 study by Giumetti, G.W. & Markey, P.M. took a focused look at the role of personality traits and anger in studies similar to Anderson’s. Giumetti and Markey used methodology that mirrored Anderson’s in order to accurately study the affects of anger on Anderson’s results. The study used the General Aggression Model (GAM) model following Anderson’s application and a methodology similar to Bushman’s and Anderson’s (2002). The participants played one of 3 violent games (Doom 3, Mortal Kombat: Deadly Alliance, or Return to Castle Wolfenstein) or one of 3 nonviolent games (Tetris Worlds, Top Spin Tennis, or Project Gotham Racing) for 15 min. After the game was finished, they provided actions for an open ended story. The model of aggression testing has been used before (Anderson and Bushman, 2002). This studied tested for short –term increases in aggression. 167 undergraduate students (79 F, 88M) studied and took an Aggression Questionnaire to determine their dispositional anger personality traits.
The violent games produced significantly more aggressive responses. But individuals with low anger traits were not significantly affected by video game violence. Low anger respondents only showed a .43 increase in aggression (not statistically significant) compared to the 2.89 increase from high anger respondents. The results support that anger personality traits are a key element in the relationship between violent video games and aggression. Respondents who were not angry were relatively unaffected by violent video game play.Giumetti, G.W., & Markey, P.M. (2007) Violent video games and anger as predictors of aggression. Journal of Research in Personality, doi:10.1016/j.jrp.2007.02.005