DOES PLAYING VIOLENT VIDEO GAMES CAUSE AGRESSIVE BEHAVOUR?

Reviews of video game research are as variable in their conclusions as the individual studies that comprise them. The same research is said to support different conclusions. For instance, Ask (1999), Funk (1993, 1995), Provenzo (1991), and Anderson & Bushman (2001) conclude that there is a causal connection between violent video games and aggressive behavior. Others think the data insufficient to support this connection (Cumberbatch, Maguire & Woods, 1993; Durkin, 1995; Griffiths, 1999; Wiegman, van Schie & Modde, 1997). Sacher (1993), reviewing mostly German research, found 5 experiments and 2 correlational studies linking violent video games to aggressive behavior, and 12 experiments and 7 correlational studies finding no such linkage.
In his overview of video game research, Barrie Gunter (1998, p. 109) concludes, "Even with experimental studies, there are problems of validity that derive from the fact that they do not measure ‘real aggression’ but rather simulated or pretend aggression."
According to British psychologist Mark Griffiths (1999) "the majority of studies on very young children tend to show that children become more aggressive after playing or watching a violent video game, but these were all based on the observation of free play."
Two recent meta-analyses (Anderson & Bushman, 2001; Sherry, 2001) report small effect sizes (r = .19 and .15, respectively). In the Sherry meta-analysis, playing time emerged as a negative predictor of effect size. That is, the more one played video games, the weaker the relation to aggressive behavior!
Meta-analysis is about the quantity, not the quality, of data. The conclusions of meta-analyses cannot be more valid than the studies that comprise them