I read several stories on violence and video games today (CNN, Washington Post). Comments to these stories note that the link between violence and video game activity is not causal, but these folks probably did not read study cited because the logic of this study cannot be dismissed with the traditional position that pre-existing characterstics were not considered. Media interest in video game violence has re-emerged in response to a study published by Craig Anderson and associates. Anderson is from the Psychology Department at Iowa State. I was trained in this department many years ago so I pay some attention when my alma mater pops up in the news.
ISU Center For The Study of Violence announcement offers some insight into the argument for causality. Researchers assess pre-exposure level of violence and use as a covariate still demonstrating that exposure to violent video games among 9-12 year-olds still differentiate later evidence of violent tendencies.
I was able to locate the study online (from Pediatrics) through our library. The key finding and the statistical method is explained in the following quote from the study.
.. the finding that across 2 very different cultures HVGV predicts physical aggression 3 to 6 months later, even after controlling for previous aggressiveness and gender.
It is important to note the method used in this study does not manipulate access to aggressive content as would be the case in a true experiment. The method attempts to discount for nonrandom assignment by accounting for a measure of the pre-existing level of aggressiveness. Pre-existing agression predicts interest in violent video, but accounting for this correlation still leaves significant group differences.
As far as the quality of the research method goes, it is difficult to offer a better example than Bandura’s bobo doll studies from the 1960’s.
[youtube:http://www.youtube.com/watch?v=uMwOexrV6fM&feature=related]Yes the narration is not in English. Watch the behavior of the children after they view the models.
