* Media Violence and Its Effect on Aggression: Assessing the Scientific Evidence. Jonathan L. Freedman. Toronto: University of Toronto Press, 2002. 227 pp. $50 hbk. $24.95 pbk.
"In 1999 I was approached by the Motion Picture Association of America [MPAA] and asked whether I would consider conducting a comprehensive review of all the research on media violence," says Jonathan L. Freedman in the Preface to Media Violence and Its Effect on Aggression. The media's cultivation of academics who disparage research showing that their products are harmful is not new: It goes back at least as far as Will Hays' lavish support of Mortimer Adler in the 1930s (Adler 1977, 193-94; Vaughn 2003). Freedman, a psychology professor at the University of Toronto who has never conducted his own research on media violence, claims that the financial support has not affected his objectivity, although he does not hesitate to see ulterior motives in the scientists and professionals who disagree with his conclusions.
Freedman expresses outrage at social scientists and public health organizations for wrongly (in his view) concluding that media violence promotes aggressive behavior. His criticism essentially boils down to two arguments. The first is that the professional organizations have exaggerated the number of scientific studies that have been conducted on the topic. The second is that a study-by-study analysis reveals that there is no consensus in the findings.
Freedman is correct that the number of studies has sometimes been overstated. Although some organizations have cited a number as high as 3,500, recent meta-analyses have placed the number between 200 and 300. Freedman explains that the inflated number originally referred to all types of articles about media effects, not just scientific studies of media violence. Somehow this number was picked up by others and misapplied. Freedman considers this "the worst kind of irresponsible behavior" and finds the use of this figure to be as "sloppy" as an economist saying that his research was based on data from "over 150 American states"!
Freedman never says how many more studies he would consider necessary. If we look at research findings in other areas, however, 200 would seem quite sufficient. For example, the finding that calcium intake increases bone mass is based on thirty-three studies (Welten, Kemper, Post, and van Staveren 1995); the conclusion that exposure to lead results in low I.Q. scores is based on twenty-four studies (Needleman and Gatsonis 1990).
The bulk of the book includes a tedious, close analysis of every published scientific study of the effects of media violence on aggression or desensitization that Freedman could find. (See the Reference List for some he missed and more recent compelling evidence). Not surprisingly, Freedman considers many studies unconvincing. Although some of his criticisms of individual studies are justified, he seems strongly motivated to find flaws. Moreover, after giving an exhaustive explanation of research methods, he forgets one basic principle: that the lack of a statistically significant difference is not the same as a finding of no effect. In addition, he disputes the fact that meta-analysis, which statistically combines all the findings in an area and eliminates the subjective interpretation of individual studies (Mann 1994), is an appropriate way to discover a research consensus. Freedman discusses two meta-analyses (Paik and Comstock 1994; Wood, Wong, and Chachere 1991) that report a clear conclusion that media violence promotes aggression, but he dismisses them. Two recent meta-analyses (Anderson and Bushman 2001; Bushman and Anderson 2001) are not included. Freedman also chooses not to cover research on media violence's effect on fear, simply claiming that "the research has not provided much support for it."
Why does Freedman think there is so much bias in the interpretation of media violence research? …