[Skip to Navigation]
June 1990

Television Violence-Reply

Author Affiliations

Division of Health Science American Medical Association 535 N Dearborn St Chicago, IL 60610

Arch Gen Psychiatry. 1990;47(6):596. doi:10.1001/archpsyc.1990.01810180096015

In Reply.—  Gadow and Sprafkin have raised questions raised questions about three of the references we cited in our recent article. The articles they comment on were cited by us as examples of studies of behavioral effects, not "exemplars" as Gadow and Sprafkin state. In one case we even explicitly state there are significant methodological criticisms of the study. First we would like to respond to their specific criticisms and then comment on review article strategies in general.First, in regard to the Feshbach and Singer1 study, our statement that "boys in some of the homes had increases in aggression, while boys in other homes had decreases in aggression as a result of a television diet high in aggressive content" was meant to indicate in a concise manner that the results of this study were quite mixed. Feshbach and Singer did indeed fin that boys in some homes had significantly lower levels, and boys in one home had a nonsignificantly higher level of aggression. Whether this pattern of significance means that voilent television lowered aggressive behaviour or, ad Gadow and Sprafkin contend, that nonvoilent television increased aggressive behaiour cannot be determined from the data as presented by Feshbach and Singer. The test of singnificance are conducted on the posttest measures, with no direct comparision with the pretest measures. Consequently, we cannot ascertain if once score is rising or the other is falling or both.The more important problem with the Feshbach and Singer study, and one to which we would have devoted more space had we written an entire monograph or, as Gadow and Sprafkin did, limited our review to field experiments with children, was what has become known as the classic "Batman" effect. Children in the nonviolent television group were forbidden to watch may of their favorite shows through this research design, and many of the respondents protested loudly about this restriction, particularly in regard to the program "Batman." (It is important to note that "aggression toward authority" was being measured during these protests.) Finally, the nonviolent television group protested so vehemently that the researchers allowed that group to include "Batman" in their nonviolent television diet. To avoid the high level of frustration associated with deprivation of favorite shows, subjects were allowed to terminate participation in the study at any time, and 38% of the subjects did so. In one home, the original 36 subjects (18 each in aggressive and nonaggressive television groups) were reduced to 7 in the aggressive and only 1 in the nonaggressive television group by posttest time. The data from all of the original subjects, however, were kept in the posttest analyses, in the form of means based on number

Add or change institution