Exposure to television as a risk factor for violence.

It has long been hypothesized that exposure to television is a risk factor for aggression and violence; the earliest public hearings on the question took place in 1952, when fewer than one quarter of US households possessed television sets (1). Over the years, there have been scores of research reports on the issue (2-4), as well as several major governmental investigations (5-7). Throughout the controversy, however, there has been minimal input from epidemiologists. The relative absence of epidemiology from the discussion is attributable to a problem confronting any epidemiologist investigating exposure to television as a risk factor for violence: Virtually 100 per cent of the general population is exposed. No valid case-control groupings can be made in the United States comparing populations exposed to television with equivalent populations not exposed to television. Those rare individuals or enclaves who are truly unexposed to television—for example, the Amish—are manifestly atypical and so contaminated with self-selection bias as to preclude meaningful comparisons. It is safe to say that if valid exposed versus unexposed groupings were possible in the United