Children begin watching television at a very early age, sometimes as early as six months, and are ardent viewers by the time that they are two or three years old. The general pattern of viewing is one of a steady rise in the number of hours viewed from early childhood through preadolescence and then a sharp drop in viewing during the adolescent years. According to audience rating surveys (Nielsen, 1988), the typical American household has the television set on for more than seven hours each day and children age 2 to 11 spend an average of 28 hours per week viewing (Andreasen, 1990; Condry, 1989; Liebert & Sprafkin, 1988). Naturally, the content viewed is more important than the amount of viewing and televised violence is one of the chief concerns.

The most extensive analyses of the incidence of violence on television are the studies conducted by Gerbner and his colleagues on the nature of American television programs. The results of these yearly analyses of the level of violence on American television for the 22-year period 1967-89 (Gerbner & Signorielli, 1990) indicate a consistently high level of violence. There were some minor fluctuations in the early 1970s followed by a steady increase to 1976, a sharp decline in 1977, and then a steady climb to an all-time high in 1982-83. According to Gerbner's initial analysis (Gerbner, 1972), eight out of every ten plays broadcast during the survey period in 1969 contained some form of violence, and eight episodes of violence occurred during each hour of broadcast time. Furthermore, programs especially designed for children, such as cartoons, are the most violent of all programming. Later analyses by Gerbner and Gross (1974, 1976a, 1976b) indicated that there was some decline in violence levels from 1969 to 1975, at least in terms of the prominence of killing. However, the level of violence dramatically increased in 1976 (Gerbner et al., 1977) and was followed by a decline to one of.the lowest levels in the 1977 season (Gerbner et al., 1978). This decline was quite dramatic. From the 'bumper-crop violence harvest' of 1976 to the relatively placid 1977, the percentage of programs containing violence fell from 90 to 75.5; the rate of violent episodes per hour fell from 9.5 to 6.7; and the rate of violence per program fell from 6.2 to 5.0 episodes. However, this downward trend was reversed in 1978 and through the early 1980s, and violence in weekend children's programs reached 30.3 violence episodes per hour in the 1982-83 season (Gerbner & Signorielli, 1990). Overall, the levels of violence in prime-time programming have averaged about five acts per hour and children's Saturday morning programs have averaged about 20 to 25 violent acts per hour.

In addition to broadcast television, cable TV adds to the level of violence through new, more violent, programs, and by recycling older violent broadcasts. A recent survey by the Centre for Media and Public Affairs (Lichter & Amundson, 1992) identified 1,846 violent scenes broadcast and cablecast between 6 a.m. to midnight on one day in Washington, D.C. The most violent periods were between 6 to 9 a.m. with 497 violent scenes (165.7 per hour) and between 2 to 5 p.m. with 609 violent scenes (203 per hour). Most of this violence is presented without context or judgement as to its acceptability. And most of this violence in the early morning and afternoon is viewed by children and youth.

What are the effects of this exposure to these levels of televised violence? What do we know about the influence of TV violence from the broad range of correlational, experimental and field studies that have been conducted over the past 40 years?