The Effects of Media Violence on Society

See allHide authors and affiliations

Science  29 Mar 2002:
Vol. 295, Issue 5564, pp. 2377-2379
DOI: 10.1126/science.1070765

Concerns about the negative effects of prolonged exposure to violent television programming emerged shortly after broadcasting began in 1946. By 1972 sufficient empirical evidence had accumulated for the U.S. Surgeon General to comment that “…televised violence, indeed, does have an adverse effect on certain members of our society” (1). Other scientific bodies have come to similar conclusions. Six major professional societies in the United States–the American Psychological Association, the American Academy of Pediatrics, the American Academy of Child and Adolescent Psychiatry, the American Medical Association, the American Academy of Family Physicians, and the American Psychiatric Association–recently concluded that “the data point overwhelmingly to a causal connection between media violence and aggressive behavior in some children” (2). In a report on page 2468 of this issue, Johnson and colleagues (3) present important evidence showing that extensive TV viewing among adolescents and young adults is associated with subsequent aggressive acts.

Despite the consensus among experts, lay people do not seem to be getting the message from the popular press that media violence contributes to a more violent society. We recently demonstrated that even as the scientific evidence linking media violence to aggression has accumulated, news reports about the effects of media violence have shifted to weaker statements, implying that there is little evidence for such effects (4). This inaccurate reporting in the popular press may account for continuing controversy long after the debate should have been over, much as the cigarette smoking/cancer controversy persisted long after the scientific community knew that smoking causes cancer.

Aggression researchers have adopted a triangulation strategy to examine the effects of violence in the media. Specifically, divergent research methods have been applied in the belief that using several unique methodological approaches yields a clearer picture than would be possible with any single method. Results of a meta-analysis of all available studies investigating the hypothesis that exposure to media violence increases aggression are displayed in the figure (4). A positive link between media violence and aggression regardless of research method is clearly shown (see the figure). Experimental studies demonstrate a causal link. Laboratory experiments yield slightly larger effects than other studies, presumably because of greater control over irrelevant factors (see the figure). Field experiments demonstrate causal effects in naturalistic settings. Cross-sectional studies demonstrate a positive association between media violence and types of real-world aggression (for example, assault) that cannot be studied ethically in experimental settings. Longitudinal studies reveal long-term effects of early media violence exposure on later aggressive acts. These effects are not trivial in magnitude. For example, they are larger than the effects of calcium intake on bone mass or of lead exposure on IQ in children (4). Interestingly, recent work demonstrates similar-sized effects of violent video games on aggression (5).

Media violence and aggression.

Effects of media violence on aggression for different types of studies. Diamond widths are proportional to the number of independent samples. There were 46 longitudinal samples involving 4975 participants, 86 cross-sectional samples involving 37,341 participants, 28 field experiment samples involving 1976 participants, and 124 laboratory experiment samples involving 7305 participants. Red lines indicate the mean effect sizes. Blue lines indicate a 95% confidence interval. Note that zero (dashed line, indicating no effect) is excluded from all confidence intervals.

The longitudinal study by Johnson and colleagues (3) is important for at least three reasons. It is the first published longitudinal study to link television exposure during adolescence and young adulthood to subsequent aggression, contradicting the common assumption that media violence affects only children. It therefore adds to extant research linking childhood TV habits to adult aggression and violence (6, 7). Second, its relatively large sample size (707 families) and time span (17 years) allowed a meaningful test of television exposure on severe aggressive behaviors (such as assault and robbery). Third, by statistically controlling for key childhood factors known to affect aggression (including childhood neglect, family income, neighborhood violence, parental education, and psychiatric disorders) the investigators were able to rule out numerous alternative explanations.

One potential problem with the Johnson et al. study is the use of hours of TV viewing, rather than hours of viewing violent TV. This is somewhat problematic because the primary source of TV viewing effects on aggression is believed to be violent content. However, about 60% of TV programs contain violence, so the number of TV hours correlates closely with the number of violent TV hours (8). Thus, the use of TV viewing hours in this study probably underestimates the effects of TV violence.

Recent theory about human aggression suggests at least two approaches to reducing media-related aggression (9). One involves reducing exposure to violent media. Robinson and colleagues reported one such intervention that significantly reduced aggression among third and fourth graders over a 6-month period (10). The other approach involves changing children's attitudes toward media violence. Huesmann successfully used this approach to reduce aggression in first and third graders over a 2-year period (11). The study by Johnson and colleagues suggests that media violence affects a larger group of people than previously believed, and that interventions for adolescents might also be beneficial. Such approaches are needed because a heavy diet of media violence contributes to a societal violence rate that is unnecessarily obese.


View Abstract

Navigate This Article