Posted by Yesterday we reported on the general reactions to studies on the impact of media violence. Today we’ll get into the specifics of those studies, as reported by Craig Anderson et al. in their report “The Influence of Media Violence on Youth.”
The history of research on media violence and its relationship to aggression (behavior intended to cause harm to others) is now more than 50 years long. The earliest studies focused on television and film, but now extensive research has also been conducted on music/music videos, news media, and video games. Nearly all the data from hundreds of studies points to a strong correlation between exposure to media violence and aggression.
Most of the experimental studies on film and television violence involve simply showing violent programming to children or young adults and then observing their behavior for a short time afterwards. For example, one Belgian study split boys in a home for delinquents into two cottages. One cottage was shown violent films every night for 5 days, while the other cottage ran nonviolent films. The “frequency of hitting, choking, slapping, and kicking their cottage-mates” was significantly higher in the violent movie cottage than in the nonviolent cottage. Even in another study of children in ordinary schools, exposure to violent films led to an increase in physical attacks in a game of floor hockey afterward.
The largest meta-analysis of experiments on media violence was conducted by Haejung Paik and George Comstock in 1994, and combined the data from 217 studies. Of these, 71 addressed physical aggression, and a significant overall correlation (r score) of .32 was found. (See this article for an explanation of correlation, and yesterday’s post for an explanation of why these results are significant). In addition, a smaller, but still significant r score of .13 was found for the 32 studies showing a relationship between media violence and criminal violence. In the same analysis, Paik and Comstock also found a correlation of .20 for the 200 cross-sectional surveys which compared violent TV viewing and aggressive behavior.
In addition to experiments and surveys covering short time periods, there have also been dozens of studies examining the relationship between exposure to violent TV and film in childhood and aggression in young adulthood. The largest study, conducted by L. Rowan Huesmann and his colleagues, found that when children were exposed to violent TV at a young age, they were significantly more likely to be aggressive as adults. For example, 42 percent of men who watched a lot of violent TV as children reported pushing, grabbing, or shoving their spouses, compared to only 22 percent of those who watched less violent TV.
The evidence for music lyrics and music videos is less convincing. There are some studies showing a limited impact of music videos on agressive thoughts and attitudes, but not behavior. The studies on violent songs are less conclusive still, but do demonstrate some trends toward aggressive or adversarial thoughts correlated with violent music lyrics.
Video games, on the other hand, have attracted considerable attention in recent years due to the increasingly realistic and violent content of many games. The Anderson and Dill study reported previously in Cognitive Daily is a good example of the type of results these studies have found. In 2004, Anderson led a group of authors in a meta-analysis of video game violence research, and found that as the methodology of the studies improved, so did the magnitude of the correlation between violent video game playing and undesirable outcomes. This graph provides a dramatic illustration of the results:
The only negative correlation they found was with “helping behavior”!
Overall, the research on media violence, whether it was experimental or correlational, has shown a significant correlation between media violence and aggressive behavior. Though the correlations are sometimes small, Anderson and his colleagues point out that they are at least as significant as other behaviors considered to be very risky, such as exposure to asbestos and smoking cigarettes.
Cognitive Daily’s coverage of the report will continue tomorrow.
April 13, 2005
April 13, 2005
April 13, 2005
November 21, 2005
April 18, 2007
December 2, 2008