Media exposure correlates most strongly with an intention to use violence. There has been too much focus on the causal relations between media exposure and violence in society, with a failure to differentiate among different kinds of influences. There are five common influences of media violence. First, many studies have focused on imitation and aggression. Small children often imitate what they see on television, as part of the socialization process, but this does not necessarily make them more aggressive. Some kinds of media violence contribute to aggression more than others. Violence contributing most to aggressive behavior tends to involve an attractive perpetrator, justified actions by a "hero", and depictions that are extensive or realistic, or which do not show the consequences of the violence. (Comstock & Strasburger, 1990)

Besides imitation and aggression, a third important influence is fear. "Entertainment violence", or fictive portrayals, often frightens children, even though violence in the news frightens more intensively. This fear, if unexpressed or untreated, can manifest itself later as depression or aggression, especially in young people without strong family supports.
A fourth influence is a skewed perception of the reality of violence. On the one hand, media exaggerate the degree and prevalence of violence, making it seem omnipresent in society. On the other hand, children take away overoptimistic impressions about the body's ability to withstand violent attacks.
The final influence on which media-violence research has focused is habituation. Studies in the USA have shown that repeated exposure to media violence desensitizes viewers to violence in real life, such that they need increasingly strong "doses" to be affected by it. (Mees, 1990)

Research on media violence is further confounded by the varieties of outcome measures, media types, and geographic location. Studies that focus solely on physical aggression as an outcome may miss important gender and class differences. And a broader conception of the types and forms of violence may be important. Country-specific data are also needed, as cultures have different understandings of what constitutes acceptable behavior.

What interventions can counteract the influence of media violence, besides pressing the "off" button? Parents can watch programs with their children, discussing and expressing their opinions of what they see. Schools can create media-education curricula, in which children not only critique programs but also participate in making them, with the goal of turning them into "critical media consumers". Finally, although showing some violence may be important for understanding the world and creating social change, the media can regulate themselves, voluntarily restricting what they show.

The Effect Media has on Children and Violence
There is a tremendous body of literature demonstrating that watching violent television is associated with increased aggressive attitudes and behaviors (Paik & Comstock, 1994). This relationship is especially pronounced in children, whose relatively limited experiences and cognitive resources make them especially vulnerable to television (Wilson, Kunkel, Linz, Potter, Donnerstein, Smith, Blumenthal, & Berry, 1997). Because children's exposure to television is virtually inevitable, an appropriate research step is to try and find ways to minimize the likelihood that adverse effects will occur.
Fortunately, a small body of literature on "television mediation" offers some hope. Television mediation has been shown to be a three-dimensional construct that encompasses the following three forms of behaviors: talking to children about television, or "active mediation"; setting rules

The Effects of Active Mediation on Children’s Responses to Television
Active mediation has been studied in the context of a variety of responses. The research reveals that children who receive active mediation learn more from television content (Collins, Sobol, & Westby, 1981; Corder-Bolz, 1980; Desmond et al., 1985; Valkenburg, Krcmar, & de Roos, 1998), better understand the distinction between television and the "real world," (Desmond et al., 1985; Messaris & Kerr, 1984), are more skeptical of television (Austin, 1993), and have less stereotyped understandings of sex roles (Corder-Bolz, 1980). More formal programs, promoting active mediation via in-school "media literacy" curricula, have shown that active mediation can help children think more critically about television (Abelman & Courtright, 1983).

In addition, studies have explored the effects of various forms of active mediation on children's responses to televised violence. Experiments have revealed that children who hear negative comments about the violent acts in the television program or restrictions regarding television viewing, or "restrictive mediation"; and watching television with children, or "co viewing" (Nathanson, in press; Valkenburg, Krcmar, Peeters, & Marseille, 1999). Previous research on restrictive mediation and co viewing has produced somewhat inconsistent findings. Whereas some work demonstrates that these forms of mediation have positive effects on children, other work suggests that they have negligible or even negative effects on children's outcomes (Desmond, Singer, Singer, Calam, & Colimore, 1985; McLeod, Atkin, & Chaffee, 1972; Nathanson, 1997; Nathanson, 1999). On the other hand, research on active mediation suggests that it is effective in a variety of domains.

Although encouraging, research on active mediation leaves a number of questions unanswered. First, it is unclear whether children who hear negative comments about TV violence really learn to view the violence from a more critical perspective. In studies where children's post-viewing aggressive attitudes were measured, it is possible that children who heard these kinds of comments simply repeated the information they were told in order to meet social desirability demands. Or, because Hicks (1968) found that differences in post-viewing aggressive behavior were evidenced only when the children knew they were being observed, it is possible that children altered their post-viewing behavior simply to please the experimenter and not because they had learned that the violence they saw was unacceptable. It remains to be seen, then, whether active mediation really helps children interpret televised violence differently.

Second, because the mediation strategy itself-telling children that the violence depicted in particular programs is bad-requires an adult to identify and condemn each antisocial act in a given program, it seems that parents or other caregivers must always watch television with their children for active mediation to have an impact. What is needed is an active mediation technique that equips children with the analytical tools that will reduce their vulnerabilities to television even when they are viewing alone.
Third, research in this area has not yet provided an explanation for the reduction of aggression brought about by active mediation. In fact, most work on television mediation has proceeded without an overarching theoretical framework. What is needed is research that addresses how theoretically-derived active mediation strategies counteract the elements of televised violence that make children more vulnerable to learning aggression in the first place.

One such technique is to encourage children to think about the consequences of violence from a victim's perspective, or increasing children's "fictional involvement" with the victims (Tamborini, Stiff, & Heidel, 1990). An individual is fictionally involved when he or she recognizes or is aware of the thoughts and feelings of fictional characters. Tamborini et al. (1990), state that fictional involvement is the same as perspective-taking (i.e., seeing the world from the perspective of another individual), except that the target is a fictional character appearing in a book, movie, or television program. Theoretical explanations for media effects imply that this kind of approach may be useful in counteracting some of the features that make children more likely to experience adverse affects from televised violence. Specifically, one of the "context cues" in a program that encourages the learning of televised aggression is the failure to illustrate the negative consequences of violence for victims (Wilson et al., 1997). Research has shown that exposure to the negative consequences of violence reduces a viewer's tendency to imitate that violence (Schmutte & Taylor, 1980; Wotring & Greenberg, 1973). As Smith, Wilson, Kunkel, Linz, Potter, Colvin, and Donnerstein (1998) suggest "... the explicit depiction of pain and harm in violent portrayals is likely to inhibit the learning of aggressive attitudes and behaviors" (p. 18). Hence, getting children to think about the consequences of violence from a victim's perspective might be a good way to prevent them from learning aggressive attitudes and behaviors from televised violence.

This approach seems especially important given that the negative consequences of violence to victims are seldom shown on television (Smith et al., 1998). In particular, this strategy should be useful for counteracting the violence that typically occurs in children's programs. Many children's programs--especially the so-called classic cartoons (e.g., "Bugs Bunny," "Woody Woodpecker")--present violence in a humorous fashion that minimizes the pain and suffering of victims (Bjorkqvist & Lagerspetz, 1985; Wilson et al., 1997). As a result, when children watch this type of depiction, they may learn that violence is funny and has little negative impact on victims.

Thus, it seems particularly important to get children to consider the consequences of violence for victims when they watch classic cartoons and other genres of television that trivialize the consequences of violence for victims. As Leyens, Herman, and Dunand (1982) suggest, getting children to focus their attention on aspects of a program that they do not ordinarily notice and attend to should influence both how they view the program and how they react to it. Perhaps by considering the violence from a victim's perspective, children become more aware of the fact that violence hurts and injures its victims. If this occurs, it is possible children will evaluate the perpetrator more negatively, regard the victim more positively, and find the violence distasteful.

Although the effect of increasing children's fictional involvement with televised violence victims have not been explored in previous research, there is some relevant empirical work that suggests this might be a useful strategy. Voojis and van der Voort (1993) created an in-school curriculum consisting of nine, 45-minute lessons designed to get 10- to 12-year olds to change how they think about television violence by getting them to appreciate the consequences of violence. The authors found that, compared to children who had not participated in this program, children who were involved in the curriculum were less likely to approve of the violent acts of "good guys" and more likely to perceive violence in a serious light. Moreover, these differences were evident when the authors re-tested the children two years later.

Voojis and van der Voort's (1993) study is encouraging; however, it leaves two questions unanswered. First, it cannot tell us whether children who learned these lessons applied them while viewing an actual piece of televised entertainment. In addition, because the children in this study received an intensive series of lessons presented during school, it is unclear whether one simple message can have the same positive effects. Research is needed, then, to determine whether increasing fictional involvement with violence victims affects children's evaluations of and reactions to actual violent entertainment programs, and whether increasing children's fictional involvement with the victims of televised violence is possible outside of the context of a formal media literacy curriculum.