Since the early days of television, people have been concerned about the possible negative effects of television on violent or aggressive behavior. One research direction to studying these effects has been to explore people's perceptions of the characters they encounter on the programs they watch. Perceived homophily, identification, and parasocial relationships with these characters are important because they may "mediate short- and long-term emotional reactions to depicted events and to characters themselves" (Hoffner & Cantor, 1991, p. 64). Connections between these relationships and trait aggression, which might help explain viewer aggression, have not yet been tested. We sought to account for these relationships with aggressive characters as they are linked to one's level of trait aggression.
Researchers have found that the effects of encountering violent television content are not uniform, but are based on viewer characteristics (e.g., Surgeon General's Scientific Advisory Committee, 1972). Among these characteristics is a disposition to behave aggressively. Aggressive dispositions are trait characteristics that develop over time and are stable across situations and time (Botha & Mels, 1990; Huesmann, Eron, Lefkowitz, & Waider, 1984). Huesmann and Eron (1986) noted that these tendencies predict serious adult antisocial behavior. Situational factors, such as familial and environmental background and observational learning conditions, interact with such aggressive tendencies to influence the probability of an aggressive response. In general, researchers have found support for the view that more aggressive people are influenced by exposure to media violence (e.g., Dorr & Kovaric, 1980).
Whereas people who are not aggressive in nature also may be influenced by media violence, the nature and extent of the effects are likely to be different from the effects on more aggressive viewers. For example, media violence may teach nonaggressive viewers aggressive attitudes, but it is likely to do this and more (e.g., reinforce existing aggressive attitudes) for aggressive persons. Violent media portrayals also might have nonaggressive outcomes, such as bringing about prosocial behaviors (Evans, 1989). Here, however, we were interested in aggressive outcomes because they are more problematic for promoting antisocial attitudes and behaviors, which are more likely to exist in aggressive individuals. We assessed viewers' self-reported aggression and their homophily, identification, and parasocial relationships with aggressive characters, which should mediate the content-effect relationship.
Some evidence for the importance of people's aggressive dispositions is provided by researchers who have noted that aggressive tendencies influence viewing preferences (e.g., Gunter, 1983, 1985). Gunter found that people with aggressive dispositions enjoy watching violent content, perceive violence in shows to be more humorous and exciting, and are more tolerant of others' violence than less aggressive people. Others have observed similar results about the influence of aggression on program preferences for children (e.g., Atkin, Greenberg, Korzenny, & McDermott, 1979; Cantor & Nathanson, 1997; Eron, 1963), for adolescents, and for adults (e.g., Diener & Defour, 1978; Heller & Polsky, 1976; Robinson & Bachman, 1972). Selective exposure provides one explanation of this effect (e.g., Zillmann & Bryant, 1985).
According to the selective exposure hypothesis, people selectively choose what they will be exposed to in the media. Although most studies have not focused on aggressive content, it is likely that selectivity influences the choice of such content, in that people prefer supportive rather than discrepant messages to validate their thoughts, feelings, or actions (Atkin et al., 1979). Social cognitive theory helps explain this selectivity in exposure to television content and the impact of viewing such content. …