Plugged In

by Patti M Valkenburg

This chapter explains the important role that cognitive development and social-emotional development play in adolescents’ media use and preferences. The broad period of adolescence can be divided into two age groups, early (ages 12–15) and late (ages 16–19) adolescence. We highlighted how the significant changes in brain development, particularly during early adolescence, lead to an increased interest in content that is fast and complex, relies on complex humor, and features riskier content such as extreme sports.

  Moreover, we highlighted how adolescents are charged with tackling three social-emotional tasks: developing an identity, learning about intimacy, and discovering their sexuality. As we have shown, these three developmental tasks have major consequences for behavior and preferences in early and late adolescence. For example, they are the reason early adolescents want to communicate constantly with peers, feel a need to belong, and seek information and validation for aspects of their identity from peers or from idols or heroes in the media.

  This triple social-emotional developmental challenge also explains why early adolescents spend so much time on social media and with entertainment media. Social media offer young teens ample opportunity to communicate endlessly with peers. In addition, more than ever before, social media provide early adolescents with the opportunity to discover and validate their identity, including their sexual identity. Entertainment media also help them in this respect. Media heroes and idols have long taught adolescents how to behave and how to deal with problematic social situations such as relationships, bullying, and falling in love.

  Late adolescents have some of the same preferences as early adolescents, such as a liking for fast-paced entertainment programs, but in other respects they begin to look much more like young adults. Their sense of autonomy and self-control increases considerably, and their media preferences are more mature. They are less concerned with accumulating as many friends as possible, but instead start to focus on the quality of their friendships and romantic relationships. It is a period typically marked by one’s first sexual relationship, and it is thus a crucial time for stabilizing one’s identity and establishing oneself as an autonomous person. This autonomy means that efforts to reach this audience through traditional “teenage approaches” are often unsuccessful. Instead, in this somewhat transitional period, it becomes increasingly important that media producers treat this audience as the autonomous individuals they are striving to become.

  7

  MEDIA AND VIOLENCE

  No one is suggesting that video games are the only reason they went out and committed those horrific acts, but was it a tipping point? Was it something that pushed them over the edge? Was it a factor in that? Perhaps. That’s a really big deal.

  —Jim Steyer, CEO, Common Sense Media, 2012

  No topic in the field of communication has been more heavily investigated than media violence and its effects on aggression. Every time a child or teenager committed an act of violence in recent years, the debate about the effects of media violence on aggression flared up again. Can children and teens indeed become aggressive, or even criminal, from seeing violence on television, in movies, or in games? And if so, are some children and teens particularly vulnerable to media violence effects? This chapter reviews the latest findings on the effects of media violence on aggression and criminal behavior. We first discuss key studies that investigated the effects of media violence on aggression. We then discuss the most important theories of why and how media violence may stimulate aggression. Finally, we reflect on how and why some children may be more—and others less—susceptible to media violence effects.

  Copycat Crimes

  The quotation by Jim Steyer that serves as the chapter epigraph concerns the tragic history of Adam Lanza. In December 2012, Lanza shot and killed twenty schoolchildren and six adults at Sandy Hook Elementary School, in Newtown, Connecticut, and then turned his weapon on himself. Following these events, some news media suggested that Lanza had used the first-person-shooter game Combat Arms to practice “head shots.” Lanza’s criminal deed was labeled a copycat crime, that is, a crime inspired by a similar crime in the past or in media.

  Copycat crimes have a history as long as the media. It is said that the publication in 1774 of Goethe’s The Sorrows of Young Werther, in which unrequited love drives the main character, Werther, to kill himself, led to a surge in suicides among readers. Newspaper reports in 1888 about the horrific murders of London prostitutes by the mysterious serial killer Jack the Ripper inspired a string of copycat crimes.

  In more recent decades, a number of alleged copycat crimes committed by youth were said to have been inspired by violent films or games. In 1993, two ten-year-olds from Liverpool murdered two-year-old James Bulger in broad daylight. The movie Child Play 3 was thought to have inspired their horrendous deed. In 1999, the world was shocked by the shootings at Columbine High School in Littleton, Colorado. Two boys, ages seventeen and eighteen, murdered twelve fellow students and a teacher. Their crime was said to have been inspired by the computer game Doom.

  In 2002, a nineteen-year-old at a high school in Erfurt, Germany, shot and killed fourteen teachers, two students, a police officer, and himself. The first-person-shooter game Counter-Strike was held responsible for his crime. The 2007 shooting spree on the Virginia Tech campus by Seung-Hui Cho, a twenty-three-year-old student, was ascribed to this game, too. In the Netherlands, Tristan van der Vlis killed himself in 2011 after carrying out an assault attack at a shopping mall. Some newspapers reported that a game he was fond of playing, Call of Duty: Modern Warfare 2, showed a gruesome resemblance to the bloodbath he caused.

  Are Copycat Crimes Evidence of Media Effects?

  In each of these horrific incidents, commentators suggested a link between violence in the media and the extreme behavior of the young perpetrators. Each of them either frequently viewed horror movies or played violent games, and each was said to have been inspired by the violence depicted in these films or games. Both suggestions are plausible. Many youth do indeed watch horror movies and play violent games. But did media violence actually cause these adolescents to commit murder?

  To establish causality between media use and crime, the circumstances must meet at least two criteria. The first is that the person’s exposure to media violence must predate his or her criminal behavior. The second is that all other possible explanations for that behavior can be excluded. The incidents described above appear to meet the first criterion: the perpetrators had all played the game or watched the movie before they committed their horrific crimes. That said, in each case, we cannot exclude significant alternative explanations for their behaviors. The boys who murdered James Bulger, for example, were problem children who had been severely neglected. One boy’s father was extremely violent. Both boys and their siblings spent their days and nights on the street and hardly ever attended school.

  The other perpetrators were either loners or had grown up in unusual and distressing circumstances. For example, Cho, the Virginia Tech shooter, had an anxiety disorder and was completely alienated from his family and friends. Personal circumstances like these are important alternative explanatory factors to consider. In fact, criminal behavior is usually the result of a complex combination of factors, including, for example, genetic predisposition, neglect, and exposure to violence in early childhood.1

  All in all, although it is plausible that media violence inspired the perpetrators, we cannot exclude alternative explanations for any of these crimes. That is why these incidents do not prove that exposure to media violence causes criminal behavior. In addition, these incidents underscore why it is difficult to conclude whether and when media violence leads to deleterious outcomes. Exposure to media violence among delinquent youth often co-occurs with a multitude of other risk factors, whose effects are incredibly challenging to disentangle.

  Research on Media Violence and Aggression

  In the case of copycat crimes, it is typically suggested that there is a relationship between media vio
lence and criminally violent behavior. Criminally violent behavior should not be confused with aggression or aggressive behavior. The lion’s share of research on the effects of media violence has focused on aggression or aggressive behavior. “Aggression” in these studies refers to aggressive thoughts and feelings, for example, the desire to punch someone or to take revenge. “Aggressive behavior” refers to a display of physical or verbal behavior (for example, fighting, cursing, or bullying) meant to hurt someone. By all estimates, more than six hundred studies have so far been published concerning the effect of media violence on aggression or aggressive behavior.2

  By contrast, there have been approximately thirty studies of the effects of media violence on criminally violent behavior.3 While criminally violent behavior is illegal behavior, and therefore punishable by law, aggressive behavior is not illegal—unless it becomes extreme. Thus, the boundary between aggressive behavior and criminally violent behavior is not always clear. Criminologists define criminally violent behavior as behavior “that transcends normal aggression and causes physical harm to others in a manner that is designated as illegal in the criminal code.”4

  Even fewer studies have looked at the effects of media violence on indirect aggression. Indirect aggression (also called social or relational aggression) involves aggression in which harm is delivered covertly, “behind the back.” Examples of indirect aggression include spreading rumors, damaging possessions, and trying to get others to exclude a peer from a social group. It has been estimated that teen television entertainment commonly depicts more indirect than direct aggression.5 And while interest in the effects of viewing dramatized indirect aggression has increased rapidly in recent years, to date only a handful of studies have investigated the effects of such content on teens’ indirect aggression.6

  In sum, most research on the effects of media violence has focused on direct aggressive behavior. Thus, the following discussion largely focuses on the empirical evidence regarding the link between media violence exposure and direct aggressive behavior. These studies are broadly divided into three categories: experiments, correlational studies, and meta-analyses. Where relevant, we consider criminal behavior and indirect aggressive behavior as well.

  Laboratory and Field Experiments

  There are two types of experiments, laboratory experiments and field experiments. In a typical laboratory experiment concerning the influence of media violence on aggression, half the study participants, the experimental group, watch a violent movie or play a violent game. The other half, the control group, watch a neutral film or play a neutral game (or neither) in the same setting. The researchers then measure aggression by observing the subjects, for example, during their play with dolls or other children. Sometimes researchers use other measurement instruments, for example, a knob or a button that allows the subjects to send a loud blast of noise to someone wearing headphones. After the test, the researchers determine whether the experimental group displayed more aggression than the control group, which most often is the case.7

  In laboratory experiments, participants are assigned randomly to the experimental and control groups. Random assignment is used to ensure that there are no a priori differences between the groups, for example, in the extent to which participants are already aggressive. If researchers detect a difference between the groups, that difference can be ascribed only to the effect of media violence. In other words, laboratory experiments have a high degree of internal validity (that is, the ability to establish a causal relationship). The disadvantage, however, of laboratory experiments is that they take place in an artificial environment, and thus lack external validity (the ability to generalize to settings typical of everyday life). For example, delivering a loud blast of noise is certainly not a “standard” form of aggression. Researchers thus can never guarantee that their results will be valid in real-life circumstances.

  This problem of external validity can be addressed by the second type of experiment, the field experiment or quasi-experiment. Field researchers often work with existing groups in their own environments, for example, with pupils at a school. A good example of a field experiment was one conducted by Jacques-Philippe Leyens and colleagues.8 In this study, children living at an institution for juvenile delinquents were shown violent movies every evening for a week while the control group watched neutral movies. Afterward, the researchers found that the children who had watched the violent films were more aggressive than those who had seen the neutral movies. Field experiments are conducted in the subjects’ natural environment, giving the results a relatively high degree of external validity. Yet they also have an insurmountable weakness in relation to the effect of media violence: they can never lead to definitive conclusions about causality. It is impossible to say that only the violent movie or game caused the experimental group’s aggression; some other factor may have played a role, such as an unforeseen incident of aggression in the group. Field experiments thus have a lower degree of internal validity.

  Correlational Research: The Chicken-or-Egg Dilemma

  Correlational (also called cross-sectional correlational) studies assume that if media violence stimulates aggression, then children exposed to high levels of media violence will be more aggressive than those who are not. In correlational studies, researchers typically collect data in schools or in families. They ask students or family members a series of questions about the number of violent movies that children see or the number of violent games they play. They also look at how aggressively a child behaves. They do this by observing the child, by asking parents or teachers to evaluate the child’s aggression, or by having children fill out questionnaires. The majority of correlational studies indicate that children who are frequently exposed to violent media are somewhat more aggressive than children who are less frequently exposed.9 The external validity of correlational studies is similar to that of field experiments, but their internal validity is minimal. Correlational studies can establish a relationship between media violence and aggressive behavior, but they cannot demonstrate that media violence causes aggressive behavior. After all, they cannot solve the “chicken-or-egg” dilemma: it is impossible to determine which came first—media violence or aggressive behavior.

  Researchers can help compensate for the chicken-or-egg dilemma by conducting causal-correlational, or longitudinal, research. Such research again involves evaluating children’s media use and aggressive behavior, but the measurements are conducted at two or more time points. In doing so, researchers can better establish whether media violence exposure may be a precursor to or a consequence of aggressive behavior. While the internal validity of these designs is certainly stronger than that of cross-sectional correlational studies, it remains possible that unmeasured third variables (for example, a child’s temperament, peer and family circumstances) may explain potential associations between media violence and aggression.

  Leonard Eron and colleagues were the first to investigate the influence of television violence and aggressive behavior in causal-correlational research.10 They observed a group of eight-year-olds to determine how much they liked watching violence on television and how aggressive they were. Ten years later, they observed the same subjects as eighteen-year-olds. They showed that watching violent movies at the age of eight predicted increased aggressive behavior at age eighteen. There was no suggestion of a reverse correlation; that is, aggressive behavior at age eight did not lead the subjects to watch more violent television programs at eighteen. Some later studies, however, did reveal a reciprocal relationship between media violence and aggressive behavior. These studies showed that media violence influences aggressive behavior, but that aggressive behavior has just as much influence on youth’s preference for media violence.11

  Meta-Analyses

  Meta-analyses are studies that use statistical techniques to summarize the results of numerous experimental and correlational studies. Meta-analyses involve entering the statistical effect sizes produced in individual empirical studi
es into a new database. This database permits researchers to determine an average effect size based on the effect sizes of the studies included in the meta-analysis. Provided they are performed properly, meta-analyses typically gain more respect from the research community than individual empirical studies because they can lead to refinements in scientific theories, show which questions have and have not been addressed in research, and identify new directions for study.12

  Starting in the 1990s, researchers carried out several meta-analyses on the influence of television, movie, and gaming violence on aggressive behavior. All have shown that media violence consumption is associated with aggressive behavior among youth. The first large-scale meta-analysis, carried out by Haejung Paik and George Comstock, encompassed 217 empirical studies. The researchers found a correlation of r = .31 between watching violent movies or television shows and aggressive behavior.13 Five meta-analyses published in the new millennium concerning the effects of video games on aggressive behavior have similarly shown a positive correlation between violent games and aggressive behavior, with effect sizes ranging from r = .08 to r = .20.14 Statisticians consider statistical effects of this size to be small to moderate. The studies on the effects of viewing indirect aggression in the media are still too scarce to justify a meta-analysis, but the empirical studies that have been conducted thus far suggest that viewing indirect aggression can also stimulate both direct and indirect aggression.15

  Two meta-analyses have studied the correlation between media violence and criminally violent behavior. In Paik and Comstock’s meta-analysis, the correlation between media violence and criminally violent behavior was smaller than that between media violence and aggressive behavior (r = .06 based on the correlational studies). This result was replicated in a meta-analysis conducted by Joanne Savage and Christina Yancey in 2008.16 In their study, the meta-analytical correlations between media violence and criminally violent behavior ranged from r = .06 in experiments to r = .12 in longitudinal research. That said, according to Savage and Yancey, the quality of the approximately thirty included studies varied significantly—which undermined the reliability of their meta-analysis. They concluded that the relationship between media violence and criminally violent behavior has yet to be established, but that this does not mean that such a relationship does not exist. More research is required into this effect, and a clear distinction needs to be made between criminal violence and other aggressive behaviors.