James D Robinson. 21st Century Communication: A Reference Handbook. Editor: William F Eadie. Sage Publication. 2009.
Communication scholars and researchers have long been interested in the way things are portrayed in media. Much of this concern stems from the pervasiveness of media in our daily lives. Currently there are more than 300 million people living in the 111,384,000 households in the United States. Recent Nielsen estimates suggest nearly every home in the United States has at least one television. The 114 million households with televisions have the sets on for about 7 hours per day, with the average viewer watching TV about 30 hours per week. Television viewing represents about 30% of the average American’s leisure time and is the single largest leisure activity for most Americans. In addition, the Internet is becoming as ubiquitous as television.
The average American spends about 2.3 hours per week using the Internet for entertainment purposes.
While an extensive review of media usage patterns is outside the scope of this chapter, it is clear that if other media choices are included, such as time spent listening to the radio, reading newspapers, magazines and books, listening to music, watching movies, and playing video games, Americans spend a great deal of time with media. With so much media stimulation going on, it is no wonder that researchers spend a great deal of time discerning what sights, sounds, and vicarious experiences audience members are likely to experience. While this chapter will focus primarily on television, those interested in studying media portrayals research can examine media content in any of the aforementioned formats or channels.
Concerns about the way things are portrayed in media generally stem from two perspectives that we can call the “mirror of society” view and the “social influence” view. Scholars operating from the mirror of society view examine media content because they believe that such examinations provide insights into the nature of our culture. They believe that media portrayals are a reflection of the way society thinks or feels about an issue. In a simple sense, the mirror of society view suggests that what you see in the media is a reflection of what society thinks or feels.
Hacker (1951) and others have suggested that examination of media portrayals is important because the social status of groups and individuals can be identified through these portrayals. Hacker argues that groups of high social status will appear more frequently than will their low-status counterparts and that the portrayals of high-status groups and individuals will be more positive. Conversely, members of groups believed to be held in low esteem or who are of relatively low social status in society—such as the elderly—appear very infrequently in media. Evidence supporting this perspective is quite strong. Even the most cursory examination of the portrayals literature demonstrates that some groups are systematically underrepresented on television over long periods of time. On television, adults tend to be disproportionately white, young, attractive, wealthy, and male.
Because of its pervasiveness, social critics often focus their attention on media portrayals on television. The ubiquity of television is often offered as an argument for studying TV content, but portrayals scholars have studied many different channels including comic books, radio, magazines, newspapers, music videos, the Internet, billboards, film, commercials, brochures, record covers, CD cases, and greeting cards.
While the media portrayals literature often focuses on television, literature reviews focusing on a particular portrayal across a variety of different channels can provide further evidence to support the notion that media are a mirror of society. For example, researchers examining portrayals of the elderly appearing in newspaper articles, magazines, prime-time television programs, children’s programming and cartoons, magazine advertisements, letters to Dear Abby, children’s literature, poetry, TV advertisements, magazine cartoons, birthday cards, and even jokes report that older adults are indeed underrepresented, occupy less prominent roles on television, and are often portrayed negatively.
By using media portrayals as a barometer of social status, researchers can make comparisons between the U.S. and other cultures. Research examining portrayals of the elderly in cultures where the elderly are held in higher esteem, such as China, Japan, and Korea, find the elderly represented quite differently. Older adult characters are found more frequently, occupy more prominent roles in the programs, and are portrayed in a more favorable light.
A much more common view adopted by media portrayals researchers is based on the notion that media portrayals affect audience members and their attitudes. From this perspective, scholars and social critics are concerned that inaccurate or stereotypical portrayals have a negative impact on audience attitudes and perceptions of reality. This social influence view is much more commonly employed in the portrayals literature and is predicated on a much wider set of theoretical underpinnings than the relatively simple notion of media mirroring society. In general, the social influence perspective or view suggests that media portrayals affect audience members but does not explain the theoretical mechanism causing these effects. Thus, a number of middle-range theories have been offered up by scholars to provide the specifics about how media portrayals influence individual audience members and ultimately society. The next section of this chapter examines some of those theoretical perspectives and discusses some of the research that has been produced from each perspective.
Theories Commonly Employed in Media Portrayals Research
Perhaps the most widely accepted theoretical explanation for media effects was first proposed by Albert Bandura (1977) to explain how television violence affects children. His theory—social learning theory—is in its most simple form a theory of modeling or imitation. Bandura believed, and much subsequent research has demonstrated, that in
addition to learning through the “trial and error” of their own behavior, children can also learn vicariously through the observation or modeling of others. In fact, Bandura suggests that anything that can be learned from direct experience can also be learned vicariously or by watching someone else—real or imaginary—enact the behavior. He further contends that people can often learn things more efficiently through modeling or imitation than they do through their own trial and error efforts. Research into the social learning process clearly indicates that audience members can and do learn vicariously from media models.
Social learning is far more than the mere imitation of mediated models, however. In addition to being able to watch the model behave, audience members can also learn about the consequences or outcomes of performing the model’s behavior. Thus, from a social learning perspective, it is important not only to identify the types and frequencies of particular behaviors occurring but in addition to identify the outcomes or consequences of those behaviors. For example, a researcher interested in portrayals of sexual harassment might also examine the consequences of sexually harassing behaviors. If the harasser is punished or socially ostracized, the audience member may learn not to engage in sexually harassing behavior. If, on the other hand, the mediated portrayal of harassment is accompanied by a laugh track indicating harassment is just a “joke” and sanctioned by other characters, the audience members may learn that harassment is not a big deal. Clearly, that is not the message anyone involved in the writing of a television show intends to send, but it is nonetheless a distinct possibility—particularly within the realm of the situation comedy. The relevance of this theory to the study of media representations is even clearer when you realize the use of multiple models and models of high social status, when the situations or context of the model performing the behavior is realistic, and when audience members find themselves in situations similar to those presented in media.
The effects of social learning are not short-lived and have been shown to last as long as a month. Furthermore, when media portrayals provide detailed accounts of some target behavior and portray the consequences of such actions as positive, even if the positive outcomes are transitory and the deviant action is later punished, people can and do learn the antisocial behavior. The ultimate negative consequences can be easily forgotten or misunderstood by some audience members, and others may attribute the failure to unrelated factors (e.g., bad luck). In fact, Bandura (1986) suggests that audience beliefs about the nature of the behavior and its consequences often outweigh direct experience with the behavior and its consequences. Reliance on the source for information further increases the likelihood of social learning occurring.
Thus, from a social learning theory perspective, media portrayals provide models for audience members to acquire new behaviors and insights into the consequences of those behaviors. Social learning theory, however, assumes that audience members have the ability to control their behavior and do so based on their understanding of the consequences of the behavior as well as their own moral and experiential worldview. So audience members do not mindlessly imitate what they see on television but rather use what they have learned—that is to say, the behavior-consequence link in their decisions about how they should behave along with what they believe to be right or wrong in their decision to enact or not engage in a particular behavior.
Another very common theoretical rationale employed in media portrayals research is the cultivation hypothesis. First proposed by George Gerbner (1969), the cultivation hypothesis suggests that when people watch television, they are acquiring or cultivating a view of the “real world” that is shaped by the way things are portrayed on television. Unlike a more traditional conception of learning, where the individual audience member does not merely learn the information presented to them via media, Gerbner argues that people do not intentionally or voluntarily adopt attitudes based on the information provided by a single television program, but rather, they acquire or “cultivate” a view of reality that is based on underlying cultural themes that occur throughout television programming. It is the underlying themes that cause audience members to cultivate a view of reality that more closely approximates the portrayal of reality in media. These underlying themes are things such as “The world is a scary place,” which arise because so many programs are predicated on the assumption that bad things happen all the time. Other such themes include “Might makes right,” which is based on the notion that violence is a common solution to problems; “It is ok to be impolite if you are being funny,” which assumes that being funny is valued more than other social norms; and “The elderly are a dying breed,” which suggests that the elderly are not valued.
The most interesting and perhaps the most controversial aspect of cultivation hypothesis is that it is the underlying myths or themes that run throughout media that are of concern to audience members. This is quite different from more traditional learning theories, which suggest that people learn about the world from television in much the same way they learn about anything else. They are exposed to the content, attend to the content, retain or store the content in their heads, and are able to recall the information when they need it. Cultivation effects occur without audience awareness of the process. Cultivation effects are particularly pernicious because the audience need not be aware that they are being exposed to those underlying themes and are consequently less able to defend themselves against the effects of media exposure. A further complicating factor is that it makes very little difference what audience members watch on television since the myths and themes are so pervasive that they run throughout television. In theory, the audience will be influenced just as much by watching the news or educational programming as they will from watching Jerry Springer or Dog the Bounty Hunter. Television is seen as a primary vehicle of story in our culture, and television is seen as a primary vehicle for the inculcation of our young. Whether the story is Hansel and Gretel or Live Free or Die Hard, the moral, myth, or theme is the same.
So the cultivation hypothesis has been widely used as a theoretical rationale for studying media content and is particularly popular as an explanatory mechanism because it allows researchers a great deal of latitude in deciding not only what they will study but also how they will study it. Studying manifest variables such as the age of the actors is one thing, but with cultivation as a rationale, researchers can also study underlying or more interpretative behaviors, such as character values or incidents of being polite.
More recent conceptualizations of the theory have begun including two additional concepts—mainstreaming and resonance. While these are outside the scope of this article, suffice it to say mainstreaming and resonance are concepts used within cultivation to explain exceptions to the general cultivation rule, which is “The more you watch, the more likely you are to view the world as it is portrayed on television”—regardless of other factors such as education and personal experience. Resonance occurs when viewers have direct experience in addition to the symbolic experience they gain through media exposure. People who have been mugged and watch a lot of television tend to think that the world is an even more dangerous place than do those individuals who watch a lot of television but have never been the victims of violence. Mainstreaming refers to those instances when heavy viewers from different backgrounds view the world similarly—even when their backgrounds suggest that they should not. For example, audience members with a high socioeconomic status (SES) should view the world as being less dangerous than would audience members from a lower SES. This is because people from a low-SES background are more likely to have personally witnessed or experienced violence. Thus, mainstreaming and resonance are used by scholars conducting cultivation research to explain those anomalies that occur among respondents who are either more different than they should be or less different than they should be, based on their experiences.
While social learning theory and the cultivation hypothesis are most commonly employed, a variety of other theories can be and have been employed by media portrayals researchers. Most often, researchers employing other theories focus less on media portrayals and more on the impact of those portrayals on audience members. Agenda-setting research also places an emphasis on media portrayals but generally focuses on the representation of issues by media more than character demography or behavior. Maxwell McCombs and Donald Shaw (1972) first proposed their theory of agenda setting in 1972 in an article examining the impact of media on the 1968 presidential campaign. They suggest that media do not so much affect audience behavior or even attitudes toward a particular topic. But media are most effective, according to McCombs and Shaw, at influencing the importance or salience of issues in the minds of audience members based on the prominence the issue receives in media content. In short, media can raise audience awareness levels about an issue by frequently discussing an issue. Thus, agenda-setting research often entails analyzing the content of some media channel (e.g., the newspaper) to determine which issues have been discussed, how many column inches were devoted to the issue, and where the stories were located within the newspaper (e.g., front page and above the fold). Once the prominence of the issues has been identified, the researchers typically survey audience members to find out how important they think those issues are. Often this is done by simply asking the respondents to rank the issues in importance. Agenda-setting effects are determined by finding a correlation between the prominence of the issue in media content (e.g., the number of times the issue was written about) and the importance audience members place on the issue (e.g., the rank they assign to the issue). In short, when an issue receives a great deal of attention, audience members should rate that issue as being more important than other issues that have received less attention.
Researchers initially focused on the first level of agenda setting. This first level corresponds to the preceding description of agenda setting. Such researchers simply identified an issue—say gun control—and then determined the prominence of that issue in media content. More recently, researchers have begun examining what is often described as the second level of agenda setting. While scholars are not in 100% agreement on the definition of the second level of agenda setting, it is reasonably safe to say that researchers examining the second level go beyond the frequency of issue portrayals in media to include the characteristics or attributes related to the issue presented by media. So in the case of the issue of “gun control,” portrayals can be framed as constitutional arguments (e.g., “The Constitution guarantees the right to bear arms”), self-defense arguments (e.g., “The police don’t come until the crime has been committed, so you must protect your own family”), or firearms-as-the-devil’s-tool arguments (e.g., “The widespread availability of guns in the United States contributes to or causes crime”). How the argument or issue is framed, as well as the characteristics of the issue that receive attention by media, is included in the analysis of media content. In terms of content analytic investigations, this means that the category systems are more complex and do not just include measures of issue portrayal frequency but also categorize issue portrayals by the type of depiction. This extension of the theory provides researchers a more complex and more specific measure of media portrayals. Generally, the more specifically two variables are measured, the higher the degree of correspondence you will find—assuming that a relationship actually exists between the two variables.
One criticism that has been levied against agenda-setting research has been the lack of a theoretical mechanism to explain why the prominence of an issue is related to audience perceptions of issue importance or salience. This shortcoming appears to have been remedied by Iyengar and Kinder (1987), who suggested that the mechanism underlying agenda setting may be the same mechanism that is used in cognitive psychology to explain “priming effects.” The simplest way to understand priming is to think of the mind as a device that needs exercise to be effective. When an individual gains information, that information is stored in an area of related information so it can be found or recalled in the future. When information in one storage area is activated or employed, the other information in that area is also exercised or activated. Consequently, when issues and their attributes are framed in a particular way, all that information is stored together within the mind and is recalled together on subsequent recollection efforts. Thus, when the National Rifle Association (NRA) pairs the issue of gun control with family safety and frames the issue in the Second Amendment of the Constitution, all those issues and related issues are stored together and are all activated when any part of that issue is relevant. Priming goes further, suggesting that not only are the issues stored together, but with exercise or activation, those beliefs are more easily recalled than other, less often considered beliefs.
It is important to note that agenda setting has traditionally focused on the prominence of issues in media content and the corresponding levels of salience or importance audience members place on that issue. Priming is a cognitive theory that focuses on how information is organized, stored, and retrieved within the brain. Thus, the correlations observed between media portrayal frequency and audience perceptions of salience are actually attributable to the way information is stored within the brain. This extension certainly does not reduce the value of the theory; rather, the addition of priming as an explanatory mechanism for agenda setting helps us better understand why issue prominence in media influences audience estimates of issue salience. It also helps explain why issue prominence does not affect all audience members in the same way. Some audience members have different types of information stored within that cognitive schema, and consequently, mentions of issues by media activate different information based on their idiosyncratic methods of storage.
While agenda setting, cognitive priming, and issue framing are not the same phenomenon, they often occur together. Examining not only the issue of prominence (agenda setting) but in addition the characteristics of the issue employed in the message (framing) and the other issues organized within the cognitive structures of the audience members (priming) provides a strong theoretical rationale for the studying of media content and portrayals.
Bradley Greenberg (1988b) proposed the “drench hypothesis” as an explanatory vehicle for the effects of media portrayals, though this is far less commonly employed as a theoretical rationale for studying media content or portrayals. Greenberg’s argument is clear and persuasive. In the drench hypothesis, Greenberg suggests that a single event can be far more influential or life changing than a series of smaller events. Events such as the shootings at Columbine High School, the Kennedy assassinations, or the space shuttle disasters (Challenger and Columbia) can have a much more pronounced effect on audiences than the stalagmitelike effects that occur over time with repeated exposures to much smaller and less meaningful events such as shootings by fictional characters on prime-time programming.
Greenberg is suggesting that highly memorable and impression-leaving events—real or fictional—can be more influential than repeated exposure to the small, less-memorable portrayals that occur on television. Seeing a single automobile accident—such as the death of Princess Diana—can be much more influential than the cumulative effects of a season full of NASCAR accidents or a lifetime of chase scenes from Hollywood. Proponents of the drench hypothesis recognize that, in terms of media portrayals, sometimes less can be more.
Greenberg also recognized that the argument underlying the cultivation hypothesis is weakened by the fact that audience members are still affected differentially by media portrayals. Some heavy viewers of violence, for example, view the world quite differently from other heavy viewers. This suggests that audience responses to portrayals differ, to some degree, by audience member. Thus, any theory focusing on the impact of the influence of media portrayals must take into consideration that audience members are affected differentially by the same message or series of messages. If audience members are affected differentially, then it is a logical necessity to recognize that different portrayals have different levels of influence on audience members.
While only a few studies have attempted to empirically test the drench hypothesis, the empirical evidence supporting the position is promising (see Bahk, 2001; Reep & Dambrot, 1989). There is, of course, potential for tautological reasoning when using the drench hypothesis. Portrayals that are highly memorable or have a high impact influence audience members more than low-impact portrayals do. Of course, the problem stems from the fact that if drench effects are observed, then the assumption is that the portrayals were high impact. If drench effects are not observed, however, it is not an indictment of the theory but rather evidence that the image was not impactful enough. This potential does not negate the utility of the theory but rather reminds us to scrutinize carefully the assumptions of the theories before we employ them.
What is very important here is the recognition that media portrayals can and undoubtedly do affect audience members differentially and that single events or portrayals can be just as important as or more important than the cumulative effects of media portrayals. This is not to say that the cumulative effects of media portrayals are unimportant. Rather, it is to suggest that the potential for media effects is a complex and multifaceted issue. The effects of portrayals may influence audience members cumulatively as well as from a single exposure. Similarly, audience members may also go relatively uninfluenced by media portrayals as well.
It is also important to note that while all of the theories discussed here examine the same issue—media portrayals—they are all quite different. Agenda setting focuses on how issue prominence in media content influences audience perceptions of issue salience. Priming theory focuses on how information is stored inside the head of viewers and accessibility. Accessibility here means that people are able to recall information because they have been primed or provided the opportunity to exercise that recall through exposure to media portrayals. Cultivation suggests that media portrayals influence audience members as a cumulative effect. Furthermore, cultivation theory suggests that it is not the information in the portrayals per se causing the audience to cultivate a particular worldview. Rather, cultivation suggests that the themes that run throughout media content are the culprit and not the portrayal itself. This nuance is often lost in discussions of portrayals. Audience members do not learn that elderly people are not highly valued in our culture from the content within media depictions of older adults. Rather, the audience members cultivate a negative perception of the elderly because they are underrepresented or depicted in a negative or stereotypical fashion. Those perceptions come from shows that contain older adults as well as shows that do not. By favoring younger adults on programs, the show is helping audience members to cultivate the perception that older adults are not so important. Finally, social learning theory suggests that audience members can and do learn.
Research Methods and Portrayals Research
Communication scholars employ a variety of different research methods in their quest to understand media portrayals, but the vast majority of the studies employ content analysis. Perhaps the most important scholar writing about content analysis was Ole Holsti. In his classic treatise on the subject, Holsti (1969) defines content analysis as “any technique for making inferences by objectively and systematically identifying specified characteristics of messages” (p. 14). Holsti was one of the first scholars to recognize that the technique could be used with text as well as any other durable data including photographs, oral communication, Web sites, brochures, or television programs. Content analysis is a research technique that takes samples of media content (e.g., a television program, newspaper article, Web site) and reduces the content into more manageable categories. For example, a scholar looking at media portrayals of race would watch the television program and every time a character appears on the screen would classify that character by their race. Thus, at the end of the study, the researcher would be able to make claims about the racial composition of television such as “12.5% of the characters on television were black.” In this way, researchers can examine how closely the characteristics of television mirror the characteristics of reality. For example, Robinson and Skill examined television portrayals of the elderly and found that only about 2.5% of the prime-time television viewers were 65 years of age or older. Obviously, the key to meaningful content analytic studies includes the development of content categories (such as the racial cohorts on television or the types and frequency of sexual harassment behaviors on television). In addition, it is imperative that anyone trained as a coder can reproduce the same or nearly the same results from the same data. This reproducibility of results is called reliability in the realm of research methods and is critical to the researcher employing content analysis as a research tool.
Of course, content analysis need not focus exclusively on such “obvious” categories of interest. In fact, one common criticism of content analytic research is that too often content analytic schemes focus on easily observable phenomena instead of the most important images. It is relatively easy to count the number of door knobs that appear during the course of a television program, but it is not particularly interesting or useful. It is much harder, for example, to identify all the acts of altruism that occur within a television program. Acts of altruism are a much more complex phenomenon and much more difficult to identify or code. Someone telling a “little white lie” so that someone else is not embarrassed could be coded as an act of dishonesty, or an act of kindness to save the face of the other interactant can be an act of kindness in one coding scheme. Similarly, researchers can not only examine the portrayals of some behaviors but also content analyze the consequences of those behaviors, thereby providing a more complex and potentially more useful coding system.
Critical to content analysis are the techniques employed by the researcher in developing the sample. Anyone content analyzing the front pages of newspapers during the month of September in 2001 would undoubtedly find that nearly every story focused on the terrorist attacks on the World Trade Center. Anyone looking to see how often issues such as education, the war on drugs, or the economy occur would conclude that there is little or no interest in those issues, based on their frequency of occurrence in the newspaper. So it is critical to use sufficiently large samples of media content and to draw those samples randomly if at all possible. Much like survey methods, however, there are of course times when such efforts are not possible or even desirable. The goals of the researcher should guide the sampling techniques employed by that same researcher.
Of course, content analysis is not the only method employed by researchers examining media portrayals. For example, researchers employing the cultivation hypothesis as the theoretical rationale for their study describe the content analysis component of their investigation as “cultivation analysis.” Once they have documented the content on television relevant to their study—either through their own content analysis or through previous studies that have analyzed media content—they may begin the second phase, which is often called “cultural indicators.”
Cultural indicators research typically involves conducting a survey of audience members to determine how much media they consume and their knowledge about and/or their attitudes toward an issue. For example, cultural indicators researchers ask respondents to estimate the proportion of the population that are police officers (e.g., 1 in 10, or 1 in 100, or 1 in 1,000) and generally find that heavy viewers believe there are more police officers than do light viewers. Thus, cultural indicators research is really survey research that attempts to find out if audience member perceptions of reality are indeed based on media portrayals.
The other theoretical perspective discussed earlier may employ such survey methods as well. Researchers employing social learning or priming theory as their theoretical rationale for the study are somewhat more likely to employ experimental or quasi-experimental designs. Nevertheless, most research media portrayals use content analysis to determine what content the audience is consuming and then employ survey techniques to examine the attitudes and beliefs of audience members.
Media Portrayals Research
It would behoove anyone interested in studying the media portrayals literature to read two books—Life on Television (Greenberg, 1988a) and Life on Daytime Television: Tuning in American Serial Drama (Cassata & Skill, 1983). Both books are excellent examples of scholarship, provide a historical context for current study of media portrayals, and address many of the theoretical, methodological, and coding issues that face media content researchers. The work by Gerbner and his colleagues in the area of cultivation is also well worth reading for the same reasons. Gerbner’s work (1969) is also useful because it will help students see how a theory and a method develop over time and how the theory guides researchers in their media content investigations.
If, however, a student were to be allowed only one reading on the topic of media portrayals, DeFleur’s (1964) article “Occupational Roles as Portrayed on Television” would be the definitive article to read. It is long enough and detailed enough to provide a wonderful introduction to the area. It is also an example of rigorous research that contains descriptions of the portrayals that are thick enough or sophisticated enough to give insight into the occupations and explains why such portrayals are potentially influential to children. Finally, it helps researchers see how decisions about issues such as sampling are actually made—DeFleur discusses why particular shows or types of shows were omitted from the analysis and how decisions about show inclusion were made. When two shows that were intended to be included in the analysis occurred at the same time, DeFleur used the flip of a coin to determine which show would be coded and which show would be excluded. It is important to recall that DeFleur was content analyzing the 250 half-hour programs without the benefit of a VCR and is quite careful to describe the decisions about coding and sampling that were made a priori and on the fly.
More recently, the National Television Violence Study (NTVS) was published and represents a spectacular example of content analysis as well as studies investigating the impact of those portrayals on audiences. This multivolume set of books is literally the last word on television portrayals of violence. The coding details and sampling techniques and statistical decisions are so carefully explained and well thought out that the books literally represent a graduate course in media portrayals research.
The NTVS constitutes the largest and most systematic content analysis of television programming ever conducted and reported in a single investigation. Researchers examined more than 10,000 hours of TV content gathered between 1994 and 1997 that were randomly selected from 23 channels aired between the hours of 6:00 a.m. and 11:00 p.m. Rather than simply coding acts of violence, NTVS researchers coded nine contextual features that may increase or decrease the risk of harmful effects, including the attractiveness of the perpetrator and victim, types of weapons, realism of the portrayals, whether or not the violence was justified or not, how graphic the portrayals were, and the outcomes of the violence.
Just more than 60% of the programs analyzed contained some violence, and the levels of violence were relatively stable over the 3-year period. Just under 40% of the violence was perpetrated by the “good guys,” and 33% of the “bad guys” who engaged in violent behaviors were not punished. It is interesting to note that in more than 70% of the scenes containing violence, aggressors were not criticized by the other characters, did not appear to feel any remorse for their actions, and were not punished at the time. Data from the 1996-1997 television season suggests that children watching 2 hours of cartoons a day will view approximately 10,000 violent incidents in a year. Typically, these acts of violence are sanitized and/or trivialized and do not depict the long-term consequences of the violence. The typical violent program contained at least six acts of violence per hour, and more than 40% of the violent scenes were either couched in a humorous context or contained an element of humor. These acts of violence spill over into programming oriented to children. In fact, children watching 2 hours of cartoons a day will see somewhere in the area of 10,000 violent incidents in a calendar year.
Even though most topics have not been as well studied as violence, this body of literature is growing at an astonishing rate. Even a cursory review of the journals illustrates that this is a very popular area of study and one that is no longer strictly within the domain of media scholars.
Studies discussing how illness and health care professionals are portrayed appear in medical journals just as studies examining portrayals of the elderly appear within the social gerontology literature. These studies examining portrayals of institutions, professions, and individuals are generally offered as explanations for audience attitudes and stereotypes. While a detailed review would be far too long to include here, a more general overview will provide an insight into the kind of work currently being done and an opportunity to discuss some of the common or stable findings within the media portrayals literature.
The demography of television has always been a popular topic of study. Typically, these studies make comparisons between media portrayals and reality and often use census data for such comparisons. For example, Robinson and Skill (1995) report that research has consistently shown that the elderly are underrepresented on television (and in nearly every other media channel ever examined). Most studies find that elderly characters represent fewer than 3% of the characters on prime-time television, and in the United States more than 12% of the population is 65 years of age or older, and they tend to be relegated to minor roles. Elderly characters also tend to be portrayed in a negative or stereotypical light. Characters under the age of 18 also tend to be underrepresented on television, and characters between the ages of 25 and 45 tend to be over-represented. This pattern exists on Saturday morning programming as well as prime-time programming.
Historically, portrayals research has shown that even though females represent 51% of the U.S. population, female characters have been underrepresented throughout the history of television. While estimates vary, most studies conducted before the 1990s indicated that females were outnumbered by males at a rate of 2 to 1, although this disparity appears to be decreasing. More recent estimates (see Glascock, 2001) of female characters on prime time suggest that about 63% of the characters on TV are male and only about 40% of all central or main characters are female. With well over 70% of the producers, directors, writers, creators, and executive directors being male, it is clear that this trend is just as true behind the camera as it is in front of the camera.
In addition to being underrepresented, female characters are more likely to be identified by their marital status than are male characters and less likely to be identified by their profession or occupation. Similarly, female characters are typically younger and occupy less prestigious and lower-paying jobs than do their male counterparts. While the percentage of female characters has improved significantly since the 1970s, they are portrayed in a different and often less positive light than are male characters.
Ethnic groups and minorities have also historically been underrepresented on television. Latino, Asian, American Indian, and almost all other groups appear on television very infrequently. They tend to be invisible, and when they do appear, they are cast in small and noncentral roles. They are also likely to be cast in stereotypical or negative roles when they do appear. Television portrayals of blacks have increased fairly dramatically since the beginnings of television. Recent census data suggest that approximately 12.1% of the U.S. population is black, and during the years 1990 to 1995, about 14% of all prime-time characters were black.
A recent estimate of the 2001 prime-time season (Glascock, 2001) suggests that those trends have remained relatively stable. Glascock reports that 81.1% of all prime-time characters were white, 13.9% were black, 3.0% were Hispanic, 1.5% were Asian, and 0.5% were Native American. The demography of main or central characters changes these percentages slightly. Glascock reports that 82.2% of all main characters were white, 13.1% were black, 3.3% were Hispanic, 0.7% were Asian, and 0.2% were Native American. The most recent census data suggest that the demographic profile of the United States looks slightly different from that of television. In the 1997 census, 72.7% of all U.S. residents were white, 12.1% black, 10.9% Hispanic, 3.6% Asian, and 0.7% Native American. Keep in mind that small fluctuations can be a function of sampling as well as the popularity of a particular type of show and do not necessarily indicate a cultural trend.
Other topics have, of course, received varying levels of attention from media portrayals researchers. A quick scan of the literature produced the following list of topics that have been addressed within the past few years:
- Demography of television: 1) Age, 2) Gender, 3) Race, 4) Socioeconomic status, 5) Occupations and professions, 6) Sexual orientation.
- Demography of television: 1) Illness, 2) Mental health, 3) Handicaps, 4) Suicide, 5) Diet, 6) Suicide, 7) Eating disorders and obesity, 8) Exercise, 9) Alcohol consumption, 10) Drug use.
- Families and relationships on television: 1) Dating and romance, 2) Marriage, 3) Families, 4) Divorce, 5) Relationships, 6) Friendships, 7) Grandparents, 8) Stepparents, 9) Self-disclosure.
- Institutions on television: 1) Religion and religious behavior, 2) Politics and political candidates, 3) Science and scientific phenomena and skepticism, 4) Legal system and criminal justice system, including violence, crime, terrorism, hate crimes, gangs, and firearms portrayals, 5) Health care system and health care professionals, 6) The press and the media, 7) Educational system
This list is not meant to be exhaustive by any means and does not come close to identifying the variety of studies that would be subsumed under each topic. In most cases, a study that examines a particular portrayal also gathers data on a number of other variables in the process. Studies looking at violence also report gender and race and may report other variables such as presence and types of weapons, how graphic the violence is, or whether or not the violence was justified. This list is just an indication of some media portrayals that have been examined recently to give you an idea of what kind of work is being done within this area. Much of the work that has been done has focused on specific behaviors (e.g., sexual behaviors and violence), but other work has focused on topics such as portrayals of altruism or depictions of agencies. The ability to create or discover a way of identifying the portrayals you are interested in is really the only limitation to studying media portrayals.
Even though a great number of studies have examined media portrayals and representations, there is still a great deal of work that needs to be done. Certainly, media portrayals can and have changed over time, so there is a continuing need to maintain accurate records about media content. Studies that compare portrayals from different countries are becoming more and more common. Similarly, studies that examine portrayals in media other than television are becoming more popular again. Future research needs to examine portrayals across multiple media channels and cultures to determine how these portrayals affect audience members. In addition, future research needs to examine the differences in media representations between central characters and peripheral characters more closely. Central characters may be more important as potential sources for social learning, but the characteristics and behavior of peripheral characters may shed more light on attitudes held by audience members. Finally, the strengthening of the theoretical rationale for these media content studies will benefit both the content analysts and the social critics and help scholars better understand the relationship between media and society. Anyone studying media portrayals would benefit greatly from considering these programs as narratives. Such an examination may illuminate the nature of what makes a good story as well as the potential for those portrayals to affect audience members.