Sexual messages on television: comparing findings from three studies.

Author: Kunkel, Dale.; Cope, Kirstie M. Biely, Erica. Source: The Journal of Sex Research v. 36 no3 (Aug. 1999) p. 230-6 ISSN: 0022-4499 Number: BSSI99032575 Copyright: The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited.

 


Sexual socialization is influenced by a wide range of sources, including parents, peers, and the mass media (Hyde & DeLameter, 1997). In trying to understand the process by which young people acquire their sexual beliefs, attitudes, and behaviors, the study of media provides information about potential socializing messages that are an important part of everyday life for children and adolescents (Greenberg, Brown, & Buerkel-Rothfuss, 1993). The significance of media content in this realm stems from a number of unique aspects surrounding its role in the lives of youth, including its early accessibility and its almost universal reach across the population.

Electronic media, and television in particular, provide a window to many parts of the world, such as sexually-related behavior, that would otherwise be shielded from young audiences. Long before many parents begin to discuss sex with their children, answers to such questions as "When is it OK to have sex?" and "With whom does one have sexual relations?" are provided by messages delivered on television. These messages are hardly didactic, most often coming in the form of scripts and plots in fictional entertainment programs. Yet the fact that such programs do not intend to teach sexual socialization lessons hardly mitigates the potential influence of their portrayals.

While television is certainly not the only influence on sexual socialization, adolescents often report that they use portrayals in the media to learn sexual and romantic scripts and norms for sexual behavior (Brown, Childers, & Waszak, 1990). Indeed, four out of ten (40%) teens say they have gained ideas for how to talk to their boyfriend or girlfriend about sexual issues directly from media portrayals (Kaiser Family Foundation, 1998).

Just as it is well established that media exposure influences social behaviors such as aggression and social stereotyping, there is a growing body of evidence documenting the possible effects of sexual content on television (Huston, Wartella, & Donnerstein, 1998). For example, two studies have reported correlations between watching television programs high in sexual content and the early initiation of sexual intercourse by adolescents (Brown & Newcomer, 1991; Peterson, Moore, & Furstenberg, 1991), while another found heavy television viewing to be predictive of negative attitudes toward remaining a virgin (Courtright & Baran, 1980). An experiment by Bryant and Rockwell (1994) showed that teens who had just viewed television dramas laden with sexual content rated descriptions of casual sexual encounters less negatively than teens who had not viewed any sexual material.

Another important aspect of sexual socialization involves the development of knowledge about appropriate preventative behaviors to reduce the risk of infection from AIDS or other sexually-transmitted diseases. When teenagers begin to engage in sexual activity, they assume the risk of disease as well as the risk of unwanted pregnancy, and it appears that many lack adequate preparation to avoid such negative consequences.

Two Americans under the age of 20 become infected with HIV every hour (Office of National AIDS Policy, 1996). Almost one million teenagers become pregnant every year in the United States (Kirby, 1997). In the face of these sobering statistics, it is important to consider the extent to which media portrayals engage in or overlook concerns such as these, which are very serious issues in the lives of young people today.

In summary, media effects research clearly suggests that television portrayals contribute to sexual socialization. And despite the increasing availability of media options, young people still devote more time to television viewing than to any other type of mass communication (Nielsen Media Research, 1998). Thus, if television is an important source of information about sex, then it is important to identify the prevailing patterns used for presenting sexual messages on television.

In this report, we present a summary view across three independent studies of sexual content on television that we and our colleagues at the University of California, Santa Barbara have recently conducted (Cope, 1998; Kunkel, Cope, & Colvin, 1996; Kunkel et al., 1998). All of the studies share the same conceptualization and operational definition of sexual content, and employ many of the same measures assessing the context surrounding sexual messages on television. Comparing the pattern of findings across these three studies affords a unique opportunity to identify clear patterns in the portrayal of sexual messages on television.

METHODOLOGYThe three studies examined in this report include the Family Hour Study (Kunkel, Cope, & Colvin, 1996), which evaluated the programs aired on broadcast networks during the first hour of prime time; the Teen Study (Cope, 1998), which assessed three episodes each of the 15 most popular TV series viewed by adolescents aged 12-17; and the V-Chip Study (Kunkel et al., 1998), which compared the V-chip rating assigned to TV programs with the actual levels of violence, sex, and adult language found in each show.

The V-Chip Study is significantly broader in scope than the other two projects, with a total of more than 1000 programs examined overall, and includes two distinct data sets which we consider here separately. The first of these encompasses most hours of the day across a full range of 10 channels (including network and independent broadcast as well as basic and premium cable), which we label here as V-Chip Study: Composite Week. The second sample is limited to prime time programming on the commercial broadcast networks, which we label here as V-Chip Study: Prime Time. A summary of the key aspects of the sample for each of the studies is presented in Table 1.

DEFINING AND MEASURING SEXUAL CONTENTAll of these studies defined sexual content as any depiction of talk or behavior that involves sexuality, sexual suggestiveness, or sexual activities or relationships. Both heterosexual and homosexual portrayals were included, although we do not report these categories separately here. Sexual behavior and talk about sex are considered fundamentally distinct categories, and we report them separately in our findings below.

To be considered a sexual behavior, actions must convey a sense of potential or likely sexual intimacy. For example, a kiss of greeting between two friends or relatives would not be coded as sexual behavior, whereas a passionate kiss between two characters with a discernible romantic interest would be. The lower threshold for sexual behaviors measured by the studies was physical flirting, which refers to behavioral actions such as a woman licking her lips provocatively while gazing intently at a man in a bar. This example underscores that our measurement in this realm encompasses the full range of sexually-related behaviors, and should not be equated strictly with the consummate sexual behavior of intercourse. Other behavior categories included passionate kissing, intimate touching, sexual intercourse strongly implied, and sexual intercourse depicted.

Sexual dialogue, or what we term talk about sex, involves a wide range of types of conversations that may involve first-hand discussion of sexual interests and topics with potential partners, as well as second-hand exchanges with others that convey information about one's prior, anticipated, or even desired future sexual activities. For purposes of measuring talk about sex, the topics of both reproductive issues (such as contraception or abortion) and sexually-transmitted diseases (including but not limited to AIDS) were considered as sexual.

For both the Family Hour Study and the Teen Study, a range of different contextual features were examined. These included relationship characteristics for people involved in sexual behaviors, consequences of sexual behavior experienced by characters, and the presence of any risk or responsibility themes such as sexual patience, sexual precautions, and depictions of the negative consequences of sexual behavior (e.g., AIDS, STDs, unwanted pregnancy). Measures were gathered at both the scene and the program level.

CODING AND RELIABILITYJudgments about the programming were recorded by undergraduate student coders, who received between 40-60 hours of training before they began to generate data for the projects. An elaborate codebook of rules and examples for coding was developed at the outset of the first project to guide the content judgments. Each of the studies reported extensive reliability assessments, employing a framework devised for the National Television Violence Study (Wilson et al., 1997) which represents the largest scientific study of television content yet conducted.

For all three studies, reliability was assessed at two distinct levels: (a) agreement on unitizing decisions (i.e., identifying sexual content within scenes) and (b) agreement on the applicable context measures used to describe the sexual content identified. For unitizing decisions, the coders demonstrated CIAM (Close Interval Around the Agreement Mode) of between 79% and 100% across all tests on all the studies.(FN1) For identifying types of talk about sex and sexual behavior, the agreement ranged from 83% to 100%. Agreement on other scene-level measures ranged from 90% to 100%, and agreement on program-level variables ranged from 76% to 100%. Overall, given the large number of coders involved and the complexity of the coding tasks, the reliability analyses establish strong confidence in the consistency of the data reported in the studies.

RESULTSTALK ABOUT SEXWe start our assessment of sexual content by addressing the question of what proportion of television programs contain talk about sex. The data make clear that a majority of programs on television include some such material. The most broad-based sample is found in the V-Chip Study: Composite Week, which includes all programs aired between 7 a.m. and 11 p.m. across both boradcast and cable channels with the exception of news and sports. This study indicates that 57% of programs contain talk about sex, with those shows averaging more than two (2.3) scenes per hour with such content (see Table 2).

Data from the V-Chip Study: Prime Time indicate that prime time network programs are somewhat more likely to include talk about sex, with two of every three shows (67%) including some. Programs most popular with teens contain talk about sex just as frequently (67%) as prime time shows overall, while programs that air in the first hour of prime time (Family Hour Study) have nearly the same proportion (59%) that include talk about sex.

Table 2 also makes clear that talk about people's prospective interests in sex is the predominant type of exchange portrayed on television. Conversation in this category may address specific targets of sexual interest, either for the self or others, or may involve general statements or views on sexual topics or issues that guide one's sexual behavior. Such comments account for nearly three-fourths (74% and 72%) of all talk about sex in the two V-chip Study samples, although they are not quite as frequent in the Family Hour Study (59%) and Teen Study (62%) data.

Example: Talk about prospective sexual interests.

After attending a class on sexual harassment, the office workers grow paranoid about a new "no tolerance" policy. Matt is frustrated because he has romantic interests for Jana, one of the women in his office. In front of the entire staff, he loudly declares his affection for her. "I like you. I like you a lot and it's okay if you like me too." She replies earnestly, "I just want to sleep with you!" Matt encourages the other staff members to express their true sexual feelings. Jana invites Matt for an apparent tryst, asking "Do you want to go to the edit room?" He responds eagerly and they scurry off holding hands. Another man asks the girl standing next to him, "Supply closet?" She nods and he wraps his arm around her, leading the way. Suddenly, everyone leaves the room coupled with the exception of one fellow who sees the last two women remaining in the corner. Trying to look suave, he struts over to them and makes a sexual advance. He alone receives a rejection. (Working, NBC, November 12, 1997).

The other common type of talk about sex involves comments about sexual experiences that have already occurred, which account for between 18-21% of all cases across the different studies.

Example: Talk about past sexual experiences.

Jerry, George, and Elaine are among a crowd in a church, awaiting the start of a friend's wedding ceremony. Earlier in the day, George learned that Jerry had previously slept with George's date for the wedding, a woman named Nina. As they all sit and wait for the ceremony to begin, George grows angry thinking about the situation, and says quietly to Jerry that he demands "reparations" to even the score. Since Jerry has slept with Nina, and Nina is George's date, George insists that he should be allowed to sleep with Jerry's date, Elaine. Elaine then joins the conversation and says she absolutely will not sleep with George, and what is the difference who slept with who? Elaine whispers to George, "Nina slept with him, he slept with me, and I slept with Pinter," referring to the man getting married. "Nobody cares," adds Elaine. George responds at the top of his voice, "You slept with the groom?" The crowd gasps in shock and the wedding is disrupted. (Seinfeld, NBC, November 20, 1997).

Interestingly, comments classified as expert advice or technical information, which could range from a visit with a doctor to a news magazine feature on sexual issues, appear only rarely. These cases accounted for 5-6% of all talk about sex in the Family Hour Study and Teen Study. The two program samples from the V-Chip Study yielded even fewer scenes with such material, although this is likely due to the exclusion of news programming from this research (i.e., such material is exempt from the industry's V-chip system for rating programs).

SEXUAL BEHAVIORWe now turn to the second major aspect of sexual content, and address the question of what proportion of television programs contain sexual behaviors. The data here are more complicated to interpret because of a methodological shift that was implemented for the V-Chip Study. The Family Hour Study and the Teen Study reported all sexual behaviors that were observed in every program. In contrast, the V-Chip Study measured sexual behaviors only when they were judged to be the primary emphasis of the scene in which they were located. This more conservative approach was pursued for the V-Chip Study so that we could be confident that a sexual scene identified by our measures would be significant enough to warrant identification by an S content rating symbol, which indicates sexual content in the new television industry rating system.

When all aspects of each program are taken into account, including minor events within scenes, we see that a solid majority of shows (61% for Family Hour Study; 62% for Teen Study) include some sexual behavior (see Table 3). In contrast, when only scenes with a primary emphasis on sexual behavior are considered (V-Chip Study), the proportion shrinks substantially to 28% of all programs examined. These two perspectives are actually complementary, and enrich our understanding of the content. Certainly, the primary emphasis scenes would be more likely candidates for influence on the audience, yet the cumulative presence of the minor portrayals may also contribute to socialization effects, particularly when consistent patterns are found across large samples of programming.

The V-Chip Study findings, which we again note are based on the broadest sample of program content, establish that, on the whole, television includes more than one scene per hour with a major focus on sexual behavior. Yet while the depiction of sexual behaviors is relatively common on television, the vast majority of the behaviors portrayed are what we term precursory in nature. They consist of physical flirting, passionate kissing, and intimate touch meant to arouse sexual interest. Such actions account for 86-97% of all sexual behaviors shown across each of our studies (see Figure 1).

Example: Precursory sexual behavior.

Rob and Gina, two young adults, meet at a party where they are introduced by their friends. Rob has recently split up from his girlfriend and Gina is unattached and looking to meet someone. They immediately like one another and spend the party talking and laughing together before deciding to leave and go back to Rob's apartment. Once inside, they begin kissing passionately and convey their intent to sleep together when they are interrupted by Rob's ex-girlfriend knocking at the door. (Almost Perfect, CBS, March 3, 1996).

Although scenes with sexual intercourse represent only a modest proportion of all of the sexually-related behaviors shown on television, the presence of intercourse on TV is nonetheless quite substantial. The frequency with which programs included one or more scenes with intercourse depicted or strongly implied held remarkably stable across all four of the data points in the three studies reviewed here, at 12-13% of all shows (see Table 3). In other words, roughly one of every eight shows sampled in each of our studies presented characters engaging in sexual intercourse.

In order for a portrayal to be considered an example of intercourse strongly implied, a scene must depict a couple's actions immediately before or after an act of intercourse that is clearly inferred by narrative device. For example, a couple might be shown passionately kissing as they undress one another in a darkened bedroom, followed by a fade-to-black that then leads to a scene in which the couple awaken in each other's arms the next morning. In contrast, scenes are classified as intercourse depicted if any portion of the body of those engaged in sexual intercourse is shown while the act is occurring. Such depictions need not be explicit in terms of nudity, as discreet camera views may focus close-up on the face, or the upper torso may be shown with the lower body partially covered by a sheet.

Example: Sexual intercourse strongly implied.

Michael and Kimberly, two young lovers, are in bed together, apparently naked but with their bodies discreetly covered by disheveled sheets. He sighs and says, "What a great way to spend an afternoon." She replies, "Do you realize that's all we've been doing lately?" Kimberly tells Michael she is concerned that he might find sex with her boring because he has had so many previous lovers. She asks how they could spice things up. Michael explains a fantasy he would like to act out with him as a plumber and Kimberly as a 'hot-to-trot housewife. They agree to act out the fantasy tomorrow. (Melrose Place, Fox, March 4, 1996).

Scenes in which sexual intercourse is strongly implied represent the most common approach for television stories to convey that love-making has occurred (see Figure 2). Again, the consistency across studies in the findings is remarkably stable, with the range holding between 83-89% of all instances of intercourse on television. This approach seems to be the standard formula for incorporating sexual intercourse within television plots while still leaving most of the details to the viewer's imagination.

IMPORTANT CONTEXTUAL FEATURESThe nature of the relationship between individuals who engage in sexual behaviors (including but not limited to intercourse) on television may offer important socialization lessons for young viewers. Two of our projects, the Family Hour Study and Teen Study, examined this topic. First, our data indicate that most sexually-related behaviors involve unmarried couples (71% for the Family Hour Study; 79% for the Teen Study). Though unmarried, most characters who engage in these behaviors have an established relationship with their partner (67% for the Family Hour Study; 71% for the Teen Study). Nonetheless, relationship infidelity, or cheating, also occurs occasionally, accounting for between 10% (Teen Study) and 17% (Family Hour Study) of the sexual behaviors observed overall.

One of the most potentially important contextual factors is the extent to which the risks or responsibilities associated with human sexual activity are presented on television. The depiction of sexual risk concerns in television content holds obvious potential for raising the salience of these issues in the eyes of viewers, in particular with young audiences who have more limited experience with sexual topics.

Here again the Family Hour Study and Teen Study provide us with data. We measured the presence of three possible types of risk or responsibility themes: (a) sexual patience, waiting until a relationship matures and both people are equally ready to engage in sex; (b) sexual precaution, pursuing efforts to prevent AIDS, STDs, and/or unwanted pregnancy when sexually active; and (c) depiction of risks and/or negative consequences of unplanned or irresponsible sexual behavior. The presence of such themes was measured both at the level of individual scenes as well as across programs as a whole.

Table 4 presents a summary of our findings on the treatment of risk or responsibility concerns amidst all of the sexual messages found on television. In general, the predominant share of sexual messages on television are conveyed without any element of risk or responsibility concerns. Fewer than one of every 10 scenes presenting some sexually-related behavior (8% for the Family Hour Study; 3% for the Teen Study) included any reference at all to these issues. The treatment of risk topics was slightly more common in scenes that contained only talk about sex (10% for the Family Hour Study; 14% for the Teen Study), particularly in the programs most popular with teens.

Several of the examples of risk and responsibility messages found in scenes from teen programming focused on pro-social messages such as waiting until one is ready to have sex.

Example: Sexual patience.

Zachary and his girlfriend Chloe, two high-school-aged teens, are making out on the couch. Zach wants to have sex but Chloe isn't sure that she's ready. He moves his hand under her shirt and she pushes it away. Frustrated, he asks "What is it now?" Chloe replies "A month from now I don't want to be taking a pregnancy test." Zach says he will use protection but Chloe says she's afraid because protection is never 100% effective. A friend of Chloe's recently thought she was pregnant and now Chloe is scared of being in the same predicament. Zach says, "It's OK. I can wait. As long as it takes. I can wait. I don't want you to do something you're not ready to do." (Malibu Shores, NBC, March 30. 1996).

This scene provides a good example of a pro-social message regarding human sexuality. Chloe, recognizing the dangers of becoming sexually active even with the use of birth control, asserts her right to wait until she's ready. Her boyfriend, frustrated at first, respects her wishes.

Across both studies, fewer than one of every ten shows with sexual content (6% for the Family Hour Study; 9% for the Teen Study) placed primary emphasis on a risk or responsibility concern throughout the show. Programs which qualified on this measure addressed such serious, life-altering outcomes of sex as unwanted pregnancy and abortion, as well as persistent anxiety about the prospect of contracting AIDS from unprotected sexual intercourse. Interestingly, however, none of the shows that featured intercourse depicted or strongly implied across both studies (N = 21) included any mention of risk or responsibility concerns at any point in the program.

DISCUSSIONAs with most other aspects of media influence, the effect of viewing sexual content on television is not thought to be direct and powerful, with a single exposure to a particular program leading a viewer to think or act in any given way. Rather, the effects of televised messages about sex are conceptualized more as the product of a slow and cumulative process. Because such influence tends to be gradual in nature, it is the overall pattern of messages across the television landscape that is of primary interest for explaining the effects of long-term exposure.

Across these three studies, we see convergence on a number of key findings. First, sexually-related messages are a common element in television entertainment. The data we have reported here suggest that the odds are better than 50-50 that any given program on television will include some talk about sex. In contrast, the odds are about one in four that a given program will contain a scene devoting primary emphasis to one or more sexual behaviors. While both of these findings support the conclusion that sexual messages are an everyday aspect of television content, they also make clear that talk about sex appears far more often than the portrayal of sexually-related behaviors.

Talk about sex is included in these content studies because of its socialization potential to convey information about sexual norms, practices, motives, and the like even when no overt sexual activity is directly involved. Indeed, research confirms that this talk is meaningful for adolescents in helping to shape their perceptions about sexual matters (Ward, 1999).

Television industry officials sometimes criticize academic studies of sex on TV, asserting that behaviors like flirting or kissing are not what "reasonable people" think of as sex (Farhi, 1998). Of course, just as talk about sex conveys information about sexual relationships, so too may these precursory behaviors that typically precede the act of intercourse in an evolving sexual relationship. However, even if one were to restrict the definition of sex on television to only those instances in which intercourse is included in the program's plot, it can still be concluded that sex on TV is commonplace. Across four different samples encompassing more than 1200 programs, we consistently found that the odds are about one in eight that a given program on television will include sexual intercourse either depicted or strongly implied.

It is this relatively common occurrence of intercourse that makes the issue of risk and responsibility concerns particularly salient. Our data indicate that themes such as sexual patience, sexual precuations, and serious negative consequences from sexual behavior remain highly infrequent on television. This finding is hardly novel, as other studies which have examined these issues have produced similar results (Greenberg & Buselle, 1996; Lowry & Shidler, 1993).

The one encouraging sign in our data is that the programs most popular with adolescent audiences show some indication of addressing these issues more frequently than other shows. Nearly one of every ten programs with sexual content in the Teen Study placed primary emphasis on a risk or responsibility issue throughout the show, and roughly one of every seven scenes of talk about sex included some mention of risk concerns. As portrayals of this type become more prevalent, they would be expected to contribute to greater awareness of the risks of human sexual activity in the eyes of adolescents, and perhaps even the normalizing of preventative behaviors (e.g., use of condoms) when teens engage in intercourse.

Our next step in this research program is to employ the composite week design to sample a broad range of television programming (as was done with the V-Chip Study), and subject it to more extensive analysis for sexual messages, emphasizing risk and responsibility concerns in particular. This study (Kunkel et al., 1999), which is supported by a grant from the Kaiser Family Foundation, will provide a benchmark for tracking sexual content on television over time, and is expected to be repeated on a biennial basis in the future.

In conclusion, sexual content is a common, if not prevalent, aspect of the overall television landscape. Portrayals of talk about sex, as well as sexually-related behaviors, are a potential source of socialization for most young viewers. Although most sexual behaviors shown on television are relatively modest, intercourse is frequently included. Collectively, these sexual messages provide an opportunity for the television industry to communicate an important and realistic view of the true risks associated with human sexual activity. Our research suggests that this opportunity has not yet been tapped very often by most segments of the industry.

Added material.

Dale Kunkel, Kirstie M. Cope, and Erica Biely University of California Santa Barbara.

The Family Hour Study was supported by the Henry J. Kaiser Family Foundation (Menlo Park, CA) and Children Now (Oakland, CA). The Teen Study was the Master's Thesis for Kirstie M. Cope. The V-Chip Study was supported by the Henry J. Kaiser Family Foundation. The authors wish to thank Carolyn Colvin, Ed Donnerstein, Wendy Jo Farinola, Ulla Foehr, Jim Potter, Vicky Rideout, and Emma Rollin, each of whom made significant contributions to one or more of the studies summarized here.

Address correspondence to Dr. Dale Kunkel, Department of Communication. University of California Santa Barbara, Santa Barbara, CA 93106; e-mail: kunkel@alishaw.ucsb.edu.

Table 1. Summary of Sample Characteristics Across Studies.

(TABLE) Family Hour Study Teen Study V-Chip Study: Composite Week V-Chip Study: Prime-TimeSample design First hour of prime-time Teens Top 15 shows Composite Week (7 a.m. - 11 p.m.) Prime-time (8 - 11 p.m.)Channels Commercial broadcast Commercial broadcast Full range of broadcast Commercial broadcast networks networks and cable channels networksYear shows aired 1996 1996 1997-1998 1997-1998Number of programs 182 45 840 245Number of hours 84 28.5 828 209.5.

Note: V-Chip Study excluded news and sports programs because these categories are exempted from the TV industry's rating system.

Table 2. Frequency and Distribution of Talk About Sex.

(TABLE) Family Hour Study Teen Study V-Chip Study: Composite Week V-Chip Study: Prime-Time Frequency of Talk About SexPercentage of programs with any talk about sex 59% 67% 57% 67%Average scenes per hour n/a 2.8 2.3 2.9Total number of programs 128 45 840 245 Distribution of Types of Talk About SexTalk about prospective interests 59% 62% 74% 72%Talk about past sexual experiences 21% 20% 18% 21%Expert advice/technical info 5% 6% 1% 0%Other 15% 12% 7% 6%Number of cases 261 95 2203 722.

Table 3. Frequency of Sexual Behavior Within Programs.

(TABLE) Family Hour Study Teen Study V-Chip Study: Composite Week V-Chip Study: Prime-TimePercentage of programs with any sexual behavior 61% 62% 28%(FNa) 28%(FNb)Average scenes per hour with any sexual behavior n/a 2.2 1.2 1.1Percentage of programs with sexual intercourse depicted or strongly implied 12% 13% 12% 12%Total number of programs 128 45 840 245.

FOOTNOTESa Family Hour Study and Teen Study included major as well as minor depictions of sexual behavior. The V-Chip Studies reported only major depictions. When minor depictions are included, this statistic increases to 42%.

b When minor depictions are included, this statistic increases to 44%.

Table 4. Treatment of Sexual Risk or Responsibility Concerns.

(TABLE) Family Hour Study Teen StudyPercentage of talk-onlyscenes with risk orresponsibility 10% 14%Total number of talk scenes 182 80Percentage of behaviorscenes with risk or responsibility 8% 3%Total number of behavior scenes 267 99Percentage of programs with sexualcontent that place primary emphasison risk or responsibility 6% 9%Total number of programs 128 45.

Figure 1. Percentage of Sexual Behaviors That are Precursory.

Figure 2. Use of Different Approaches for Portraying Intercourse.

FOOTNOTE1 The Close Interval Around the Agreement Mode (CIAM) is a technique for assessing reliability of media content coding when large numbers of coders are employed to make conceptually-based unitizing judgments at multiple levels of analysis (Potter et al., 1998). The Agreement Mode refers to the number of identical scenes within a program that the greatest number of coders recognize as containing the relevant content, which in this case is sexual talk or behavior. The Close Interval is defined as +/- 1 from the Agreement Mode for most programs. This technique was devised for use in the National Television Violence Study (Wilson et al., 1997). which explicates the development of the procedures in much greater detail. This approach reflects the most current methodological innovation for calculating reliability across large numbers of coders who are rendering content-based judgments at multiple levels of analysis.

REFERENCESBrown, J. D., Childers, K. W., & Waszak, C. S. (1990). Television and adolescent sexuality. Journal of Adolescent Health Care, 11, 62-70.

Brown, J. D., & Newcomer, S. F. (1991). Television viewing and adolescents' sexual behavior. Journal of Homosexuality, 21, 77-91.

Bryant, J., & Rockwell, S. C. (1994). Effects of massive exposure to sexually oriented prime-time television programming on adolescents' moral judgment. In D. Zillman, J. Bryant, & A. C. Huston (Eds.), Media, children, and the family: Social scientific, psychodynamic, and clinical perspectives (pp. 183-195). Hillsdale, NJ: Lawrence Erlbaum.

Cope, K. M. (1998). Sexually-related talk and behavior in the shows most frequently viewed by adolescents. Unpublished master's thesis, University of California, Santa Barbara.

Courtright, J. A., & Baran, S. J. (1980). The acquisition of sexual information by young people. Journalism Quarterly, 1, 107-114.

Farhi, P. (1998, October 21). Most new TV ratings missing, study finds. Washington Post, p. B7.

Greenberg, B. S., Brown, J. D., & Buerkel-Rothfuss, N. L. (1993). Media, sex, and the adolescent. City, NJ: Hampton Press.

Greenberg, B., & Buselle, R. (1996). Soap operas and sexual activity: A decade later. Journal of Communication, 46, 153-160.

Huston, A. C., Wartella, E., & Donnerstein, E. (1998). Measuring the effects of sexual content in the media. Menlo Park, CA: Kaiser Family Foundation.

Hyde, J. S., & DeLameter, J. (1997). Understanding human sexuality (6th ed.). New York: McGraw Hill.

Kaiser Family Foundation (1998). Kaiser Family Foundation and YM Magazine national survey of teens: Teens talk about dating, intimacy, and their sexual experiences. Menlo Park, CA: author.

Kirby, D. (1997). No easy answers: Research findings on programs to reduce teen pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy.

Kunkel, D., Cope, K. M., & Colvin, C. (1996). Sexual messages on family hour television: Content and context. Menlo Park, CA: Kaiser Family Foundation.

Kunkel, D., Cope, K. M., Farinola, W. J. M., Biely, E., Rollin, E., & Donnerstein, E. (1999). Sex on TV: Content and context. Menlo Park, CA: Kaiser Family Foundation.

Kunkel, D., Farinola, W. J. M., Cope, K. M., Donnerstein, E., Biely, E., & Zwarun, L. (1998). Rating the TV ratings: An assessment of the television industry's use of V-chip ratings. Menlo Park, CA: Kaiser Family Foundation.

Lowry, D. T., & Shidler, J. A. (1993). Prime-time TV portrayals of sex, "safe sex" and AIDS: A longitudinal analysis. Journalism Quarterly, 70, 628-637.

Nielsen Media Research (1998). 1998 report on television. New York: author.

Office of National AIDS Policy (1996, March). Youth & HIV/AIDS: An American agenda. Washington, DC: author.

Peterson, J. L., Moore, K. A., & Furstenberg, F. F. (1991). Television viewing and early initiation of sexual intercourse: Is there a link? Journal of Homosexuality, 21, 93-119.

Potter, W. J., Linz, D., Wilson, B., Kunkel, D., Donnerstein, E., Smith, S., & Blumenthal, E. (1998). Content analysis of entertainment television: New methodological developments. In J. Hamilton (Ed.), Media violence and public policy (pp. 55-103). Ann Arbor, MI: University of Michigan Press.

Ward, L. M., & Rivadeneyra, R. (1999). Contributions of entertainment television to adolescents' sexual attitudes and expectations: The role of viewing amount versus viewer involvement. The Journal of Sex Research, 36, 237-249.

Wilson, B., Kunkel, D., Linz, D., Potter, W. J., Donnerstein, E., Smith, S., Blumenthal, E., & Gray, T. (1997). Violence in television programming overall: University of California Santa Barbara study. In National Television Violence Study: Vol. 1. Thousand Oaks, CA: Sage Publications.

Manuscript accepted April 14, 1999.