Introduction

Television exposure has been identified as a potentially important influence on the sexual socialization of children and adolescents (Collins et al. 2003, 2004; Ward and Friedman 2007). With the average American youth exposed to nearly 14,000 sexual references, innuendos, and behaviors on television annually (Strasburger 2004), television programming offers many opportunities for observational learning. Recent longitudinal findings for 12- to 17-year olds indicate that adolescents who viewed more sexual content on television were more likely than lighter viewers to initiate sexual intercourse or engage in more advanced non-coital sexual activities in the subsequent year (Collins et al. 2004). Research conducted to date has focused primarily on media influences on sexual attitudes and sexual behaviors, with little attention paid to contextual factors associated with adolescent sex such as partner relationship status. The current paper aims to address this gap in the research by ascertaining whether and to what extent television exposure is associated with two relationship contexts: committed/romantic (or exclusive) versus uncommitted/non-romantic (or casual) sexual partnerships.

Relationship Contexts Depicted on Television

Content analyses suggest that youth may be receiving unrealistic messages about sex from the media. Sexual behaviors are presented in a largely favorable light, and are seldom accompanied by references to risks or negative consequences such as pregnancy or sexually transmitted infections (STIs) (Baxter et al. 1985; Brown and Steele 1995; Buerkel-Rothfuss 1993; Fisher et al. 2004; Kunkel et al. 1999, 2001, 2003; Lowry and Towles 1989). With regard to relationship context, sex on television occurs 4–8 times more often between unmarried persons than married couples (Brown and Newcomer 1991; Lowry and Shidler 1993). A recent content analysis of a large sample of programming likely to appeal to youth found that sexual intercourse not only occurred more often between unmarried persons than married persons at a rate of 5–1, but in nearly 60% of all portrayals of intercourse the couple was in a casual rather than a committed romantic relationship (Fisher et al. 2004).

From a theoretical perspective, Bandura’s social cognitive theory suggests that youth may use such media portrayals to learn new behaviors if the models are attractive, identified with, or their behaviors are rewarded (Bandura 1977). To the extent that youth—who are relatively inexperienced in sexual matters yet are beginning to enter into romantic relationships—are particularly likely to use the sexual scenarios portrayed by the media as scripts to guide their own sexual practices, such depictions may lead them to adopt norms favorable to non-exclusive or uncommitted sex. Similarly, cultivation theory suggests that heavy viewers of television incorporate attitudes and behaviors portrayed on screen as representative of the real world (Gerbner and Gross 1976; Hawkins and Pingree 1983). For example, previous research indicates that heavy viewers of soap operas report a greater perceived incidence of divorce and illegitimate children, common themes found in soap operas, than lighter viewers (Carveth and Alexander 1985). It is reasonable to propose that youth who have greater exposure to casual sex, through discussion, allusion, or behaviors among television characters, are more likely to view this behavior as normative and therefore may adopt attitudes and values that make participation in this behavior more likely compared to peers with lower exposure.

Casual Sex and Youth

Several large national studies indicate that youths’ sexual experiences frequently occur outside of established relationships. For example, nearly one-quarter (23%) of adolescent girls who participated in the 1995 National Survey of Family Growth indicated that they had their first sexual experience with someone they had just met, with whom they were “just friends,” or with whom they had gone out “once in a while” (Elo et al. 1999). Another study found that although over a 1-year period most sexually active teenagers (62%) reported having a sexual experience only within a romantic relationship, 24% reported having had only a non-romantic sexual partner, and 14% reported having had both a romantic and non-romantic partner (Manning et al. 2005). A random sample of undergraduate students found that 48% had engaged in sexual intercourse with a stranger or brief acquaintance in college and 78% had some sexual interaction (without sexual intercourse) with a stranger in college (Paul et al. 2000).

A small number of studies have examined associations between casual partnerships and attitudes, normative beliefs, intentions and behavior (Herold et al. 1998; Levinson et al. 1995; Maticka-Tyndale et al. 1998; Paul et al. 2000). One recent study assessed predictors of non-romantic sexual behaviors over a 1-year period among adolescents aged 15 and older. A multinomial analysis found that being male, black, having parents with more than 16 years of education, and prior sexual experience increased the likelihood of having sex with a non-romantic versus romantic partner. Alternatively, adolescents who were older and had a higher level of self esteem were less likely to have sex with a non-romantic partner than a romantic partner (Manning et al. 2005).

The distinction between exclusive romantic versus casual sexual relationships is important given that adolescents engaging in sex within casual partnerships may be at increased risks of disease transmission due to involvement with a greater number of sexual partners, lack of knowledge of their partners’ sexual or drug use history, and less frequent use of condoms (Bender and Kosunen 2005; Crosby et al. 2000; Manlove et al. 2004; Manning et al. 2000; Misovich et al. 1997; Rosengard et al. 2005). Results from one national study indicated that young people who had engaged in a one-night stand (one-time sex partner) were significantly more likely to report an STI than those who had not had a one-night stand (Tanfer et al. 1995). Aside from increased risks for negative health consequences, early sexual experiences may have important implications for socio-emotional outcomes (such as correlations with self-esteem) and establishing of patterns of sexual behavior that extend into adulthood.

Current Investigation

Given that adolecents watch an average of 20.5 hours of television a week (Roberts et al. 2005), and current research suggests that most sexual encounters between television characters often occurs in a casual relationship (Fisher et al. 2004), the current longitudinal study will examine whether and to what extent exposure to various types of high sexual content television programming predicts relationship status (exclusive versus casual) at last sexual encounter. This study is one of the first to investigate the association between the media and adolescent sexual relationship contexts. In addition, the current study expands the focus beyond vaginal intercourse and examines these predictors in relation to partner relationship status for oral sex.

Method

Participants

This study is based on data from the second and third annual waves of a survey funded by the National Institute of Child Health and Human Development (NICHD) to study media exposure and adolescent sexual behavior. A list-assisted sample of households from 10 California counties was used to recruit study participants aged 12–16 at Wave 1. Potential participants and their parents were first contacted through a mailed letter and fact sheet that described the study and invited them to take part. The estimated response rate at Wave 1 was 75%. The cooperation rate (N completed interviews/N known eligible numbers) was considerably higher (88%).

Of the 1,105 youth participants in Wave 1, 1,012 (92%) completed the annual survey at Wave 2, and 932 (84%) participants completed the survey at Wave 3. Written parental consent was obtained for all respondents as per the approved institutional review board protocol. Data were collected using in-home computer assisted self interviews (CASIs) that averaged between 25 and 35 min in length. The CASIs were administered to adolescents in a private setting. While the youth completed the CASI, one parent was asked to fill out a paper-and-pencil survey.

A total of 824 adolescents had complete data for the variables examined in the current study; 68 cases (7%) were missing data for key variables and were deleted from the analyses. A review of the data found that a majority of the missing values were linked to one variable, parents’ failure to complete the item on their annual income on the parent survey. The average age of the sample at Wave 3 was 15.02 (SD = 1.43) with 50.1% (n = 440) males and 49.1% (n = 424) females. The sample comprised Latinos (8.9%), African Americans (5.8%), Asian Americans (5.7%), European Americans (72.3%), and individuals of other ethnicities (7.6%).

Measures

Sexual Behavior (T1 and T2)

Adolescents’ participation in oral sex and vaginal intercourse was measured with two dichotomous (yes/no) items. Oral sex was measured by asking respondents, “Have you ever had oral sex with a girl/boy (When a girl/boy puts her/his mouth or tongue on your genitals or you put your mouth or tongue on a girl’s/boy’s genitals)?” Vaginal intercourse was measured by asking participants, “Have you ever had sexual intercourse? By sexual intercourse, we mean when a boy puts his penis into a girl’s vagina.”

Relationship Status (T2)

Adolescents who indicated that they had engaged in vaginal intercourse and/or oral sex were asked about their relationship status with their most recent opposite-sex sexual partner. Specifically, respondents who had participated in oral sex were asked, “Thinking about the last time you had oral sex (with a boy/girl), how would you describe the relationship you had at the time (please make your best guess).” Response options included: (a) We were in an exclusive romantic relationship (not seeing or dating other people), (b) We were in an open romantic relationship (could see or date other people), (c) We knew each other in a non-romantic way (such as friends, classmates, neighbors, or other acquaintances), and (d) We had just met. Adolescents who had experienced vaginal intercourse were similarly asked about their relationship status at last sexual intercourse.

Overall Television Exposure (T1)

Adolescents were asked two questions to assess their weekly television viewing habits: “In general, on Mondays through Thursdays, about how much television do you usually watch each day?” and “On Fridays through Sundays, about how much television do you watch each day?” Response options ranged from “none” (1) to “more than 8.5 h” (7). Responses were recoded to the category midpoints (e.g., a respondent who indicated watching television “5–6.5” h per week was given a score of 5.75 h). The responses to these two items were weighted by 4 and 3, respectively, and summed to calculate “total number of hours of television watched per week”.

Exposure to Sexually Suggestive Cable Programming (T1)

Respondents were also asked a series of items that measured their exposure to several types of cable programming were known to be sexually provocative (Fisher et al. 2004).

Music Videos

Respondents were asked how many hours they watched televised rock, rap, or hip-hop music videos like those on MTV, BET or VH-1 in a typical week [Don’t watch any of these (1) to More than 21 h a week (8)]. Responses were coded into weekly hours using category midpoints.

Premium Cable

Respondents were asked how many hours they watched HBO, Showtime, or Cinemax in a typical week using the same scale as for music videos.

Adult Programming

Respondents were asked how many hours in a typical week they watched adult cable channels such as Playboy or Spice or adult Pay-Per-View movies using the same scale and coding as for music videos.

Frequency of viewing sexually suggestive broadcast programming (T1)

The CASI presented respondents with video clips from eight current fictional prime time shows on network television. The eight shows were selected because they were popular among 12–17 year olds based on television ratings provided by Nielsen Media Research and were high in sexual content based on a separate content analysis conducted a few months prior to the survey (e.g., Fisher et al. 2004). The eight series used in the Wave 2 CASI were Scrubs, Will and Grace, Friends, That 70’s Show, According to Jim, One on One, My Wife and Kids, and Boston Public. If respondents correctly identified what show a clip came from, they were asked how often they watched the show. Those who did not recognize the clip and those who said they had never watched the show all the way through were given a 0 viewing frequency score for that show and skipped over follow-up questions about it. Viewers were asked how frequently they had watched the show during the past 6 months using a four-point scale from “not at all” (1) to every time it’s been on (4). Frequency of viewing high sexual content prime time broadcast shows was the sum of the frequency values for each of the 8 shows. Therefore, this variable had a potential range of 0–32.

Background Variables (T1)

Respondents were asked their current age, gender, and ethnicity. They were allowed to select from among eight racial/ethnic categories and more than one category could be selected. Because of the small numbers of ethnic minorities in the sample, ethnicity was dummy coded into a single variable White non-Hispanic (1) or Non-White (0). We also control for parent-reported annual household income which correlates highly with parental education, a predictor of involvement in non-romantic sexual partnerships.

Analyses

Because of the small number of cases in some of the relationship context categories (e.g., only 12 adolescents indicated having had vaginal intercourse with someone they had just met), the relationship types open romantic, non-romantic, and someone just met were combined into a single category, non-exclusive/non-romantic (or casual), for the primary analyses. We first conducted separate ANOVAs to determine whether media use at time 1 differed across the three groups of youth—sexually active/causal relationship, sexually active/exclusive relationship, and non-sexually active youth (measured at time 2). Next we conducted a series of hierarchical multinomial logistic regressions to ascertain which types of sexually explicit television exposure at time 1 uniquely predicted being in (1) an exclusive relationship at last oral sex at time 2 and (2) an exclusive relationship at last sexual intercourse at time 2. Model 1 for both analyses included sexual experience at time 1, age, gender, parental income, and ethnicity, with non-Whites and females serving as the reference groups for ethnicity and gender, respectively. We added all of media variables in Model 2. Chi-square difference tests were used to evaluate improvement in fit for successive models. To allow us to directly contrast youth from all three groups to one another, each analysis was executed twice with the only difference being the selected reference group.

Results

Television Use and Adolescent Sexual Status and Sexual Partnership

A total of 344 adolescents indicated having had oral sex at time 2. Of those, 63.2% indicated that this sexual behavior took place most recently within the context of an exclusive romantic relationship. About a tenth of the sample (10.5%) last experienced oral sex within an open romantic relationship, 18.2% last had oral sex with someone they knew in a non-romantic way, and 8.0% last had oral sex with someone they had just met. Fewer adolescents indicated having ever had vaginal intercourse (n = 269). Of those, 62.6% had last had sexual intercourse within an exclusive romantic relationship, 14.2% within an open romantic relationship, 16.5% with someone they knew in a non-romantic way, and 6.6% with someone they had just met.

Regarding media exposure, adolescents who had no vaginal sex experience watched significantly fewer hours of music videos, premium cable, adult cable, and sexually suggestive broadcast television shows than adolescents who reported being in a causal partnership at last vaginal intercourse (see Table 1). Adolescents who were in a committed relationship at last vaginal intercourse reported watching significantly less premium and adult cable than those who reported being in a casual partnership at last sexual intercourse. However, adolescents who had no sexual experience reported significantly more television viewing overall and less exposure to music videos than those who were in a committed relationship at last vaginal sex. A similar pattern emerged when examining oral sex and television use (see Table 2).

Table 1 Mean differences in media exposure across sexual experience and relationship status, vaginal intercourse, time 2
Full size table
Table 2 Mean differences in media exposure (T1) across sexual experience and relationship status, oral sex, time 2
Full size table

Results from a multinomial logistic regression found that overall television use, music video, premium cable, and adult cable exposure were significant predictors of adolescent sexual and relationship status 1 year later, controlling for previous sexual experience and background demographic characteristics. Specifically, adolescents who had not had vaginal intercourse in the past year were more likely to report having watched more television overall at time 1 than youth who were in a committed relationship at last vaginal intercourse (OR = 1.03, p < .01) and more television than youth who were in a casual relationship (OR = 1.03, p < .01; not shown in Table 3). Overall television use did not differ between youth who were in a committed relationship at last vaginal intercourse and those who were in a casual relationship at last vaginal intercourse.

Table 3 Logistic regression predicting factors associated with being in an exclusive relationship versus non-exclusive relationship at last vaginal intercourse
Full size table

Adolescents who had not had sexual intercourse in the past year were significantly less likely to have been exposed to music television (OR = .92, p = .02) and adult television (OR = .34, p < .01) at time 1 than adolescents who reported being in an exclusive relationship at last vaginal sex. Similar findings were observed when comparing adolescents who had not had sexual intercourse in the past year and adolescents who were in a casual relationship at last vaginal sex (Music television: OR = . 91, p < .01; adult television: OR = .30, p < .01; not shown in Tables 3 and 4).

Table 4 Logistic regression predicting factors associated with being in an exclusive relationship versus non-exclusive relationship at last oral sex
Full size table

The only media exposure measure that differentially predicting being in a casual versus a committed relationship for last sexual intercourse was premium cable TV. Youth who reported being in a casual relationship at last vaginal intercourse at time 2 were more likely to have been exposed to more premium cable television at time 1 than youth who reported being in a committed relationship at last vaginal intercourse at time 2 (OR = 1.09, p < .05). A similar pattern emerged when examining relationship status at last oral sex and media exposure measures.

Discussion

This investigation is one of the first studies to include media exposure among the broad range of possible correlates to sexual partner status. Results of the study partially support the proposed hypotheses. First, a series of ANOVAs found that media use within specific genres known to have relatively high sexual content differed significantly among adolescents who had not had sexual intercourse, those who had engaged in sex with an exclusive partner, and those who had engaged in sex with a casual partner, with respect to last oral sex and last vaginal intercourse separately. Multinomial logistic regressions confirmed these findings. Specifically, when comparing youth who had had sexual intercourse, exposure to greater amounts of premium cable programming was predictive of having been in a casual partnership at last vaginal sex. Although not significant, the results also suggest a trend in which greater exposure to premium cable and adult programming may also be predictive of being in a casual relationship at last oral sex.

Overall these findings suggest that exposure to specific genres of television influence the context within which sexual behavior occurs. Although additional research should be conducted, the results highlight the need for prevention programs to address not only specific sexual behaviors and contraceptive use, but also the context within which these behaviors occur as context may also increase risk. Second, parents are in the unique position of engaging in television co-viewing with their child and perhaps having the opportunity to deflect the effects of sexually suggestive media on casual relationships by discussing potential physical and emotional consequences that many television shows fail to include.

In particular, the similarity in the patterns of media exposure effects across both vaginal intercourse and oral sex is worth noting. In content analysis studies, relationship status has typically been examined with regard to sexual intercourse exclusively and portrayals of oral sex on television have been found to be rare. The current findings, however, suggest that the influence of sexually explicit portrayals of one behavior may be generalized to other sexual behaviors, especially those at the higher end of the sexual intimacy continuum that are not often depicted, yet may pose sexual health risks.

Despite the novelty of the findings, limitations of the current study should be noted. First, relationship context was assessed only for the last event within each sexual behavior, thus providing a snapshot of the possibly shifting and diverse nature of adolescents’ sexual partnerships. To the extent that these last encounters are atypical and do not reflect the usual relationship context for sex, our findings about predictors of relationship status may underestimate the true effects. Second, the study was conducted in California with a sample that was largely white and middle class. Thus, it is unknown whether and to what degree the results are generalizeable to other youth populations. Nonetheless, a number of important and significant relations were found and the current study serves as a starting point.

It is important to note that although casual sexual relationships may be associated with greater health risks, being in a committed or exclusive sexual relationship does not eliminate these risks. It is possible that adolescents in committed sexual relationships engage in sexual behaviors more often. Adolescents in committed relationships may also be more likely to use forms of contraception that do not reduce the transmission of STIs. In addition, some adolescents may have a series of short duration (i.e., a week, a month) “committed” sexual relationships with many different partners. These adolescents engaging in high frequency serial monogamy may be at greater risk for STIs than adolescents who have a small number of sexual partners outside of committed relationships. Finally, incidences of infidelity within committed relationships are common among youth (Feldman and Cauffman 1999a, b), though understudied as a phenomenon. Thus, many adolescents who believe that they are in an exclusive relationship may be exposed to risks of casual sex due to a partner’s behavior about which they are unaware.

Research on the relationship context for adolescent sex is formative and thus much remains to be learned about this important issue. Casual sex has been operationalized in many ways, from a non-committed sexual relationship among friends, to a sexual encounter between strangers (Herold et al. 1998; Paul et al. 2000). Young adults and researchers alike may define casual sex differently depending on the behavior(s) that occur within a sexual encounter, perceived intimacy, commitment level, and/or the length of the relationship. Future research should examine how adolescents define committed and casual relationships, the motivations of adolescents who engage in casual sexual encounters, and the physical and psychological consequences of such encounters.

Regarding media exposure, although the current study focused on sexually suggestive television content, it is also necessary to assess the impact of unregulated sexual content (via the internet or mobile device) on the formation of adolescents’ beliefs and behaviors around casual encounters. Additionally, future research should focus on potential gender and age differences. It may be that older adolescents interpret sexually explicit media in a differential manner than younger youth, which in turn affects sexual behavior. A deeper understanding of the context of adolescent sexual risk taking can help guide efforts to educate youth about the consequences of such behavior and contribute to the design of interventions to reduce the negative consequences associated with adolescent risky sexual behavior.