URGENT - NEED ASSISTANCE W/ DISCUSSION POST ON DIVERSITY - DUE TOMORROW

Sexual References and Consequences for Heterosexual, Lesbian, Gay, and Bisexual Characters on Television: A Comparison

Content Analysis

Bradley James Bond Department of Communication Studies

University of San Diego

Brandon Miller Department of Communication

University of Massachusetts Boston

Jennifer Stevens Aubrey Department of Communication

University of Arizona

A content analysis was conducted to examine sexual references and consequences among lesbian, gay, bisexual (LGB), and heterosexual characters on television. The sample was composed of programs portraying an LGB lead or reoccurring character.

Bradley James Bond (Ph.D., University of Illinois, 2011) is an associate professor in the Department of Communication Studies at the University of San Diego. His research interests include parasocial relationships and media effects related to sex and sexuality.

Brandon Miller (Ph.D., University of Missouri, 2016) is an assistant professor in the Department of Communication at the University of Massachusetts Boston. His research interests include the use and effects of mass and interactive media among LGBTQ individuals and mainstream audiences.

Jennifer Stevens Aubrey (Ph.D., University of Michigan, 2004) is an associate professor in the Department of Communication at the University of Arizona. Her research interests include media effects on young people's sexual socialization and body image.

Correspondence should be addressed to Bradley James Bond, Department of Communication Studies, University of San Diego, Camino Hall 126-C, 5998 Alcala Park, San Diego, CA 92110. E-mail: bond@sandiego.edu

Mass Communication and Society, 22:72–95 © 2018 Mass Communication & Society Division of the Association for Education in Journalism and Mass Communication ISSN: 1520-5436 print / 1532-7825 online DOI: https://doi.org/10.1080/15205436.2018.1489058

72

Results showed that heterosexual and LGB characters engaged in sexual talk and behavior in similar contexts. When discussing LGB sexualities, however, heterosexual characters were disproportionately likely to make jokes; LGB characters were dispro- portionately likely to discuss coming out. LGB characters depicted in sexual references were more likely to be in dialogue with a heterosexual character than another LGB character. Sexual consequences were more common for heterosexual characters than LGB characters. No gender differences existed in frequency of sexual references or consequences among LGB characters, evidence that the sexual double standard found in previous research may not apply to LGB characters. Results are discussed in terms of potential effects of exposure.

Communication scholars have long been interested in the portrayal of sex on television given the considerable amount of time that individuals spend watching television and the empirical evidence that suggests television exposure is among the socialization agents that can influence audiences’ sexual attitudes, beliefs, and behaviors (Strasburger, 2012). One contextual element of sex on television that has received increasing attention is the depiction of sexual consequences. The study of sexual consequences is often justified by social cognitive theory (Bandura, 2009), which argues that audiences are more likely to model observed behaviors if they are rewarded than if they are punished. According to the theory, depicting consequences to sexual references on television could alter messages’ influence on viewers. Content analyses of sexual consequences on television consistently conclude that depictions of sexual consequences are infrequent even though sexual references are pervasive, negative consequences generally appear with greater frequency than positive consequences, and emotional or social consequences are more likely to be depicted than physical con- sequences (Aubrey, 2004; Eyal & Finnerty, 2009; Kunkel et al., 2007; Ortiz & Brooks, 2014). Although these studies are important contributions to understanding the scope of sexual content on television, they are consistently heteronormative.

The foundational Sex on TV studies (e.g., Kunkel, Eyal, Finnerty, Biely, & Donnerstein, 2005) make no mention of lesbian, gay, or bisexual (LGB) char- acters, nor do any of the studies just cited. The absence of such research is problematic given that LGB youth seek out LGB-inclusive media messages (Bond, Hefner, & Drogos, 2009), depictions of LGB characters are increasingly common on television (Bond, 2014; GLAAD, 2016), and LGB youth may be particularly likely to be influenced by media depictions of sex because they lack other sources of information about LGB sex and sexualities (Winderman & Smith, 2016). Moreover, heterosexual audiences can learn about LGB sexualities from televised depictions of LGB individuals (Schiappa, Gregg, & Hewes, 2005). The objective of the present study was to better contextualize the depic- tion of LGB sexual references on television by using content analytic methods to

SEXUAL REFERENCES AND CONSEQUENCES ON TV 73

investigate sexual references and sexual consequences for heterosexual and LGB television characters on programs inclusive of LGB sexualities.

LGB SEXUALITIES ON TELEVISION

LGB characters accounted for 5% of all scripted series regulars on broadcast, cable, and streaming services from June 2016 to May 2017, representing the highest percentage of LGB characters in the history of television (GLAAD, 2016). Although LGB characters are more visible on television than ever, the uniformity of LGB depictions is potentially problematic. LGB characters are often portrayed in heteronormative situations or in highly stereotypical contexts (Streitmatter, 2009).

Few studies have empirically investigated the sexual references made by LGB characters on television. Fisher, Hill, Grube, and Gruber (2007) used a coding scheme that was adapted from Kunkel’s Sex on TV content analyses (e.g., Kunkel et al., 2005) to analyze the portrayal of LGB characters in a 3-week composite sample of prime-time programs from 11 networks. LGB characters engaging in sexual talk appeared in 13% of programs, a significant discrepancy from the 70% of programs containing heterosexual sexual talk reported in the Sex on TV studies (e.g., Kunkel et al., 2005). Bond (2014) also used definitions stemming from the Kunkel et al. (2005) study in his analysis of LGB depictions on television shows popular with LGB youth. LGB talk was added as a variable in this study because LGB sexualities may be discussed in contexts that would not be coded using Kunkel et al.’s (2005) definitions. Results from Bond’s content analysis sug- gested that the frequency of television programs depicting LGB sexual behavior had gone unchanged since Fisher et al.’s analysis but that sexual talk among LGB characters had appeared in an increasing number of television programs. However, when LGB individuals were depicted talking about sex, it was most likely to be about their sexual identities, not about their sexual desires, activities, or experiences. The most common type of LGB talk was insults/jokes/mocking references sourced by heterosexual characters. Sexual behavior was rarely por- trayed between LGB characters. The humorous depictions of LGB sexual talk combined with the rarity of LGB sexual behaviors were used to justify the claim that LGB sexualities on television are sanitized to appease heterosexual audi- ences. As Bond concluded, “LGB characters in media are sexual only by proclamation, void of any sexual expression” (p. 117).

LGB characters are appearing on television with increasing regularity, but their sexualities may be more ornamental than authentically integrated compo- nents of their characterizations. Although no study to date has examined the depiction of sexual consequences for LGB characters on television, the conclu- sions from previous studies that have examined sexual consequences for

74 BOND, MILLER, AUBREY

heterosexual characters may lend insight into the potential similarities and differences between television’s portrayal of LGB and heterosexual characters.

Sexual Consequences on Television

Sexual references among heterosexual characters are prevalent on American televi- sion, but consequences of sex are far less frequent (Aubrey, 2004; Eyal & Finnerty, 2009; Kunkel et al., 2005; Ortiz & Brooks, 2014). Kunkel and colleagues (2005) coded “risks and responsibilities” of sex, which they defined as depictions of safe sex issues or negative consequences resulting from sexual activity. Among shows with sexual content, only 14% had references to risks and responsibilities. Aubrey (2004) broadened the definition of sexual consequences to include emotional and social consequences of sex. Approximately 30% of scenes containing sexual instances also contained sexual consequences in her analysis, of which 75% were emotional or social. Sexual consequences were significantly more likely to be negative than positive, a finding reinforced by more recent research (Ortiz & Brooks, 2014). A study examining only sexual intercourse also found that emotional consequences far outweigh physical consequences and that consequences to inter- course tend to be positive in valence but that they are also formulaic and that “television largely ignores the complex outcomes of sexual intercourse” (Eyal & Finnerty, 2009, p. 163).

When sexual consequences are depicted on television, the gender of the characters involved in the narratives can influence the context of the consequences. Although men are more likely to initiate sexual references on television than women, women are more likely to receive negative consequences when they initiate sexual references (Aubrey, 2004). Women experience more diverse sexual consequences and are more likely to be the victims of nonconsensual sexual relations than men (Eyal & Finnerty, 2009). This gender discrepancy has been referred to as television’s own sexual double standard, suggesting that gender plays a significant role in determining the style with which sexual consequences are depicted on television (Aubrey, 2004). Whether LGB characters face sexual consequences in a similar fashion to heterosexual characters, or whether the sexual double standard transgresses sexual orientation of the characters, is unknown. Investigating sexual consequences of LGB characters not only would shed light onto the context of LGB characters on television but also could provide a groundwork for further research examining the effects of exposure to LGB characters’ sexual references and consequences.

EFFECTS OF EXPOSURE

Theories used to predict and explain media effects have demonstrated the potential influence of television’s depiction of sexual consequences on viewers’ sexual behaviors. The predominant theory utilized to explain these potential effects is social

SEXUAL REFERENCES AND CONSEQUENCES ON TV 75

cognitive theory. According to social cognitive theory, individuals can learn not just through direct experience but also by observing the behaviors of others in their immediate environment (Bandura, 2009). Not every behavior observed is subse- quently modeled. The consequences of observed behaviors are critical to predicting the likelihood of modeling. Punishments associated with observed behaviors lessen the likelihood that acts will be modeled; rewards associated with observed behaviors strengthen the likelihood. Social cognitive theory is particularly explanatory when applied to a television context: viewers of television models who see behaviors rewarded are more likely to acquire the observed behaviors. The theory has been used to frame the importance of understanding the context of televised depictions of sex in several content analysis studies (Aubrey, 2004; Eyal & Finnerty, 2009; Ortiz & Brooks, 2014). The predictions of social cognitive theory are also supported by studies examining exposure to on-screen sexual behavior.

Eyal and Kunkel (2008) found that posttest attitudes toward premarital sex were significantly more negative than pretest attitudes after young adults viewed television dramas that portrayed the negative emotional consequences of sexual intercourse. In a similar study, Finnerty (2007) found that previous experience moderated the influence of exposure to sexual consequences. Participants who had little to no past sexual experience reported safer sex outcomes after exposure to televised negative sexual consequences. No such effect was found for parti- cipants with extensive past sexual experience. Together these findings support the premise of social cognitive theory by suggesting that associations created between sexual behaviors and their negative consequences lead to negative dispositions toward sexual behaviors. Although no studies have examined how sexual consequences on television influence the sexual behaviors of LGB audi- ences, social cognitive theory could be applied to predict similar outcomes. Thus, a stronger empirical understanding of the depictions of sexual consequences among LGB characters on television would prove useful in studying potential effects of exposure.

Exposure to LGB characters in sexual situations could also alter heterosexual viewers’ beliefs about LGB sexualities. Cultivation theory predicts that heavy television viewers will, over time, develop attitudes and beliefs that mirror those presented on television. The world as depicted on television becomes symbolic for reality (Morgan, Shanahan, & Signorielli, 2009). Thus, exposure to LGB characters in sexual situations could inform the worldviews of heterosexual viewers and potentially alter the beliefs that heterosexual audiences have about LGB sexualities.

The parasocial contact hypothesis also predicts that heterosexual viewers may be influenced by depictions of LGB characters on television (Schiappa et al., 2005). According to the hypothesis, television viewers can form relationships with television characters similar to those that they form with real-life others. In turn, exposure to social minorities on television may have a similar effect in reducing stereotyping, prejudice, and discrimination as real-life intergroup

76 BOND, MILLER, AUBREY

contact. Ortiz and Brooks (2014) argued for the value of quantifying sex on television by noting that the development of parasocial relationships with sexual television characters might influence sexual outcomes of audience members who develop strong connections with those respective characters. Regardless of the theoretical perspective taken, media scholars have long argued that television exposure to out-group individuals can serve as one of many environmental factors influencing individuals’ perceptions of those different from themselves (Harris, 2004).

The Present Study

Although previous studies have quantified sexual references among LGB characters (Bond, 2014; Fisher et al., 2007), the depiction of LGB characters in the media is changing rapidly (GLAAD, 2016), and no previous research has quantified the sexual consequences of LGB characters on television. As such, the following research questions are proposed:

RQ1: Will heterosexual and LGB characters differ in the way they are depicted engaging in sexual talk or behavior on LGB-inclusive television programs?

RQ2: Will heterosexual and LGB characters differ in the way they are depicted engaging in LGB talk on LGB-inclusive television programs?

RQ3: Will sexual consequences differ for heterosexual and LGB characters depicted on LGB-inclusive television programs?

The present study also examines whether the sexual double standard (Aubrey, 2004) holds true for LGB characters by investigating the intersec- tion between gender and sexual orientation through the following research question:

RQ4: Will gender differences exist in the ways that LGB characters are depicted engaging in sexual references?

In addition, studies have found significant differences in the depiction of sex and the portrayal of LGB characters based on the origin of the program. Premium cable networks like HBO and Showtime portray the greatest fre- quency of sexual intercourse (Eyal & Finnerty, 2009). LGB characters are also depicted in sexual references more frequently on basic and premium cable networks than on broadcast networks (Netzley, 2010). The television landscape is arguably undergoing the most significant transformation since the golden age of the medium, given the growing popularity of on-demand streaming services like Hulu and Netflix. Both premium cable networks and streaming services may be more likely to depict LGB characters in sexual

SEXUAL REFERENCES AND CONSEQUENCES ON TV 77

situations given their freedom from advertisers. GLAAD (2016) reported that streaming services had the highest percentage of LGB characters in the 2016–2017 television season compared to broadcast and cable networks. Given potential content carrier differences, the following research question is posited:

RQ5: Will program origin differences exist in the frequency of LGB characters depicted engaging in sexual references or experiencing sexual consequences?

METHOD

Sample

GLAAD (2012, 2013, 2014) publishes an annual report detailing the television series that depict LGB characters. Any television series that depicted LGB characters in leading or reoccurring roles as reported in the 2012–2015 GLAAD annual reports were included in the purposive sample created for the present study. The GLAAD annual reports were an appropriate sampling frame for this study because they were representative of the variance in the television landscape, including broadcast, basic cable, premium cable, and popular streaming services such as Netflix and Hulu. The GLAAD reports have also been used in previous research as valid databases of LGB characters on television (e.g., Netzley, 2010). A graduate student screened two randomly selected episodes from every nonanimated series in each season from 2012 to 2015 to ensure that the episodes depicted at least one sexual reference. Given the objectives of the content analysis to examine patterns of sexual references and sexual consequences, episodes that did not contain sexual references were removed from the data set. After removing episodes that did not adhere to the screening criteria (n = 23), the final sample consisted of 132 television episodes that contained at least one LGB character and at least one sexual reference. Examples of television series in the sample are Modern Family (broadcast), Broad City (basic cable), Shameless (premium cable), and Orange Is the New Black (streaming). The full list of television show titles included in the sample is located in the appendix.

Coding

Reliability. Six undergraduate students served as coders for the present study. Three of the six coders analyzed sexual references and character

78 BOND, MILLER, AUBREY

demographics under the direction of the first author. The remaining three coders analyzed sexual consequences under the direction of the third author. Coders were trained to understand and apply the codebook of definitions over an 8-week period. Training involved pilot coding on extraneous episodes from the same television programs included in the sample. All six coders were required to reach acceptable unitizing agreement, but only the three coders assigned to code specific variables needed to reach acceptable intercoder reliability, defined as α > .67 (Krippendorff, 2013). Coders began coding the sample once agreement was reached during the pilot. Approximately 10% of the sample was analyzed by all coders to conduct intermittent reliability checks and to monitor coder drift. The average reliability coefficient for the variables coded was .77 (range = .67–.87). Race/ethnicity was not analyzed in the results because the variable fell below acceptable intercoder reliability during coding (α = .65 source, α = .54 target). The alpha coefficients for all other variables accompany their definitions.

Unit of Analysis. Each sexual reference was coded as a separate unit of analysis. A sexual reference was defined as a discrete instance of sexual talk, sexual behavior, or LGB talk. A new unit of analysis began when there was a change in category, source, or scene. When the source or scene did not change but the sexual behavior became progressively more sexual, coders were instructed to code only the most explicit sexual behavior depicted in an effort to avoid overcoding behaviors, a practice in line with previous research (Kunkel et al., 2005). All six coders reached a total agreement of 73% on their unitization decisions.

Measures

Sexual References. Sexual talk (α = .84) was defined as firsthand conversations or discussion of sexual interests, or secondhand conversations about prior, anticipated, or desired sexual topics or issues. Sexual talk was coded into the following categories: talk about sexual interests, talk about past sexual experiences, talk toward sex, sex crimes, and other. Sexual behavior (α = .87) was defined as actions that convey a sense of potential, likely, or actual sexual intimacy. Sexual behavior was broken down into categories of behavior: physical flirting, romantic kissing, intimate touching, sexual intercourse, and other (Kunkel et al., 2005).

LGB Talk. LGB talk (α = .67) focused on the social and cultural context of LGB sexualities rather than talk about sexual interests or behaviors. LGB talk was broadly defined as any dialogue specific to LGB experiences. LGB talk was coded into the following categories of talk: equality, coming out, gay culture,

SEXUAL REFERENCES AND CONSEQUENCES ON TV 79

stereotypes, insults/jokes/mocking, distancing/reaffirming, speculation, and other (Bond, 2014).

Negative Consequences. Negative sexual consequences (α = .72) were defined as repercussions resulting from sexual activity, sexual relationships, or potentially sexual relationships (Aubrey, 2004). Negative consequences could be visually depicted or verbally discussed. After determining that a sexual reference was accompanied by a negative consequence, coders then indicated what type of negative sexual consequence was portrayed: emotional (α = .72), physical (α = .93), romantic relational (α = .74), or nonromantic relational (α = .77). Emotional consequences included distress, guilt, anxiety, regret, shock, shame, anger, grief, and discomfort. The expression of emotion needed to be unmistakable and explicit. Negative physical consequences were defined as the portrayal or discussion of the following physical risks: unplanned and unwanted pregnancy, sexually transmitted infections, sexual assault, or aggression. Romantic relational consequences were defined as expressed or observed detriments to existing romantic relationships, such as the dissolution of a relationship or a serious conflict. Nonromantic relational consequences were defined similarly but the referent relationship could be any nonromantic relationship. Because a single consequence could fall into more than one of these categories, the coders were instructed to indicate whether the negative sexual consequence conformed to each of the four categories as yes/no dichotomous coding decisions.

Positive Consequences. Positive sexual consequences (α = .69) were defined as beneficial outcomes from sexual activity, sexual relationships, or potentially sexual relationships (Aubrey, 2004). Like negative sexual consequences, positive sexual consequences could be visually depicted or verbally discussed. Coders indicated what type of positive consequences the characters experienced: emotional (α = .77), physical (α = .75), romantic relational (α = .71), or nonromantic relational (α = .67). Emotional consequences included happiness, excitement, satisfaction, and contentment. Positive physical consequences were explicit and observed expressions of physical pleasure or planned, wanted outcomes of sexual activity (e.g., planned pregnancy). Romantic relational consequences included benefits to a romantic relationship, such as stronger relational bonds. Nonromantic relational consequences benefited others such as friends, families, and teachers. Similar to negative consequences, coders were instructed to indicate whether the sexual consequence conformed to each of the four categories as yes/no dichotomous coding decisions so that more than one consequence could be coded per sexual reference.

Demographics. The demographic characteristics of those engaging in sexual references were also coded. Gender of the source (α = .83) and target (α = .72) was

80 BOND, MILLER, AUBREY

coded as male, female, male-to-female transgender, or female-to-male transgender. Transgender characters (n = 11) were dropped from the analyses due to low frequencies. Age of the source (α = .85) and target (α = .80) was coded as child, teen, adult, or older adult. Sexual orientation of the source (α = .82) and target (α = .72) was coded as heterosexual, gay male, lesbian, bisexual, questioning, or asexual. Character demographics were determined primarily by using the content and context of the episode being coded. However, coders were also provided with a list of characters that were determined to be LGB by the GLAAD reports (2012, 2013, 2014). Coders were told to reference the character list if they were unsure of a character’s sexual orientation during coding.

RESULTS

A total of 1,338 sexual references were coded. Sexual talk accounted for 53% (n = 715) of all sexual references. Sexual behavior (34%, n = 451) and LGB talk (13%, n = 172) composed the remaining sample of sexual references. Heterosexual characters initiating sexual references (n = 1,068) were equally likely to be male (51%, n = 546) and female (49%, n = 522), and were predominantly adults (78%, n = 832). When LGB characters were the source of sexual references (n = 270), they were most likely to be gay (56%, n = 152) adult (82%, n = 221) males (57%, n = 155). Lesbians (22%, n = 60) and bisexuals (21%, n = 56) were less likely to be depicted.

The source and target’s sexual orientations were also examined together to provide insight into the context of sexual references on television. A majority of sexual references occurred between two heterosexual characters (71%, n = 951). LGB characters were more likely to be depicted in a sexual reference with a heterosexual character (18%, n = 238) than with another LGB character (11%, n = 146).

Sexual Talk and Behavior

Heterosexual characters sourced 83% of sexual talk (n = 590); LGB characters sourced 17% (n = 125). There was no significant difference in the distribution of sexual talk categories between heterosexual and LGB characters, χ2(4, N = 715) = 1.15, p = .89. Regardless of sexual orientation, characters were most likely to discuss sexual interests when talking about sex, followed by past sexual experiences. Other categories of sexual talk composed less than 10% of all sexual talk references (see Table 1). As displayed in Table 2, z scores for differences in proportions suggest that sexual talk was significantly more likely to occur between two heterosexual characters than between a heterosexual character and an LGB character (z = 27.76, p < .001) or between two LGB characters (z = 31.67, p < .001). Sexual talk was significantly less

SEXUAL REFERENCES AND CONSEQUENCES ON TV 81

TABLE 1 Distribution of Sexual References by Sexual Orientation of Source

Heterosexual LGB

Sexual talk Sexual interest 63% (374) 66% (82) Past sexual experiences 22% (127) 21% (26) Talk toward sex 6% (37) 4% (6) Sex crimes 8% (47) 7% (9) Other 1% (5) 2% (2) Total 100% (590) 100% (125)

Sexual behavior Physical flirting 29% (111) 30% (20) Romantic kissing 51% (198) 49% (32) Intimate touching 6% (21) 1% (1) Implied intercourse 8% (31) 11% (7) Depicted intercourse 5% (19) 6% (4) Other 1% (5) 3% (2) Total 100% (385) 100% (66)

LGB talk Equality 11% (10) 10% (8) Coming out 10% (9)a 34% (27)b Gay culture 19% (18) 17% (13) Stereotypes 17% (16) 18% (14) Insults, jokes, mocking 24% (22)a 10% (8)b Distancing/Reaffirming 3% (3) 0% (0) Speculation 14% (13) 6% (5) Other 2% (2) 5% (4) Total 100% (93) 100% (79)

Note. Raw frequencies are in parentheses. Sexual talk, χ2(4, n = 715) = 1.15, p = .89; sexual behavior, χ2(5, n = 451) = 3.60, p = .61; lesbian, gay, bisexual (LGB) talk, χ2(7, n = 172) = 22.93, p < .01, V = .37. Differing subscripts across rows represent statistically significant differences of propor- tions as determined by pairwise comparisons using the Bonferroni method.

TABLE 2 Distribution of Sexual References by Source/Target Sexual Orientation

Heterosexual Source: Heterosexual Target

Mixed Source/Target Sexual Orientation

LGB Source: LGB Target

Sexual talk 71% (951)a 18% (238)b 11% (146)c Sexual behavior 82% (369)a 6% (28)b 12% (54)c LGB talk 41% (70)a 38% (66)a 21% (36)b

Note. Rows total 100% of sexual references. Raw frequencies are in parentheses. Differing sub- scripts across rows represent statistically significant differences of proportions as determined by z scores. LGB = lesbian, gay, bisexual.

82 BOND, MILLER, AUBREY

likely to occur between two LGB characters than between a heterosexual character and an LGB character (z = 5.07, p < .001).

Analyses for sexual behavior mirrored those of sexual talk. Heterosexual characters were more likely to be the source of sexual behavior (85%, n = 385) than LGB characters (15%, n = 66). Much like sexual talk, there was no significant difference in the distribution of sexual behaviors between heterosexual and LGB characters, χ2(5, N = 451) = 3.60, p = .61. Regardless of sexual orientation, characters were most likely to be depicted kissing when engaged in sexual behaviors followed by physical flirting (see Table 1). Table 2 reveals that sexual behaviors between two heterosexual characters (82%, n = 369) were significantly more likely than those between a heterosexual character and an LGB character (6%, n = 28; z = 22.87, p < .001) or between two LGB characters (12%, n = 54; z = 21.02, p < .001). Sexual behaviors between two LGB characters were significantly more likely to be depicted than those between a heterosexual char- acter and an LGB character (z = 3.01, p < .01).

LGB Talk

There were 172 references coded as LGB talk. Heterosexual characters initiated a majority of LGB talk (54%, n = 93). When engaged in LGB talk, the type of talk differed by the sexual orientation of the source, χ2(7, N = 172) = 22.93, p < .01, Cramer’s V = .37. The Bonferroni method for pairwise comparisons revealed that heterosexual characters were more likely to engage in insults/jokes/mocking (24%, n = 22) than LGB characters (10%, n = 8). LGB characters were more likely to be depicted talking about coming out (34%, n = 27) than heterosexual characters (10%, n = 9). Distributions are depicted in Table 1. An examination of the source/target dyad depicted in Table 2 suggests that LGB talk occurred significantly less between two LGB characters than between two heterosexual characters (z = −3.97, p < .001) or between a heterosexual and LGB character (z = −3.54, p < .001).

Sexual Consequences

Of the 1,338 sexual references, 28% (n = 374) were depicted as having at least one sexual consequence. A smaller percentage of sexual references had multiple conse- quences (10%, n = 141). The proportion of sexual references receiving any conse- quence was significantly higher for heterosexual characters (35%, n = 315) than for LGB characters (26%, N = 59), χ2(1, N = 1,124) = 6.80, p < .01, Cramer’s V = .08. Analyses of negative and positive sexual consequences separately revealed that heterosexual characters were more likely to receive negative consequences (28%, n = 252) than LGB characters (20%, N = 45), χ2(1, N = 1,124) = 6.38, p < .05, Cramer’s V = .08. However, no significant difference existed in the depiction of

SEXUAL REFERENCES AND CONSEQUENCES ON TV 83

positive consequences between heterosexual characters (8%, N = 68) and LGB characters (8%, N = 17), χ2(1, N = 1,124) = .01, p = .96. Separate chi-square analyses were run to examine proportion differences in the types of negative and positive consequences depicted. Results suggest no difference by sexual orientation. Sexual consequences were most likely to be emotional in nature regardless of the sexual orientation of the sexual reference source (see Table 3).

Analyses of the source/target dyads also suggest that two heterosexual characters engaged in a sexual reference were more likely to experience a negative consequence (28%, nN= 226) than two LGB characters (18%, N = 21), χ2(2, N = 1,122) = 6.78, p < .05, Cramer’s V = .08. Two heterosexual characters (8%, N = 64) and two LGB characters (10%, N = 12) were equally likely to be shown receiving positive con- sequences, but both source/target dyads were significantly more likely to attain positive consequences than interactions between a heterosexual and LGB character (3%, N = 6), χ2 (2, N = 1,122) = 7.86, p < .05, Cramer’s V = .08.

Gender Differences in LGB Depictions

Only sexual references initiated by LGB characters were included in analyses to examine gender differences. Chi-square analyses suggested no gender differences in the source of sexual talk, LGB talk, or sexual behavior. No gender differences were found when examining the distribution of the various categories of sexual talk, LGB talk, or sexual behavior that were coded. The frequency of sexual consequences also did not differ based on gender of the source. Negative consequences were equally likely among LGB male sources of sexual references

TABLE 3 Proportions of Sexual Consequence Categories by Sexual Orientation of Source

Heterosexual LGB

Negative consequences Emotional 89% (223) 80% (36) Physical 8% (21) 13% (6) Relational–Romantic 15% (38) 20% (9) Relational–Nonromantic 23% (58) 18% (8)

Positive consequences Emotional 65% (44) 88% (15) Physical 10% (7) 12% (2) Relational–Romantic 49% (33) 35% (6) Relational–Nonromantic 9% (6) 18% (3)

Note. Raw frequencies are in parentheses. Columns do not equal 100 because one sexual reference could have multiple consequences. Separate chi-square analyses for each consequence showed no difference in proportions of sexual consequences that were coded into each category by sexual orientation of source. LGB = lesbian, gay, bisexual.

84 BOND, MILLER, AUBREY

(26%) as they were among female sources (26%). LGB male sources of sexual references also experience positive sexual consequences (6%) at the same rate as LGB female sources (6%). The sexual double standard is traditionally defined as males initiating more sexual references than females but females facing more consequences for sexual references than males. Neither was found in the current study. The sexual double standard may be more nuanced when referencing LGB individuals. For example, differences in the gender of LGB targets of sexual references coded as “insults/jokes/mocking” might reveal a sexual double stan- dard related to sexual orientation rather than sexual talk or behavior. Results suggest, however, that LGB females (46%) were no more likely than LGB males (54%) to be the target of insults, jokes, or mocking. The sexual double standard may simply not be applicable to LGB television characters.

Program Origin

Nonparametric chi-square goodness-of-fit tests were run to examine differences by program origin. Streaming services and premium cable were bundled into one category to represent programming that is free from the traditional advertising structure of broadcast and cable. All sexual references sourced by LGB characters were included in the analyses. Expected values were set to 42%, 35%, and 23% to represent the proportion of episodes in the sample from broadcast networks, cable networks, and streaming/premium, respectively. The proportions significantly dif- fered from the expected values for sexual talk, χ2(2, N = 134) = 20.86, p < .001; LGB talk, χ2(2, N = 90) = 8.76, p < .05; sexual behavior, χ2(2, N = 69) = 12.90, p < .01; negative sexual consequences, χ2(2, N = 50) = 6.94, p < .05; and positive sexual consequences, χ2(2, N = 22) = 16.46, p < .001. Figure 1 visualizes the residuals that represent the difference between the expected frequency and the observed frequency for each coded variable. Results suggest that streaming/ premium overrepresented sexual talk, LGB talk, sexual behavior, and sexual con- sequences initiated by LGB characters in the sample. Cable networks underrepre- sented sexual talk, LGB talk, and positive consequences. Broadcast networks underrepresented sexual behavior.

DISCUSSION

The goal of the present study was to examine the depiction of sexual references on television with a specific focus on LGB characters and sexual consequences. When LGB characters were shown engaged in sexual references, the types of sexual talk or behaviors were similar to the sexual references of heterosexual characters. The frequency of LGB characters appearing on television is increas- ing (GLAAD, 2016) and the present study suggests that, when depicted in sexual

SEXUAL REFERENCES AND CONSEQUENCES ON TV 85

references, LGB characters are being depicted in a similar fashion to hetero- sexual characters. Such a finding could be interpreted as evidence for the normalization of LGB sexualities on entertainment television programming.

There were, however, significant differences between heterosexual and LGB characters’ talk about LGB sexualities. Heterosexual characters were most likely to be engaged in dialogue coded as insults, jokes, or mocking when talking about LGB sexualities. Charlie, a heterosexual male character on FX’s Anger Management, responded to a question about bus arrival times by saying, “Well, the straight bus gets here at three o’clock. The lesbian one gets her at three-thirty. It takes a long time to load all of those golf clubs.” Early research studies found that gay-themed jokes originated on television primarily from LGB characters, plau- sibly because audiences wanted to laugh at LGB sexualities but felt more comfor- table laughing at self-deprecating humor originating from LGB characters (Raley & Lucas, 2006). The results from the present study refute these early findings and support more recent research suggesting that LGB sexualities are often used for comedic objectives by heterosexual characters on television (Bond, 2014). LGB stereotypes framed as humor seem to remain a prevailing theme in narratives depicting heterosexual characters talking about LGB sexualities. Although media

Sexual Talk

LGB Talk

Sexual Behavior

Negative Consequence

Positive Consequence

FIGURE 1 Representation of sexual references involving lesbian, gay, bisexual (LGB) characters by program origin. Note. Nonparametric chi-square goodness-of-fit tests run with references originating from L char- acters. Values represent residual difference between expected values and observed values. Bars represent the residuals, or differences between expected frequency and observed frequency. Positive numbers represent overrepresentation; negative numbers represent underrepresentation.

86 BOND, MILLER, AUBREY

are only one of many societal resources for sexual information, exposure to such content could have a negative influence on LGB viewers by highlighting the differences between heterosexual and LGB sexualities, degrading the latter. Fisher and colleagues (2007) concluded their content analysis of LGB characters on television by arguing that the perpetual depiction of LGB sexualities in comedic contexts inhibited LGB audiences from cultivating attitudes and beliefs about “healthy and responsible” same-sex relationships (p. 185).

LGB characters in the sample were most frequently depicted talking about coming out when engaged in dialogue about LGB sexualities. Lily, the adopted child character of two gay men on ABC’s Modern Family, discloses that she thinks she is gay. The episode then revolves around her fathers trying to explain their coming-out stories to her in an effort to understand her out-of-character disclosure. The “coming-out” narrative may be so prevalent in entertainment television because LGB sexualities have no physical marker other than showing romantic affiliation with another character. Given that television producers are cautious about depicting same-sex sexual behaviors, verbal disclosure of sexualities must occur for audiences to interpret characters as LGB. In essence, television characters must make procla- mations about their sexualities to come out in much the same manner as real-life LGB individuals. Depicting coming-out narratives on-screen could be validating to LGB audiences. Social cognitive theory might presume that exposure to coming-out narratives provides LGB audiences with strategies for self-disclosure in their own lives, particularly if coming-out references are void of negative consequences (Gomillion & Giuliano, 2011). Oftentimes, however, coming-out narratives are the primary or even sole storyline for LGB characters. Consider Teddy on the CW’s 90210. Teddy was noted as a series regular in the 2010–2011 GLAAD report on LGB representation but was demoted to a reoccurring character in the 2011–2012 GLAAD report. Teddy’s coming-out narrative played out in the season of 90210 that originally aired in 2010–2011. Essentially, Teddy was less important to the plot once his closet door was opened. The perception that LGB individuals are of less value once they have openly disclosed their sexualities could be the conclusion of audiences exposed to this narrow focus on coming-out narratives.

The present study was the first to examine the target of the sexual reference in addition to the source, lending insight into the context of LGB characters on- screen. LGB characters were almost twice as likely to be shown engaging in sexual references with heterosexual characters than with other LGB characters. This finding may support the idea of a “token” LGB character. Producers are comfor- table depicting one LGB character in television programs, as the visibility might be a signal of social progress and acceptance that can be marketed to young, affluent audiences that are becoming harder to reach in today’s changing television land- scape. Chunk on CBS’s Bull, William on NBC’s This Is Us, Kevin on the CW’s Riverdale, and Blake on CBS’s Madame Secretary are just a few examples of sole LGB characters whose sexualities are arguably tokenistic. Fear of offending a

SEXUAL REFERENCES AND CONSEQUENCES ON TV 87

larger, more influential heterosexual audience may keep producers from depicting LGB sex and sexualities in authentic contexts that might require an LGB character to speak about sex or be engaged in sexual behavior with another LGB character (Bond, 2014). Indeed, writing LGB characters into realistic sexual situations is so novel that when it does occur, headlines are made and people take notice. Consider the relationship between Connor and Oliver on ABC’s How to Get Away With Murder. The characters are often shown in steamy, intimate sex scenes, sparking outrage from some conservative pundits, organizations, and news media (O’Neil, 2014). Shonda Rhimes responded to the anger by tweeting, “There are no gay scenes. There are scenes with people in them” (Shondarhimes, 2014). Rhimes’s effort at normalizing same-sex sexual references on her show acknowledges that such content is still anomalous. The presence of insults, jokes, and mocking LGB sexual references and the lack of LGB characters speaking to other LGB characters also supports Bond’s (2014) argument that LGB sexualities are sanitized on television, void of authentic sexual expression that would develop complex, mean- ingful LGB characters on-screen.

The focus on heterosexual–LGB dyads discussing LGB sexualities in humor- ous contexts may also influence heterosexual viewers’ perceptions of LGB individuals. The parasocial contact hypothesis predicts that heterosexual viewers can vicariously experience intergroup contact with LGB others through media (Schiappa et al., 2005). Majority, in-group members (e.g., heterosexual viewers) form scripts for interacting with the minority out-group (e.g., LGB individuals) from watching these interactions on-screen. If LGB characters are shown enga- ging with heterosexual characters in ways that belittle LGB sexualities, this is the lesson that heterosexual viewers could take from the televised interaction.

An objective of the present study was to examine sexual consequences given the salience of vicarious motivations like rewards and punishments as predictors of learning observed behaviors according to social cognitive theory (Bandura, 2009). Sexual consequences occurred for 28% of all sexual references in the sample, consistent with previous research suggesting that approximately 30% of sexual references are depicted as having consequences (Aubrey, 2004). Negative consequences occurred for one in five sexual references inclusive of LGB characters. For example, Frank ended his relationship with Agustin on HBO’s Looking after he learned of Agustin’s escapades with a sex worker. Although significantly less frequent, positive consequences also existed for LGB charac- ters. For example, Cam was granted approval from his boyfriend Mitch’s dis- cerning father following a conversation in which Cam talks about the sexual prowess of older men in a flirtatious fashion on ABC’s Modern Family. Heterosexual characters received more negative consequences for sexual refer- ences than LGB characters, but there was no sexual orientation difference in the representation of positive consequences for sexual talk or behavior. Sexual consequences may be a manifestation of relevance. Heterosexual characters’

88 BOND, MILLER, AUBREY

sexualities are likely more central to the overall plot than LGB characters’ sexualities, which may only be central to their own personal storyline. As such, heterosexual sexual talk and behavior has more narrative depth and must be given more attention, including the depiction of sexual consequences.

There was also no difference in the proportion of negative or positive sexual consequence categories by sexual orientation; emotional consequences were the most frequently depicted among both heterosexual and LGB sexual references. This equity in consequences could suggest that LGB characters are gaining respect by being depicted similarly to heterosexual characters. Alternatively, this finding could also be explained by the general heteronormative assumptions of the television industry. Research suggests that real-life sexual experiences and consequences differ between heterosexual and LGB individuals (Ward, Dahlhamer, Galinsky, & Joestl, 2014). For example, pregnancy is a prominent sexual consequence for heterosexual individuals, but it is irrelevant for gay and lesbian individuals. Televised sexual consequences may be similar between heterosexual and LGB characters because writers and producers do not have the knowledge or experience with LGB sexualities to create nuanced storylines that may address sexual consequences among LGB characters with greater sincerity and realism.

LGB males and females did not differ in the frequency with which they received sexual consequences or the types of consequences they received for their sexual references. LGB males and females were also equally likely to be the target of insults, jokes, or mocking related to sexual orientation. The lack of gender differences suggests that television’s sexual double standard (Aubrey, 2004) may not apply to characters identifying with sexualities other than hetero- sexual. The intersection between sexual orientation and gender muddies the sexual double standard, an important finding that suggests previous heteronor- mative studies investigating gender on television should be challenged when studying the visibility and representation of LGB sexualities on television.

Previous studies have found that lesbian and bisexual women are more likely to be involved in sexual references than gay men (e.g., Netzley, 2010). The gender discrepancy has been rationalized by noting that heterosexual men, who produce media content, are more accepting of lesbians than gay men. Others have found that gay males dominate in LGB representation (e.g., Fouts & Inch, 2005), arguing that such a finding maps onto larger societal gender inequalities: men maintain power by depicting other men, even if they are gay men. The present study found no significant differences in the gender of LGB characters initiating sexual talk, LGB talk, or sexual behavior. Future research should continue to dissect the intersection of gender and sexuality and the potential effects of exposure to LGB characters: Does a bisexual male identify with a bisexual female television character simply because the character is bisexual? Effects research should attempt to parse out these potential intersections.

SEXUAL REFERENCES AND CONSEQUENCES ON TV 89

The current study also reinforces previous research that suggests the content carrier matters (Fisher et al., 2007; Netzley, 2010). Streaming services and premium cable were the most progressive content carriers, depicting a higher proportion of same-sex sexual talk, LGB talk, sexual behavior, and sexual consequences than broadcast or cable networks. HBO, Showtime, Netflix, and Amazon Prime have all garnered critical praise and accolades for the sexual inclusivity in many of their original programs. From Showtime’s hit series Queer as Folk to the more recent Netflix drama Sense8, many original series on streaming services and premium cable outlets feature prominent LGB characters in sexual situations and engaged in dialogue about their sexual experiences and desires. The visibility of LGB sexualities on streaming services and premium networks could be explained using the economic rationale presented earlier. Advertisers are an essential factor in the economic model of broadcast and cable networks that are absent from the economic model of streaming services and premium cable. Relying on audiences as the primary revenue source allows producers to develop content that is not regulated by advertisers who may be fearful of how progressive content will be interpreted by potential consumers. The influence of advertisers may also explain the underrepresentation in broad- cast and cable networks. Broadcast networks underrepresented same-sex sexual behavior; cable networks underrepresented same-sex sexual talk and LGB talk. Although pressure from advertisers may explain the underrepresentation, the nature of the sample may also play a role in the cable findings in particular. A review of the series titles reveals that LGB characters on cable were more likely to be reoccurring characters and those on streaming/premium services were more likely to be lead characters. LGB characters on cable series may have played a lesser role and were likely coded with less frequency. Moreover, many of the cable shows were procedurals. In procedurals, less emphasis is placed on char- acter development and more emphasis is placed on resolving the initial conflict in 40 minutes or less. Explicit sexual references are less likely to occur in the procedural drama, a commonality on cable.

The underrepresentation on cable may also be explained by shifts in narrow- casting. Streaming/premium services may be more sophisticated in focusing their content to niche audiences, thereby permitting greater diversity in depicting sexualities than seen on broadcast or cable series that are still trying to garner larger audience segments. Regardless of the explanation, the finding that basic cable networks are particularly problematic spaces for LGB representation is ripe for future research that could confirm this finding, further dissect differences between networks and premium/streaming services, investigate the potential differences in access (particularly for youth) given the varying financial models, and ultimately study effects of exposure to programs based on their content carriers.

90 BOND, MILLER, AUBREY

It is worth noting that the original content analysis codebook included defini- tions for gender identities. As noted in the methods, transgender characters were dropped from analysis because of their infrequent appearance. Although gender identity was ultimately not investigated in the present study, transgender char- acters are becoming more visible on both scripted television (e.g., Amazon’s Transparent) and unscripted television (e.g., I Am Cait). Future research should consider taking a more a qualitative approach to understanding the context of transgender representation on television and the potential effects of exposure on both cisgender and transgender audiences.

Although this study operationalized variables not previously included in content analyses of LGB characters on television, the design did present several limitations worth noting. First, external generalizability is limited because the sample contained only programs that were inclusive of LGB characters. The sample is not representative of all television, just television programs that depict LGB characters and contain sexual references. Comparisons between heterosexual and LGB characters must be interpreted with caution.

It is likely that many of these programs, particularly the ones that have been noted for their LGB-inclusiveness, have a more LGB-friendly fan base and thus more interest and latitude in depicting same-sex sexualities than television programming as a whole. Television shows like American Horror Story, Looking, True Blood, Orange Is the New Black, and Glee have an atypical number of LGB characters. Future research should examine the demographics of the audience and potential selective exposure to LGB-inclu- sive content.

The sample excluded unscripted programs and animated programs. However, LGB characters appear regularly on unscripted programs (e.g., Andy Herren in CBS’s Big Brother) and on animated programs (e.g., Big Gay Al in Comedy Central’s South Park). Future research studies should expand on the parameters of the sampling frame for more generalizable results regarding both LGB and heterosexual sexual references.

The codebook for the present study allowed coders to quantify insults/jokes/ mocking of LGB sexualities but not for coding humor related to heterosexuality. Investigating if heterosexual sex and sexualities are discussed in a similarly humorous context would allow for insightful comparisons between heterosexual and LGB sexualities on-screen that this study could not speak to. In addition, this study sought to measure the race of television characters involved in sexual references. Unfortunately, the variable did not retain reliability through data collection and was dropped from analyses. Raley and Lucas (2006) found that LGB characters were likely White. LGB Whiteness would reinforce the argu- ment that LGB characters are presented in contexts that producers believe majority audiences would find most discernable and accepting; however, the

SEXUAL REFERENCES AND CONSEQUENCES ON TV 91

present study cannot speak to race. Other variables that were coded reached acceptable levels of reliability but were low. Future research studies should attempt to update insights on the humorous context of representation, investigate the intersections between sexual orientation and race/ethnicity as they are repre- sented on-screen, and improve training procedures to increase reliability coefficients.

In sum, when LGB characters are depicted as the source of sexual references, they are likely discussing issues pertinent to LGB individuals, and their sex- ualities are represented in similar contexts to their heterosexual counterparts. When heterosexual characters are depicted engaged in LGB talk, it is most likely to be talk deemed a joke, insult, or mocking situation that adds comedic effect to the narrative, potentially belittling LGB sexual identities. Sexual consequences are proportionally more likely to occur for heterosexuals engaged in sexual references than LGB characters engaged in sexual references, arguably because discussions about their sexualities are more authentic and grounded. Although frequency of LGB representation is increasing, the quality of those depictions needs to be carefully considered by producers given the potential influence on both heterosexual and LGB audiences’ beliefs, attitudes, and behaviors as they relate to LGB individuals and their sexualities.

REFERENCES

Aubrey, J. S. (2004). Sex and punishment: An examination of sexual consequences and the sexual double standard in teen programming. Sex Roles, 50, 505–514. doi:10.1023/B: SERS.0000023070.87195.07

Bandura, A. (2009). Social cognitive theory of mass communication. In J. Bryant & M. B. Oliver (Eds.), Media effects: Advances in theory and research (pp. 94–124). New York, NY: Routledge.

Bond, B. J. (2014). Sex and sexuality in entertainment media popular with lesbian, gay, and bisexual adolescents. Mass Communication and Society, 17, 98–120. doi:10.1080/15205436.2013.816739

Bond, B. J., Hefner, V., & Drogos, K. L. (2009). Information-seeking practices during the sexual development of lesbian, gay, and bisexual individuals: The influence and effects of coming out in a mediated environment. Sexuality & Culture, 13, 32–50. doi:10.1007/s12119-008-9041-y

Eyal, K., & Finnerty, K. (2009). The portrayal of sexual intercourse on television: How, who, and with what consequence? Mass Communication and Society, 12, 143–169. doi:10.1080/ 15205430802136713

Eyal, K., & Kunkel, D. (2008). The effects of sex in television drama shows on emerging adults’ sexual attitudes and moral judgments. Journal of Broadcasting and Electronic Media, 52, 161–181. doi:10.1080/08838150801991757

Finnerty, K. (2007). Risky sexual intercourse on entertainment television: Comparing audience responses to different types of negative consequence portrayals (doctoral dissertation). The University of Arizona Campus Repository.

Fisher, D. A., Hill, D. L., Grube, J. W., & Gruber, E. L. (2007). Gay, lesbian, and bisexual content on television: A quantitative analysis across two seasons. Journal of Homosexuality, 52(3/4), 167–187. doi:10.1300/J082v52n03_08

92 BOND, MILLER, AUBREY

Fouts, G., & Inch, R. (2005). Homosexuality in TV situation comedies: Characters and verbal comments. Journal of Homosexuality, 49, 35–45. doi:10.1300/J082v49n01_02

GLAAD. (2012). Where we are on TV ’12-’13. New York, NY: Author. GLAAD. (2013). Where we are on TV ’13-’14. New York, NY: Author. GLAAD. (2014). Where we are on TV ’14-’15. New York, NY: Author. GLAAD. (2016). Where are we on TV ’16-’17. New York, NY: Author. Gomillion, S. C., & Giuliano, T. A. (2011). The influence of media role models on gay, lesbian, and

bisexual identity. Journal of Homosexuality, 58, 330–354. doi:10.1080/00918369.2011.546729 Harris, R. J. (2004). A cognitive psychology of mass communication (4th ed.). Mahwah, NJ: Lawrence

Erlbaum. Krippendorff, K. (2013). Content analysis: An introduction to its methodology (3rd ed.). Thousand

Oaks, CA: Sage. Kunkel, D., Eyal, K., Finnerty, K., Biely, E., & Donnerstein, E. (2005). Sex on TV 4. Menlo Park, CA:

Kaiser Family Foundation. Kunkel, D., Farrar, K. M., Eyal, K., Biely, E., Donnerstein, E., & Rideout, V. (2007). Sexual

socialization messages on entertainment television: Comparing content trends 1997-2002. Media Psychology, 9, 595–622. doi:10.1080/15213260701283210

Morgan, M., Shanahan, J., & Signorielli, N. (2009). Growing up with television: Cultivation processes. In J. Bryant & M. B. Oliver (Eds.), Media effects: Advances in theory and research (pp. 34–49). New York, NY: Routledge.

Netzley, S. B. (2010). Visibility that demystifies: Gays, gender, and sex on television. Journal of Homosexuality, 57, 968–986. doi:10.1080/00918369.2010.503505

O’Neil, L. (2014, October 20). Shonda Rhimes responds to intolerance over “gay scenes” in her TV shows. The Hollywood Reporter. Retrieved from www.hollywoodreporter.com

Ortiz, R. R., & Brooks, M. E. (2014). Getting what they deserve? Consequences of sexual expression by central characters in five popular television teen dramas in the United States. Journal of Children and Media, 8, 40–52. doi:10.1080/17482798.2014.863477

Raley, A. B., & Lucas, J. L. (2006). Stereotype or success? Prime-time television’s portrayal of gay male, lesbian, and bisexual characters. Journal of Homosexuality, 51, 19–38. doi:10.1300/ J082v51n02_02

Schiappa, E., Gregg, P. B., & Hewes, D. E. (2005). The parasocial contact hypothesis. Communication Monographs, 72, 92–115. doi:10.1080/0363775052000342544

Shondarhimes, (2014, October 19). @Dabdelhakiem there are no GAY scenes. There are scenes with people in them [Twitter post]. Retrieved from https://twitter.com/shondarhimes

Strasburger, V. C. (2012). Adolescents, sex, and the media. Adolescent Medicine, 23, 15–33. Streitmatter, R. (2009). From “perverts” to “fab five”: The media’s changing depiction of gay men

and lesbians. New York, NY: Routledge. Ward, B. W., Dahlhamer, J. M., Galinsky, A. M., & Joestl, S. S. (2014). Sexual orientation and health

among U.S. adults: National health interview survey 2013. In National Center for Health Statistics (Ed.), National health statistics reports (Vol. 77). Hyattsville, MD: Centers for Disease Control and Prevention.

Winderman, K., & Smith, N. G. (2016). Sexual minority identity, viewing motivations, and viewing frequency of LGB-inclusive television among LGB viewers. Sexuality & Culture, 20, 824–840. doi:10.1007/s12119-016-9361-2

APPENDIX

List of Television Show Titles in Sample

SEXUAL REFERENCES AND CONSEQUENCES ON TV 93

Anger Management Army Wives Awkward Being Human Bones Broad City Californication Chasing Life Chicago Fire Defiance Degrassi Episodes Game of Thrones Glee Gotham Grey’s Anatomy Happy Endings How To Get Away With Murder Looking Lost Girl Major Crimes Marry Me Mistresses Modern Family Nashville Nurse Jackie One Big Happy Orange Is the New Black Orphan Black Portlandia Pretty Little Liars Ray Donovan Revenge Rookie Blue Scandal Sean Saves the World Shameless Sirens Southland The Bridge The Carrie Diaries The Flash

94 BOND, MILLER, AUBREY

The Fosters The Good Wife The Killing The L.A. Complex The Millers The Office The Secret Life of an American Teenager True Blood Two and a Half Men Undateable Web Therapy White Collar

SEXUAL REFERENCES AND CONSEQUENCES ON TV 95

Copyright of Mass Communication & Society is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.