Introduction

LGBTQ + characters seem to be more prevalent in the media than ever before. For example, 10% of the regular characters on broadcast primetime television in 2019–2020 were LGBTQ, representing a peak over 15 years of tracking this representation (GLAAD, 2019). This media representation may both reflect and generate a parallel increase in acceptance toward the LGBTQ + population that has been observed over the same time period (Brown, 2017; Callender, 2015). The content of these portrayals, both in terms of who is portrayed and how they are portrayed, is less clear from previous research. Identities at the intersection of oppressions are often invisible (e.g., Crenshaw, 1989; Nolke, 2018; Purdie-Vaughns & Eibach, 2008). For example, gay men may be overrepresented in portrayals of LGBTQ characters, while heterosexual women are overrepresented in portrayals of women, leaving lesbian women less visible. Indeed, gay men are the most commonly represented on primetime television among LGBTQ + identities (GLAAD, 2019), and content analyses have revealed that gay men are depicted significantly more than lesbian women or bisexual individuals in LGBTQ + media (Bond, 2015; Nolke, 2018). This discrepancy is echoed in the scholarly literature, where a quick search in Google Scholar (conducted July 1, 2020) returned approximately three times as many results for the search term “gay men” as it did “lesbian women,” both overall (2.1 million vs. 713,000) and when limited to the last five years (97,000 vs. 29,800). Meanwhile, studies of women have centralized heterosexual women, creating an invisibility of lesbian women (e.g., no mention of lesbian women in Ward’s, 2016 review of sexualization in the media) in the media overall. This invisibility is exacerbated by the trope that lesbian identities are transient, such that lesbian women will eventually partner with men (Diamond, 2005; Jackson & Gilbertson, 2009). Thus, focusing exclusively on the portrayals of lesbian women can address an important gap in the literature.

Visual media’s portrayal of lesbians could affect the way lesbians are perceived in the public domain as well as how they perceive themselves, as echoed by scholarly literature and popular culture as exemplified in the documentary The Celluloid Closet based on the 1981 work of Vito Russo (Epstein et al., 1996). Queer women have reported feeling disempowered and marginalized at the invisibility of certain queer identities (e.g., masculine-presenting women, or butch lesbians) in the media (Randazzo et al., 2015). This finding is representative of a larger issue that seems to affect how minority populations become portrayed in music videos, films, or television: invisibility leads to the push for more media portrayals, and the subsequent portrayals are often stereotypical and sexualized to mirror “cookie-cutter” heterosexual identities and relationships (e.g., Epstein et al., 1996; Jackson & Gilbertson, 2009; Nolke, 2018). In this regard, the current study aimed to identify how lesbians are presented in the media and the resulting effects on lesbians’ mental health and well-being.

Because sexual orientation is often unmarked compared to more visibly recognized identities based on race or gender, the media may be an especially powerful force in shaping public perceptions of lesbians (Epstein et al., 1996; Gross, 1991; Raley & Lucas, 2006). McLaughlin and Rodriguez (2016) introduced the concept of stereotyped identification, uncovering in a nationally representative sample that relatable media portrayals of lesbian and gay individuals promote acceptance via identification with familiar characters while simultaneously reifying stereotypes. Furthermore, depictions of lesbians may shape identity development among lesbians via cultural ideology and discourse (Hammack, 2005). Young women who look to popular visual media to understand their sexuality are seemingly pushed toward heterosexuality, though such ideas are often viewed as a personal choice rather than the heteronormativity of mainstream media (Diamond, 2005; Rich, 1980). Women’s “to-be-looked-at-ness" fostered by the male gaze is indicative of the pervasive objectification of women that is woven into visual media depictions (Mulvey, 1989). This may have deeper implications for the lesbian community, as an absence of positive role models could push young lesbians to learn from the media what a lesbian “looks like,” how she acts, and what to expect from romantic relationships (Epstein et al., 1996; Tukachinsky & Dorros, 2018). This may be especially true for young lesbians of color, as previous research has shown that Black and Latina youth are often more cautious about coming out and have a slower identity integration process compared to White youth (Rosario et al., 2004).

The racial diversity in women’s experiences is central to Kimberly Crenshaw’s (1989) multi-dimensional model of intersectional feminism inspired by the works of Richard Wasserstrom (1977), Angela Davis (1983), Taub & Schneider (1982), Phyliss Palmer (1983), and Paula Giddings (1984), among others. Her focus on the invisibility of Black women’s experiences illuminates limitations in traditional explorations of sexism (focused on White women) and racism (focused on Black men). Intersectionality is a key tenant of Critical Race Theory (CRT), which asserts that race, at the intersection of class, mobility, gender, power, sexuality, ideology, and other systems of subordination, has an impact on the way an individual interacts with the world around them (Rollock & Gillborn, 2011). Indeed, previous research has revealed racial and ethnic differences within the lesbian population, in that lesbians of color may feel more constrained in expressing their sexuality due to cultural norms within their ethnic groups and the perceived Whiteness of the LGBT community (Fingerhut et al., 2005; Rosario et al., 2004). CRT would predict that lesbians of color may perceive and respond to the portrayal of lesbians in the media differently than White lesbians, due to the intersection of oppressions based on heterocentric, androcentric, and ethnocentric ideologies. For example, Black women’s lived experiences with stereotypes (e.g., the Jezebel stereotype; Anderson et al., 2018) may intersect with lesbian stereotypes in ways unlike that of White women. Thus, exploring the racial diversity within lesbian perceptions of media portrayals is an additional goal of the present research.

The purpose of this study is to assess the portrayals of lesbian women in the media from the perspective of lesbian women. Previous research on this topic has utilized content analyses (e.g., Bond, 2015; Milillo, 2008; Nolke, 2018) or surveyed heterosexual participants (e.g., Levina et al., 2000; Louderback & Whitley, 1997; McLaughlin & Rodriguez, 2016). However, lesbians offer a critical viewpoint on media portrayals because of their unique standpoint in society at the intersection of gender and sexual orientation (Krane, 2001). Lesbian women are likely to be especially motivated to pay attention to and scrutinize portrayals of lesbians in media of all forms (Epstein et al., 1996). Furthermore, in past research examining issues pertaining to the LGBTQ + community, participant pools often consisted of either a mixture of gay men and women (e.g., Kite & Whitley, 1996) or lesbians and bisexual women (e.g., Randazzo et al., 2015). Yet, previous research has shown that there are meaningful differences between lesbians and other sexual minority women (e.g., Balsam & Mohr, 2007; Borgogna et. al, 2019; Dyar et al., 2015). For example, Dyar et al. (2015) and Balsam and Mohr (2007) both found that lesbian women have higher degrees of centrality and certainty in their sexual orientation identity, implying that being a lesbian is a larger part of their sense of themselves compared to bisexual and other queer women. Therefore, representations of lesbians in the media may feel more identifiable and/or relevant to lesbians rather than other sexual minority women.

The vast literature on objectification theory is an example of a body of research that has largely overlooked the lesbian experience. Numerous studies have found that objectifying media portrayals of women have a detrimental effect on women’s body image, self-esteem, and mental health (Ward, 2016). However, much of this research focuses on heterosexual women despite women of other sexual orientations being hypersexualized as well (Worthen, 2012). Most of the research that does examine lesbian women groups them together with other sexual minority women (e.g., Moradi et al., 2019), making it difficult to understand lesbian experiences of objectification (for an exception, see Haines et al., 2008). For example, there has been no prior published research linking portrayals of lesbians in the media and lesbians’ body image, either correlationally or experimentally. The present study aims to address this gap in the literature.

Portrayal of Lesbians in Media

Women are frequently sexualized in numerous forms of media (APA, 2007; Ward, 2016), with lesbians being no exception (Gill, 2009; Tebbe et al., 2018). Though previous research has not documented lesbian perspectives, content analyses and surveys of mixed groups do point to important themes in how lesbians are portrayed in the media. Lesbians have become increasingly commodified in various forms of media to sell a multitude of products by reinforcing the current perceived norm in popular culture of an ‘appropriate’ representation of lesbianism (Gross, 2001; Jackson & Gilbertson, 2009; Raley & Lucas, 2006). Historically, these portrayals have shifted from villainous presentations to comedic relief, as well as from unattractive and masculine to hot and desirable (Epstein et al., 1996; Jackson & Gilbertson, 2009). Both heterosexual and gay men identify that “beautiful blondes” are the stereotypical lesbians, with the pervasiveness of the hot lesbian subtype heightening the invisibility of masculine-presenting women who are referred to by male participants as “butch ugly” (Hegarty & Beuchel, 2011). The predominance of the hot lesbian subtype seems due to the centering of the male gaze (Mulvey, 1989), wherein popular visual media targets male consumers' ideals of being sexually involved with two (or more) women at one time (Albertson, 2018; Epstein et al., 1996). Indeed, previous research has found that sexual minority women are primarily sexualized by men (Tebbe et al., 2018).

The performative quality of lesbianism for a heterosexual male audience is widely accepted in music, movies, and television (Diamond, 2005). Heterosexual men are seemingly both intrigued by and fearful of the prospect of same-sex attraction among women (Epstein et al., 1996). The phallocentrism unpacked by Laura Mulvey lends itself to straight men’s active desire to watch lesbians while simultaneously being fearful of their apparent indifference and rejection (Mulvey, 1989; Rich, 1980). Scholars have long speculated that heterosexual men identifying with the idea of hegemonic masculinity are likely to be more fearful of homosexual male attraction and conversely fascinated by homosexual female attraction (Epstein et al., 1996). The “girls kissing girls” phenomenon was exemplified in mainstream music culture with Katy Perry’s hit song “I Kissed a Girl” (Perry et al., 2008), in which the singer openly recounts using women as a form of sexual experimentation. This marked a popular example of lesbianism being simultaneously commodified and devalued as a viable sexual orientation when Perry references kissing a girl “just to try it” rather than due to actual sexual desire (Diamond, 2005). In the case of lesbians portrayed on American television, the message is sent to viewers that lesbianism is, in part, a performance that can be enacted for the attention of heterosexual men (Randazzo et al., 2015). Heterosexual men encourage lesbianism for their own pleasure due to a common desire to be with two women at once, while simultaneously devaluing it by associating lesbianism with an act of performance (Jackson & Gilbertson, 2009). This male-centered approach contributes to assumptions of heteroflexibility, which is when primarily heterosexual women hint at or experiment with same-sex sexuality (Diamond, 2005). The popular media’s frequent portrayal of heteroflexibility may foster the relationship stereotype that lesbianism is merely an “experimental phase” rather than a valid relationship (Epstein et al., 1996). Notably, the heteroflexible characters which permeate the media space may be interpreted antithetically by lesbians versus bisexual or other sexual minority women. Therefore, it is worthwhile to examine lesbian perceptions of media portrayals separately from other sexual minority women, hence the recruitment of an all-lesbian sample in the current study.

Where there is variation in media portrayals of lesbians, stereotypes still dominate. Previous research has identified six lesbian subtypes that are reflected in the media: the out lesbian, the closeted lesbian, the bisexual lesbian, the feminine lesbian, the butch lesbian, and the hot lesbian (Brambilla et al., 2011; Szymanski et al., 2010). These group subtypes can quickly become stereotypes: as the portrayal of subtypes becomes more formulaic, colloquial understanding of the larger group identity (i.e., lesbians) becomes fractured into knowledge of widely consumed stereotypes (e.g., hot, hypersexualized, feminine women engaging for the androcentric viewer; McLaughlin & Rodrigues, 2016). The cyclical nature of the subtype/stereotype relationship is a double-edged sword: when a subtype is amplified in the media, it adds diversity to lesbian portrayals but becomes a stereotype that curtails the full spectrum of lesbian identities. Popular shows such as Desperate Housewives (DH) or Orange is the New Black (OITNB) often employ characters that fall into one or more of the aforementioned lesbian tropes. For example, Robin (DH), represents the quintessential hot/feminine lesbian trope and is distinctly sexualized; she is often filmed in full-body shots to emphasize her low-cut clothing while her heterosexual counterparts are most frequently shot from the waist or shoulders up. Contrarily, Big Boo (OITNB) is a classic portrayal of a butch lesbian (i.e., short hair, masculine features and tendencies) even to the point of having “Butch” tattooed on her forearm. More contemporary media like OITNB displays an array of lesbian presentations while still operating within the confines of certain stereotyped identities, such as the bisexual lesbian (Lorna Morello) and the out lesbian (Alex Vause), among others. A more thorough understanding of how lesbians are portrayed in the media was an important goal of the present research.

Objectification of Lesbian Women

The predominance of the hot lesbian subtype in visual media raises the question of how objectification theory applies to the lesbian community. Objectification occurs when a woman is treated as being merely a body who exists for the use and pleasure of others, separate from her person (Nussbaum, 1995). The objectification of women is pervasive, especially for Black women who are subject to the Jezebel stereotype (Anderson et al., 2018). Objectification theory posits that sexual objectification can be internalized as self-objectification, which can increase women’s opportunities for body shame as well as anxiety about their physical appearance and physical safety (Fredrickson & Roberts, 1997). Simultaneously, self-objectification can decrease awareness of internal body sensations (e.g., blood sugar levels) as well as opportunities for motivational states (e.g., rare moments during which one feels creative, joyful, and uncontrolled by others). This, in turn, can lead to negative body image, mood, and poorer self-esteem, often manifested through symptoms of depression, eating disorders, and sexual dysfunction (Calogero, 2012; Fredrickson & Roberts, 1997; Moradi & Huang, 2008; Szymanski et al., 2010). As postulated by Fredrickson and Roberts (1997), sexual objectification, both interpersonal and internalized, is likely to be influenced by factors such as sexual orientation, social class, and race/ethnicity. Indeed, Moradi (2010) elaborates on how objectification theory is well-suited to understanding differences across race, ethnicity and gender identity. However, such intersectional factors have not been adequately studied, which supports the need for research regarding how sexual objectification intersects with other aspects of women’s sociocultural identities (Szymanski et al., 2010).

Objectification, both of others and of the self, is a frequent and largely unconscious occurrence within the community of sexual minority women (Moradi et al., 2019; Tebbe et al., 2018). Yet, little research has been done to test objectification theory with an all-lesbian sample (for an exception, see Haines et al., 2008). Kozee and Tylka’s (2006) study found that the objectification theory framework proposed by Fredrickson and Roberts (1997) was a poor fit for a lesbian sample while being a strong fit for heterosexual women. This could indicate that the relationships between objectification theory variables are more distinct and complex for lesbian women than heterosexual women (Kozee & Tylka, 2006). Higher levels of body surveillance could signify that lesbians monitor their appearance more than heterosexual women, perhaps to ensure visibility or to avoid derision (Kozee & Tylka, 2006). However, there are also findings that indicate that lesbian and heterosexual women report similar levels of objectification in their life experiences from sources other than mass media (Hill & Fischer, 2008). Given the limited and mixed previous research, an additional goal of the present study was to test objectification theory in an all-lesbian sample. Previous research has primarily used qualitative and correlational methods to test objectification theory in queer women (e.g., Haines et al., 2008; Moradi et al., 2019), though research with predominantly heterosexual participants has experimentally manipulated exposure to thin-ideal and sexualized images to demonstrate their impact on self-objectification and a variety of mental health outcomes (Grabe et al., 2008; Ward, 2016).

Present Study

In summary, research from the past two decades has shown that women are sexualized in popular media characterizations, and that this sexualization has a negative impact on their body image and self-esteem (Ward, 2016). However, the vast majority of this research focuses on heterosexual women, and so the present study aims to address the gap in the literature by utilizing an all-lesbian sample in a mixed-method study with experimental, correlational, and qualitative components.

The first aim of the current study was to analyze lesbians’ perceptions of how lesbians are portrayed in the media overall. Because of extensive previous research emphasizing pervasive sexualization of women in the media (e.g., Ward, 2016), our first hypothesis was:

H1: The hot lesbian is the most commonly reported subtype.

Because little previous research has assessed stereotypes of lesbian relationships, we posed this research question instead of testing a specific hypothesis:

RQ1: What stereotypes regarding lesbian relationships are most frequently observed in the media by lesbian women?

The second aim of the present study was to test objectification theory in an all-lesbian sample using experimental methods. Previous research on objectification theory has demonstrated that exposure to media portrayals of sexual objectification leads to self-objectification and negative mental health outcomes (Grabe et al., 2008; Ward, 2016). Therefore, we hypothesized that:

H2: Compared to the control group, participants exposed to a video of a sexualized portrayal of a lesbian character will report increased self-objectification, negative mood, lower self-esteem, and negative body image.

Additionally, in an effort to make the most of the data from a uniquely all-lesbian sample, we aimed to replicate previous correlational research on objectification theory (Grabe et al., 2008; Ward, 2016) by testing whether the dependent variables (self-objectification, mood, self-esteem, and body image) from the experimental component of the study were correlated with the reported exposure to stereotypes of lesbians in the media.

H3: If a participant reports higher exposure to lesbian stereotypes in the media, then they will report increased self-objectification, negative mood, lower self-esteem, and negative body image.

Finally, this study offered an opportunity to examine racial differences in how lesbians perceive and are affected by portrayals of lesbians in the media. Race cuts across each of the research topics brought together in the present study: experiences of lesbians (e.g., Fingerhut et al., 2005; Rosario et al., 2004), media portrayals (e.g., Behm-Morawitz & Ortiz, 2013), and objectification (e.g., Anderson et al., 2018). Thus, we explored the following research question by analyzing each of the above hypotheses by race:

RQ2: Do lesbians of color have different experiences with the portrayal of lesbians in the media than White lesbians?

Method

Participants

The online platform Prolific Academic was used to recruit lesbian women who were citizens of the United States. The mean age of the 179 participants was 31.54 (SD = 11.57) with an average BMI of 28.10 (SD = 7.66). The racial makeup of the sample included 72.1% of participants identifying as White, 12.8% Black or African American, 6.1% Hispanic or Latino/a, 3.9% Asian, 2.8% identifying as ‘other’, 1.7% American Indian or Alaska Native, and 0.6% Native Hawaiian or Pacific Islander. Participants reported a wide range of education levels: 33% with some college education, 32.4% with a 4-year college degree, 13.4% were high school graduates, 12.9% with a professional or doctoral degree, 7.3% with a 2-year college degree, 0.6% with less than a high school degree and 0.6% responding as ‘other’. The employment status of participants was that those employed full- and part-time made up 67.6% of the sample, 15.1% students, 10.7% unemployed looking or not looking for work, 3.4% retired or disabled, and 3.4% identifying as ‘other’. On average, participants reported that they were moderately familiar with the show Desperate Housewives (M = 3.85, SD = 1.19), from which clips were used as the experimental stimulus. Participants were paid through Prolific Academic $1.95 each; availability of funding determined the sample size. Participants were treated in accordance with APA ethical guidelines.

Materials

Experimental Manipulation

Participants were randomly assigned to view one of two video clips from the show Desperate Housewives, Season 6, Episode 15, in order to manipulate exposure to stereotypical portrayals of lesbians. Desperate Housewives was chosen due to the availability of clips seeming comparable in terms of portraying the same characters in similar settings but with differences in stereotypicality, our independent variable. Due to Desperate Housewives’ age and target audience of heterosexual women, participants were expected to be less familiar with the chosen video clips, thus producing more authentic responses to the stimuli.

To select the video clips, four possible video clips from Desperate Housewives were pretested using a separate sample from the main study (N = 35 lesbian or bisexual women). The four clips were matched pairs (i.e., two experimental and two control clips) that were similar on objective characteristics such as the actors featured and the length of the scene. The aim of the pretest was to identify a pair of video clips that differed in stereotypical portrayals of lesbians yet were similar in terms of how engaging they were. Pretest participants were instructed to watch the four video clips (two neutral interactions and two interactions displaying one or more lesbian stereotypes) and respond to two sets of questions after each video on a scale of 1 (not at all) to 5 (very much). The first set of four questions examined the extent to which each video represented a lesbian stereotype (e.g., “To what extent did this video portray a lesbian who is hypersexualized?”). The second set of four questions measured the engagement level of the video (e.g., “How entertaining was this video?”). Upon completion of the videos and question sets, participants were asked demographic information about their age, sex, ethnicity, sexual orientation, and level of education.

A series of t-tests revealed that the video clips chosen for the main study were significantly different on all four stereotypes assessedFootnote 1: portraying performative lesbian actions for men, lesbian relationships as primarily sexual experimentation, lesbians as hypersexualized, and lesbians as unsure of their sexual orientation. The pair of videos did not significantly differ on three out of four of the engagement questions: emotionality, entertaining, and easy to pay attention to.Footnote 2 However, the control video was rated as funnier (MC = 2.71, SDC = 1.22) than the experimental video (ME = 2.12, SDE = 1.09), t(33) = 2.83, p = 0.008. Despite this one significant difference, these two clips were chosen because the other two clips pretested had fewer differences in the stereotype questions and more differences in the engagement questions.

Both the experimental and the control videos portrayed the same four characters in the same setting (Robin, Katherine, and two men at a bar) and were approximately the same length (45 & 50 s). In the experimental clip, two men at a bar unsuccessfully attempt to flirt with Robin, an out lesbian, who responds by kissing her heterosexual friend Katherine on the lips. The women walk away and leave the men gawking in their absence, and the scene cuts to Katherine and Robin discussing the men’s reaction to the event. In the control clip, Katherine and Robin are sitting with the same two men at the bar and having a conversation about Robin’s occupation, during which the sexual orientation of the characters was not revealed.

Dependent Measures

Self-objectification

The surveillance subscale of the Objectified Body Consciousness Scale (OBCS) assesses the extent to which a participant monitors their body (McKinley & Hyde, 1996) and is commonly used as a measure of self-objectification. Eight questions are answered on a 7-point scale (1 = Strongly agree, 7 = Strongly disagree). Items were slightly modified to assess state-level (as opposed to trait-level) self-objectification, as has been done by previous researchers (e.g., Betz et al., 2019), by adding qualifiers such as “right now” or “while watching the video” to the original items (e.g., “While watching the video, I thought about how I look many times,” “Right now I am thinking about how I look compared to how other people look”). Cronbach’s alpha for this scale was 0.81.

Mood

The PANAS-SF is a more concise version of the original PANAS used to assess mood (Watson et al., 1988). The aim of the questionnaire is to determine how a person feels at that moment (i.e., “Indicate the extent to which you have felt this way in the past week”) on a 5-point scale. One subscale with ten items measures a person’s positive emotion (i.e., enthusiastic, proud, attentive) and the other subscale with another ten items measures the negative emotions (i.e., jittery, hostile, upset). Cronbach’s alpha was 0.90 for the positive scale and 0.90 for the negative scale.

Body Image

Three subscales of the Multi-Dimensional Body Self-Relations Questionnaire (MBSRQ; Cash, 2015) were utilized to assess attitudes toward one’s body and appearance on a scale from 1 (definitely disagree) to 5 (definitely agree). Appearance evaluation (7 items; α = 0.91) measured general feelings about one’s physical appearance (e.g., “My body is sexually appealing”, “I like my looks just the way they are”). Appearance orientation (12 items; α = 0.88) assessed how much time and energy a person spends trying to “look good” (e.g., “I am careful to buy clothes that will make me look my best”, “I check my appearance in the mirror whenever I can”). Body area satisfaction (9 items; α = 0.87), measured how satisfied/dissatisfied a person is with specific areas of their body, such as “lower torso (buttocks, hips, thighs, legs)”.

Self-esteem

The Rosenberg Self-Esteem Scale (Rosenberg, 1965) is a ten-item questionnaire intended to assess self-esteem (global and general self-worth) via positive and negative feelings about oneself on a four-point Likert scale, from 1 (Strongly agree) to 4 (Strongly disagree). Some items include “I feel I do not have much to be proud of” and “On the whole I am satisfied with myself.” The Cronbach’s alpha for this scale was 0.94.

Exposure to Lesbian Stereotypes in the Media

As there are no validated scales that measure self-reported exposure to lesbian stereotypes in television media, an original measure (10 items; α = 0.70) was created using a 5-point Likert scale (1 = Not very often, 5 = Very often) consistent with measures of frequency of other subjects (e.g., violence) on television (Krahé et al., 2011). The original measure is composed of thirteen items: six questions about lesbian subtypes (i.e., hot, butch, feminine, out, closeted, and bisexual; Brambilla et al., 2011) and four questions about colloquial lesbian relationship stereotypes (i.e., invalid, vacation, performative, and move too quickly). For example, one item was “How often do you come across a portrayal of a “butch” lesbian (a lesbian whose gender expression is stereotypically masculine or ‘boyish’) in popular media?” See “appendix” for full text of items.

Two additional questions gauged the overall perception of lesbians and their relationships in media (i.e., “Would you say that portrayals of lesbians [lesbian relationships] in popular media are generally positive, generally negative, or is it hard to say?”) on a 5-point scale (1 = generally negative, 2 = somewhat negative, 3 = hard to say, 4 = somewhat positive, 5 = generally positive).

Qualitative Data

Participants were asked an open-response question at the end of the study: “In your experience, how would you describe the ways in which the media portrays lesbians and their relationships?” Based on her thorough reading of the data, one researcher created a list of possible codes. Then, responses from ten participants were discussed by both researchers in relation to the codes by both researchers, in order to clarify definitions and examples of the codes. Researchers then coded the responses independently. Throughout the process, definitions of codes were further clarified and additional codes were added. After three rounds of coding, researchers matched codes 97% of the time. The qualitative responses with differences in coding were not included in analyses. See Table 1 for full list of codes and definitions.

Table 1 Coding of open-ended responses: definitions, examples, and frequencies
Full size table

Procedure

The study was completed online via Qualtrics survey software and the participant-recruitment website Prolific Academic. Researchers obtained informed consent from participants prior to the beginning of the study. Participants were told that the study was a measure of the effects of media and were instructed to watch the video and complete measures that followed. First, participants were randomly assigned (via Qualtrics) to a control or experimental group and shown the corresponding video clip. Both groups then completed the measures regarding mood, body image, self-esteem, and self-objectification. Finally, participants answered the questions regarding their exposure to lesbian stereotypes in the media, included the open-ended question used for qualitative coding.

Results

As is customary with research on body image, BMI was tested as a covariate in each of the quantitative analyses below. However, the pattern of findings did not change when controlling for BMI, so the more parsimonious analyses without BMI are described below.

Reported Frequency of Lesbian Stereotypes in the Media

See Fig. 1 for the means and standard errors of each lesbian subtype and relationship stereotype assessed in the survey. A repeated measures ANOVA and pairwise comparisons indicated significant differences in the frequency of lesbian subtypes seen in the media. As predicted in H1, the portrayal of the hot lesbian (M = 3.61, SD = 1.16) was reported significantly more often than the other (5) subtypes, F(5, 177) = 35.97, p < 0.001. Additionally, participants reported that feminine lesbians (M = 3.34, SD = 1.03) were more frequently portrayed than out lesbians (M = 3.12, SD = 0.96), while out lesbians were shown more often than closeted lesbians (M = 2.88, SD = 0.92). A repeated measures ANOVA found a significant difference in the frequency of lesbian relationship stereotypes: the stereotype of moving too quickly in a relationship (M = 3.23, SD = 1.11) was reported significantly more than the other three stereotypes, F(3, 531) = 23.45, p < 0.001. No other relationship stereotypes were significantly different from each other.

Fig. 1
figure 1

Means of Lesbian Stereotypes [with standard error bars] Note. See appendix (Original Measure Items) for full text of items.

Full size image

RQ2 asked whether lesbians of color report different stereotypes observed in the media than White lesbians. Therefore, we conducted two mixed ANOVAs with one within-subjects factor (stereotype) and one between-subjects factor (race). One analysis included the six individual stereotypes and another included the four relationship stereotypes for the within-subjects factor. For both analyses, race consisted of two groups, White participants and participants of color. Specific racial groups (e.g., Black, Latina, etc.) were not analyzed due to small sample sizes that hindered statistical power. Means and standard errors for each stereotype are reported in Fig. 1 for the whole sample, White participants, and participants of color.

The mixed ANOVA assessing individual subtypes revealed a main effect of subtype, F(5, 172) = 12.92, p < 0.001, no main effect of race, F(1, 176) = 0.38, p = 0.54, and a significant interaction, F(5, 172) = 2.51, p = 0.03. Pairwise comparisons revealed a significant difference in the butch subtype, p = 0.047; White participants reported seeing the butch stereotype in the media more than participants of color. All other individual stereotypes showed no significant difference between White participants and participants of color, ps ≥ 0.10.

The mixed ANOVA assessing relationship stereotypes revealed a main effect of stereotype, F(3, 174) = 6.52, p < 0.001, no main effect of race, F(1, 176) = 3.75, p = 0.054, and a significant interaction, F(3, 174) = 5.22, p = 0.002. Pairwise comparisons showed that participants of color reported seeing the invalid relationships, p = 0.02, and vacation stereotypes, p = 0.001, significantly more than White participants, and the performative relationships stereotypes marginally more than White participants, p = 0.06. White participants reported seeing the U-Haul stereotype marginally more than participants of color, p = 0.08. Thus, some racial differences did emerge regarding lesbian stereotypes observed in the media.

Overarching Perceptions of Lesbian Media Portrayals

Mean scores on the closed-ended questions regarding overarching perceptions of lesbian media portrayals were around the midpoint of the scale, which represented “hard to say” (i.e., not positive or negative). A paired samples t-test indicated that participants believed that lesbians (M = 2.99, SD = 1.10) were portrayed more positively in the media than lesbian relationships (M = 2.84, SD = 1.14), t(177) = 2.47, p = 0.01. There were no differences between White participants and participants of color in perceptions of how lesbians or lesbian relationships are generally portrayed in the media.

In the qualitative responses, the most frequent theme apparent in over half of the participant responses was that lesbian portrayals in the media are stereotypical (56.8%), and the most frequent stereotype mentioned was that lesbian relationships are temporary (24.4%). Other common themes present in responses included that lesbian portrayals are negative (36.4%), for the male gaze (30.1%), and hypersexualized (20.5%). Though a less common response, several participants specifically mentioned the invisibility of specific stereotypes such as butch lesbians (e.g., “not enough butch dykes…”) and representative portrayals of lesbians of color (e.g., “lesbians of color…are often still hyper-sexualized or seen as exotic under the White gaze. Lesbians of color are often only relevant when they are in proximity to a White character whether that be through dating or friendship. In addition, darker-skinned lesbians in media are often portrayed as more masculine or predatory than White lesbians.”). Furthermore, several participants pointed out that they purposefully seek out LGBTQ-friendly media in order to avoid the lesbian tropes common in mainstream media. See Table 1 for full list of codes, definitions, examples from the data, as well as frequencies for each code.

Effects of Sexualized Lesbian Media Portrayal

Hypothesis 2 stated that lesbians in the experimental group who were exposed to sexualized lesbian media portrayals would report higher self-objectification, lower mood, lower self-esteem, and lower body image (as measured by appearance evaluation and body area satisfaction) than the control group. An independent samples t-test revealed no significant difference in self-objectification levels between the control (MC = 5.29, SDC = 1.16) and experimental groups (ME = 5.32, SDE = 1.04), t(177) = 0.20, p = 0.85. Similarly, t-tests conducted on differences in mood, both positive (MC = 2.35, SDC = 0.94; ME = 2.24, SDE = 0.77) t(177) = −0.85, p = 0.40, and negative (MC = 1.46, SDC = 0.64;ME = 1.36, SDE = 0.58) t(177) = −1.07, p = 0.29, yielded no significant differences. Further, there were no significant differences in self-esteem between the experimental (ME = 2.73, SDE = 0.72) and control groups (MC = 2.74, SDC = 0.73), t(177) = −0.14, p = 0.89, or between the appearance evaluation of the experimental (ME = 2.87, SDE = 1.00) and control groups (MC = 3.04, SDC = 0.99), t(177) = −1.14, p = 0.26. However, a t-test yielded a significant difference between the body area satisfaction of the experimental (ME = 2.91, SDE = 0.83) and control groups (MC = 3.17, SDC = 0.84), t(177) = −2.06, p = 0.04, indicating that the control group was more satisfied with their specific body areas than the experimental group. Thus, support for this hypothesis was limited.

Additional exploratory analyses examined whether race interacted with condition to affect the dependent variables. Because of the small sample size of minority racial groups, comparisons were made between White participants and all participants of color grouped together. No significant interactions emerged for any of the dependent variables, indicating that the manipulation did not affect White participants and participants of color differently.

Relationships between Exposure to Lesbian Stereotypes in the Media and Other Variables

H3 proposed that participants’ reported exposure to lesbian stereotypes in the media would correlate with increased self-objectification, negative mood, lower self-esteem, and negative body image. A mean was calculated for all of the closed-ended questions regarding stereotypes (i.e., six individual lesbian stereotypes and four lesbian relationship stereotypes) to represent exposure to lesbian stereotypes in the media. There were no significant relationships between higher exposure to lesbian stereotypes and self-objectification, r = −0.90, p = 0.23, self-esteem, r = −0.09, p = 0.24, body area satisfaction, r = 0.10, p = 0.18, or appearance evaluation, r = 0.05, p = 0.55. In contrast to the hypothesis, positive mood was positively correlated with exposure to lesbian stereotypes, r = 0.19, p = 0.01. However, in line with the hypothesis, there was also a marginally significant positive relationship between self-reported exposure to stereotyped lesbian media and negative mood, r = 0.14, p = 0.06. Thus, more stereotypical lesbian media exposure seemed to be related to both an increase in positive and negative mood at the time of the experiment. However, more general perceptions of the portrayal of lesbians in the media were correlated only with positive mood, r = 0.26, p < 0.001, and not negative mood, r = 0.02, p = 0.77.

Regression analyses were conducted to see if any of the above correlations varied by race. None of the interaction terms were significant, indicating that there were no significant differences between White participants and participants of color regarding the relationships between self-reported exposure to stereotypes lesbian media and self-objectification, mood, self-esteem, or body image.

Discussion

This study addressed gaps in the literature regarding lesbian perspectives on their portrayals in the media. Altogether, the findings suggest that while representations of lesbians in the media seem to be growing (GLAAD, 2019), the content of these representations may continue to promote stereotypical and sexualized portrayals of lesbian women centering the heterosexual male gaze. As hypothesized, the most reported lesbian subtypes in the media were the hot lesbian and the feminine lesbian, which is in line with previous research revealing the relative invisibility of lesbians who break gender norms (e.g., butch lesbians; Randazzo et al., 2015). Qualitative responses echoed that finding, in that participants commonly described lesbian portrayals as stereotypical, hypersexual, and performative for the male gaze. Thus, these findings add to the significant body of research on the frequent sexualization of women in the media (for a review, see Ward, 2016) and suggest that lesbianism does not shield on-screen characters from the male gaze but instead subject them more deeply to it. Similarly, the most common trope reported in the qualitative data was that of lesbian relationships as temporary explorations that later dissolve as characters return to heterosexuality. Taken together, these findings corroborate theories regarding the male-centric portrayal of lesbianism as a response to the sociopolitical threat that lesbianism poses to the patriarchal mainstream (e.g., Diamond, 2005).

Participants reported that the most commonly depicted relationship stereotype was the idea of moving too quickly in a relationship (colloquially known as ‘U-haul’ing), which is consistent with previous research emphasizing the prevalence of this stereotype (e.g., Gordon, 2006). In contrast to the hypersexual and performative portrayals of lesbians, the U-haul stereotype may better resonate with (some) lesbian women’s lived experiences. Studies regarding norms and scripts of lesbian courtship reveal that quickly moving in with their partners is unique to lesbians as a queer subgroup (Black et al., 2000; Rose & Zand, 2002). For example, lesbians show a higher percentage of cohabitating in their relationships (44%) when compared to gay men (28%; Black et al., 2000), and are more likely to emphasize the possibility of a permanent relationship in personal ads than gay men (Deaux & Hanna, 1984). Further, Rose and Zand’s (2002) study regarding lesbian dating and courtship indicated that the vast majority (69–77%) of participants upheld that distinguishing aspects of lesbian relationships included a lack of rigid gender roles, higher levels of intimacy and friendship, and rapid relationship development.

Interestingly, though there were not large racial differences in responses, participants of color were less likely to report seeing the U-haul stereotype in the media than were White participants. Participants of color also saw the butch subtype somewhat less, and the invalid and vacation relationship stereotypes more. The data collected do not offer a clear explanation as to why these subtle racial differences emerged. It may be that the participants of color are more attuned to the stereotypes that they face as lesbians of color. In addition, or alternatively, participants of color may be more likely than White participants to view media that features lesbians of color, who are likely portrayed differently than White lesbians. One of the most insightful qualitative comments suggested this was the case, pointing out that lesbians of color are often minor characters in support of more central White characters and portrayed negatively and even as more predatory than White lesbians. Indeed, women of color are hypersexualized in distinct ways and subject to racialized stereotypes. For example, Black women have been burdened with stereotypes of promiscuity and animalism (Anderson et al., 2018), and Asian American and Pacific Islander women have been subjected to exoticism and assumptions of passivity that can culminate in a racialized objectification (Azhar et al., 2021). That women of color reported seeing the invalid and vacation relationship stereotypes more and the U-haul stereotype less may reflect the intersectional marginalization of lesbians of color, as the invalid and vacation relationship stereotypes delegitimize lesbian relationships and the U-haul stereotype is arguably a more positive portrayal. To better understand the complexity of the intersection of race and sexual orientation, future research should more explicitly ask participants to report on portrayals of Black, Latinx, Asian American, and Native American lesbians, whom all face overlapping yet distinct stereotyping and marginalization.

A separate goal of this study was to test objectification theory by experimentally manipulating exposure to a sexualized portrayal of a lesbian woman and then measuring lesbian women’s self-objectification, mood, self-esteem, and body image. Most of the findings did not support Hypothesis 2. However, the control group reported more body area satisfaction than the experimental group after immediate exposure to stereotypical lesbian media portrayals, possibly because media portrayals have a more direct and immediate impact on body satisfaction than they do on self-objectification, mood, or self-esteem. It is also possible that this finding was a Type I error, given the number of statistical tests conducted in this study, so future research should aim to corroborate this finding prior to reaching firm conclusions.

Furthermore, Hypothesis 3, that exposure to stereotypical portrayals of lesbians in the media would be associated with poorer outcomes for lesbian women, was also not supported. The general lack of support for Hypotheses 2 and 3 regarding the relationship between stereotypical depictions of lesbians in the media and mental health outcomes might reflect a kind of ambivalence from lesbian audiences. Historically, depictions of lesbian sexuality have been rare, and qualitative responses highlighted that this relative invisibility persists today. Therefore, lesbian audiences might find any depiction—even a stereotypical one—empowering, while at the same time recognizing the limitations of sexualized portrayals. Furthermore, in their open-ended responses, some participants referenced purposefully seeking out alternative media in order to be exposed to more realistic representations of lesbians. Thus, those exposed to more stereotypical portrayals may also be exposed to more nuanced, realistic, and empowering portrayals. While it is not possible to tease out this relationship in the current study, future research should consider collecting data on more specific types of media that lesbians may consume. Media marketed towards the LGBTQ+ community is an actively evolving area; since the present study’s methodology focused only on one show, results could differ when using newer or more queer-centric and nuanced shows/movies. The media clip chosen for the experimental component of the study was included due to its exaggerated depiction of a sexualized and stereotypical lesbian and its impact, not based on its target audience. Including a more nuanced media clip would not have been advantageous in finding an answer to our experimental research question, which was a comparison of a sexualized stereotypical portrayal of a lesbian relationship compared to a neutral portrayal. However, future research may want to take up the question of the impact of distinct kinds of lesbian portrayals, including those aimed at queer audiences.

Previous research suggests greater body dissatisfaction in individuals exposed to sexualized media portrayals than their counterparts experiencing neutral portrayals (Grabe et al., 2008; Ward, 2016). The current study’s divergent findings could mean that these types of portrayals do not affect lesbians in the ways previous research suggests. Indeed, the research testing objectification theory in lesbian samples is quite mixed; for example, Kozee and Tylka (2006) found that correlational data from a lesbian sample was more complex than the straightforward predictions from objectification theory. Alternatively, the specific media clips chosen may have contributed to the null results. Given that, to our knowledge, this is the only experimental test of objectification theory in a lesbian sample, these hypotheses deserve further testing before forming conclusions, and findings should be published to avoid the “file-drawer problem” (Rosenthal, 1979).

Limitations

Limitations of this study include the nature of the media clips used; they may have been perceived as dated and thus not relatable, given that DH stopped airing new episodes in 2012. Additionally, the lesbian featured in the clip exemplified the hot subtype, which was necessary to operationalize a sexualized media portrayal, yet lesbian participants who identify as butch may have found this portrayal particularly unrelatable. Further, there is the lack of a validated measure to assess exposure to stereotypical lesbian media. To address this limitation, previous research was used to generate content for each item in the scale we created, and items were modeled after preexisting scales measuring media exposure in other domains. However, future research should validate a scale of reported exposure to stereotypical content in media or evaluate actual exposure to stereotypical content through methods such as tracking participants’ media intake. Given that the measure in the current study relies on self-report, it is difficult to tease apart actual exposure from other phenomena, such as sensitivity to exposure, which would lead participants to retroactively report higher levels of exposure than experienced. Another limitation concerns the collection of qualitative data at the end of the study, because participant responses could have been primed by previous sections. However, it is worth noting that many respondents wrote lengthy explanations regarding their perspectives on how lesbians are portrayed in the media, so perhaps this sample was motivated to provide a thorough report of their experiences.

The lack of diversity in our sample is a further limitation. Previous research has shown race to be a factor in both sexual objectification (e.g., Anderson et al., 2018) and queer experiences (Aranda et al., 2015; Rosario et al., 2014). Our findings did reveal some small differences between how White participants and participants of color perceived lesbian stereotypes in the media, which suggests that future research should more closely examine the role of race in the experiences of lesbian women. In particular, future research should purposefully recruit a large sample of racial minorities so that the data can be disaggregated by specific racial groups instead of grouping all women of color together, as media representation differs across specific racial groups (e.g., Black, Latinx, Asian American, Native American). A study with a more central aim of defining the intersection between race and sexuality is warranted. Additionally, experimental stimuli should include diverse representations, as a lesbian woman of color may be less likely to internalize portrayals of White women that are not relatable or seen as a source of beauty standards to this population. Overall, the intersection of race and sexuality is one which requires a more in-depth analysis than the current study was able to provide, given the data obtained.

Future Directions

Corroborating the suggestions raised by Ward (2016), future researchers should redefine how they approach media representations and media exposure. The term ‘media’ is ubiquitous and in flux, thus requiring continuous re-assessment of definitions, scales, and forms of media being studied. Future research might test the differences between using still photographs versus videos, as well as different platforms such as social media, television, and magazines. More concretely with respect to video media, an avenue for future research to explore would be using target audiences as key variable in analysis, as lesbian portrayals targeted for heterosexual audiences may differ from those targeting a queer audience. These ideas should be explored in other members of the LGBTQ + community, as each subset of the community faces a distinct set of sociocultural and intersectional hurdles. Findings and data can be used to grant more visibility to different ‘types’ of lesbians in media so lesbians as a whole feel more adequately represented and respected.

Conclusion

Depictions of lesbianism have shifted over the years from being viewed as taboo to being labeled as sexual deviance to being titillating in the context of the heterosexual male gaze, and over time they may further evolve to become an accepted and normal aspect of womanhood. Lesbian participants surveyed indicate that though the portrayals of lesbians in media are progressing, there is still much work to be done for representations to accurately reflect this diverse minority population. It is important to examine media portrayals of lesbians because they can shape how lesbians view themselves, and it is important to document lesbians’ perceptions of the media which can be shaped by their lived experiences. Media analyses should center the intersections of gender, race, and sexual orientation to understand how the varied portrayals of lesbians reflects intersecting oppressions. Targeting media representations of lesbians can have great positive effects on the way lesbians are perceived by the public, including those in power (e.g., media executives, policy makers) who can make a difference in the lives of lesbian women.