Differences Among Lesbian, Gay, Bisexual, Heterosexual Individuals, and those Who Reported an Other Identity on an Open-Ended Response on Levels of Social Anxiety

Abstract

Previous research suggests that individuals with a marginalized sexual orientation report higher levels of emotional distress (Cochran, 2001; Mayer, 2003), including higher prevalence of social anxiety (Gilman et al., 2001; Potoczniak, Aldea, & DeBlaere, 2007; Safren & Pantalone, 2006) than heterosexuals. The present study builds on previous research by examining results across sexual minority identities, including an additional write-in response option. One hundred eighty individuals participated in an online study in which they indicated their sexual orientation and completed measures of social anxiety. Results indicated that in a sample recruited in a liberal urban population, lesbian/gay, and heterosexual individuals rated similar levels of social anxiety across four Liebowitz Social Anxiety Scale subscales (fear, avoidance, social, and performance; Liebowitz, 1987). Alternatively, individuals who identified as bisexual, or indicated a write-in sexual orientation rated significantly higher levels of social anxiety than the heterosexual, and lesbian/gay groups. Findings highlight the importance of offering a write-in sexual identity option, as well as looking at differences among group experiences across sexual minorities. Future studies should investigate potential group differences in social anxiety across sexual orientations in larger samples so that comparisons can be made among subgroups of the write-in response group, as well as investigate potential contributors to these group differences.

Lesbian (L), gay (G), and bisexual (B) individuals experience social and institutional prejudice and oppression in social interactions (Mays & Cochran, 2001), less social support than heterosexual individuals (Safren & Pantalone, 2006), internalized shame, and are the targets of sexual minority violence (King, et. al, 2008). Meyer et al. (2003) theorize that the disparity in social resources (social support, privilege, etc.) contributes directly to negative mental health outcomes. Indeed, LGB populations show higher psychiatric burden, including anxiety disorders, depression, drug use, and substance use than individuals who identify as heterosexual (Cochran, 2001; Cochran & Mays, 2009; Cochran, Sullivan, & Mays, 2003; King et al., 2008). LGB youth are at higher risk of suicide attempts (hypothesized to be linked to social discrimination), both nationally (Eisenberg & Resnick, 2006; Russell & Joyner, 2001), and transnationally (Fergusson, Horwood, & Beautrais, 1999; Wichstrom & Hegna, 2003).

Though there has been an increase in research on the LGB populations, there are still many gaps in our understanding. Previous research has focused largely on individuals identifying as gay, less on lesbians, little on individuals identifying as bisexual (Rust, 2002), and barely at all on those who identify as other sexual identities. Though theories on minority stress in LG populations are often applied to bisexual and other sexual minority individuals, the research may not be generalizable due to differences in the experiences of bisexual and other sexual minority individuals (Balsam & Mohr, 2007; Brewster & Moradi, 2010). Learning about the unique social experiences and the psychological burden of each group is vital to our understanding of the impacts of oppression on mental health. For example, bisexual prejudice or biphobia (e.g. stemming from biased beliefs that bisexuals are unstable in their identity, greedy, immature, and/or immoral) is experienced from not only the heterosexual population, but also from lesbian and gay populations (Brewster, Moradi, DeBlaere, & Velez, 2013; Brewster & Moradi, 2010; Burleson, 2005; Weiss, 2003). Recent research investigating a large bisexual population found that experiences of anti-bisexual prejudice were linked to lower well-being and higher psychological distress (Brewster et al., 2013).

There is even less research on individuals who identify with other sexual identities, such as queer, questioning, asexual, or fluid. Even though some studies report using an “other” or write-in response for sexual orientation, few investigate potential unique findings for this group compared to LGB or heterosexual groups (DeBlaere, Brewster, Sarkees, & Moradi, 2010; Brewster et al., 2013). Though research is sparse, a Nurses’ Health Study revealed significant disparities between individuals who indicated a an “other” category for their sexual orientation compared to lesbian, gay, bisexual, and heterosexual groups on health and psychological measures (Case et al., 2004). Specifically, Case et al. found that participants in the write-in response category had significantly worse social functioning, physical functioning, and mental health status, hypothesized to be due to the negative impacts of discrimination. The research disparity on individuals who do not identify as LGB or heterosexual is becoming increasingly relevant. The use of alternative labels continues to grow as more individuals feel that the labels LGB or heterosexual do not adequately define their sexual identities (Cohler & Hammack, 2007; Savin-Williams, 2005). Recent research suggests that when given the option, a significant number of participants choose to identify with less rigid identity categories, like “mostly heterosexual” (Vrangalova & Savin-Williams, 2012). In addition, individuals from younger generations are increasingly resisting identifying with any label or category (Olive, 2012). The emergence of label resistance might be an acknowledgement of the likelihood of sexual orientation changing over time (Rosario, Scrimshaw, & Hunter, 2011; Savin-Williams, 2005). Indeed, research on sexual attraction is more complex than static categories, as we have come to understand the fluidity of attraction over time, especially in females (Diamond, 2000; Diamond, 2009; Fahs, 2009; Mock & Eibach, 2012). Thus, to fully understand the interaction between sexual identity and mental health, we must first expand the scope of our research to also investigate the experiences of individuals who do not identify as LGB or heterosexual, or resist labels altogether.

One area in which the differences in social experiences among sexual orientations may be particularly relevant is social anxiety. Though levels vary geographically, research has shown that LGB populations experience social oppression, defined as institutionalized collective and individual modes of behavior through which one group attempts to dominate and control another in order to secure political, economic, and/or social-political advantage (Mar’i, 1988). There is a significant body of literature exploring the relationship between social oppression and levels of social anxiety disorder (SAD), which is characterized by marked and persistent fear of negative evaluation during performance situations or social interactions often due to fears of poor performance or of showing physical signs of anxiety (American Psychiatric Association, 2013). Models of SAD (Rapee & Heimberg, 1997) suggest that SAD is maintained and exacerbated by a negative feedback loop, where individuals expect that a social situation will go poorly, experience anxiety and negative thoughts throughout the situation, and then interpret their performance as negative. Thus, each situation provides further evidence that social situations should be feared and/or avoided, leaving little to no room for disconfirming evidence (Rapee & Heimberg, 1997). Considering Rapee & Heimberg’s (1997) model of SAD, we can imagine how being a member of a group that experiences actual social oppression would add tangible evidence to the negative interpretation, fueling the anxiety response and negative feedback loop, increasing SAD symptomology. In fact, individuals with “coverable” stigma (e.g. marginalized sexual orientations) are likely to experience a high frequency of stigma-related thoughts during social situations, suggesting that non-heterosexual individuals might have more negative thoughts surrounding social situations than heterosexuals (Smart & Wegner, 1999).

Indeed, higher levels of social anxiety (along with lower self-esteem and greater fears of negative evaluation) have been found in LGB populations when compared to heterosexual individuals (gay men: Pachankis & Goldfried, 2006; LGB: Gilman, Cochran, Mays, Hughes, Ostrow, & Kessler, 2001; Potoczniak, Aldea, & DeBlaere, 2007; Safren & Pantalone, 2006). Comparable research has been replicated in samples of gay man, demonstrating that higher expectations of rejection, and more frequent experiences of discrimination are linked to higher levels of social anxiety (Hatzenbuehler, Nolen-Hoeksema, & Erickson, 2008; Pachankis, Goldfried, & Ramrattan, 2008). Indeed, research has supported a link between being rejected from the majority and social anxiety (Feinstein, Goldfried, & Davila, 2012; Hart & Heimberg, 2005; Meyer, 2003; Syzmanski, Kashubeck-West & Meyer, 2008). Additionally, lower social support has been linked to higher social anxiety and other mental health struggles, suggesting that increased avoidance of social interactions may lead to less opportunities for LGB youth to engage in positive experiences, form supportive relationships, and build community (Safren & Pantalone, 2006). Given the higher prevalence of SAD in LGB populations, we must consider the combined impacts of living with both a marginalized sexual orientation and experiencing mental health struggles, as the intersectionality could potentially make both identities more difficult.

In summary, gay and lesbian populations face significant mental health disparities compared to their heterosexual counterparts. Individuals who identify as gay or lesbian also endure social oppression and stigma, which has been linked to higher levels of social anxiety. Higher levels of social anxiety occur with greater rejection sensitivity, lower social support, and personally directed experiences of discrimination. Though greater anxiety may be warranted due to perception of actual danger, anxiety levels generalize past these situations and negatively impact quality of life. Even though research on gay and lesbian populations could theoretically extend to other sexual minority groups (i.e., bisexual, asexual, queer, etc.), these groups must be specifically studied. Underrepresented sexual minority groups could potentially experience even greater distress than lesbian and gay populations, as underrepresented sexual minorities likely have even less social support due to lower community sizes and awareness of and attention to their sexual identities.

In the present study, we will investigate levels of social anxiety in participants who identify as lesbian/gay, bisexual, heterosexual, as well as those who endorse the write-in response option to see whether rates of social anxiety vary when we break down sexual identities into these subgroups. To our knowledge, this would be the first study to investigate levels of social anxiety of individuals who use the write-in response option compared with more commonly endorsed sexual orientations. This study also supports the literature arguing the importance of increasing the inclusivity of demographic forms to collect more accurate data. Further, we provide a description of individuals in the write-in responses group, acknowledging the breadth of possible sexual orientation identities present within our sample.

We hypothesize that, similar to previous research, heterosexual individuals will report lower levels of social anxiety than all other groups (lesbian/gay, bisexual, and those in the write-in response group). In addition, we hypothesize that the bisexual group will rate higher levels of social anxiety than the lesbian/gay, and heterosexual groups, due to their added minority status within the lesbian and gay communities (Brewster & Moradi, 2010; Burleson, 2005). Finally, we hypothesize that individuals in the write-in response group will rate higher social anxiety than all other groups (lesbian/gay, bisexual, and heterosexual) because of the lower public awareness of their sexual identities, as well as potential lower social support experienced by those individuals.

Method

Participants and Procedures

One hundred eighty individuals part of a university community in the northeast were recruited via a flier emailed and posted on campus. Participants consented to participate in a larger online study examining anxiety¹. Fifty-five (31.0%) individuals self-reported their sexual orientation as bisexual, 43 (24.0%) as gay/lesbian, 28 (16.0%) within the write-in response group [included: 5 queer (3%), 5 questioning (3%), 2 pansexual (1%), and 1 each (0.5%) for asexual, mostly women, free love, continuum, mainly men, choose not to define, open, and sexual. The additional 8 did check the “other” box but did not write in a qualitative response to the question (4%), and 54 (30.0%) as heterosexual². Sixty (33.3%) participants self-identified as male, 108 (60.0%) as female, and 12 (7%) as “other” (write-in responses included gender queer, no gender, rainbow warrior, and gender fuck). Participants were asked to select as many races as were relevant. Overall, 133 (71.5%) participants identified racially as White, 33 (17.7%) as Asian, 12 (6.5%) as Black, 9 (4.8%) as White Latino, 9 (4.8%) as Multiracial, 3 (1.6%) as non-White Latino, and 3 (1.6%) as Pacific Islander/Native Hawaiian, 3 (1.6%) as Native American. Ages ranged from 18–66 (M=27.53, SD= 10.40). Highest level of education attained for the sample was as follows, 15 (8.0%) of participants had a high school diploma, 110 (58.8%) had completed 1–3 years of college, 42 (22.5%) had a college degree, 18 (9.6%) had a master’s degree, and 1 (0.5%) had a professional degree.

Measures

Participants completed the Liebowitz Social Anxiety Scale – Self-Report (LSAS-SR; Fresco et al., 2001; for original LSAS see Liebowitz, 1987), a self-report measure of fear and avoidance across 24 situations likely to elicit social anxiety (i.e. telephoning in public, talking to people in authority). Respondents rate their fear on a scale from 0 (never) to 3 (severe) and their avoidance on a scale from 0 (never) to 3 (usually – 67–100%) for each of the situations. Scoring of the LSAS yields four subscales, each with a possible score of 0–36. Based on ROC curves, Mennin, Fresco, Heimberg, Schneier, Davies, & Liebowitz (2002) suggested that total scores above 30 on a given subscale are indicative of a likely diagnosis of SAD. Subscales include: anxiety in Performance (e.g. speaking up at a meeting, being the center of attention) situations in which one is the center of attention (combined fear and avoidance ratings), anxiety in Social (e.g. going to a party, meeting strangers) situations where one is interacting with others (combined fear and avoidance ratings), total Fear of all situations (combined performance and social), and total Avoidance of all situations (combined performance and social). The LSAS has demonstrated high internal consistency (all subscales had coefficients of 0.79 or higher) and good test-retest reliability (r=.83, p<.001) on all subscales (Baker, Heinrichs, Kim, & Hofmann, 2002). Convergent validity correlations ranged from.78–.85 with other measures of social anxiety including the social phobia scale and social interaction anxiety scale within a community clinical sample (SPS & SIAS; Mattick & Clarke, 1998), the social phobia subscale of the fear questionnaire in a clinical sample (Marks & Mathews, 1979), and the fear of negative evaluation scale in an undergraduate sample (FNE; Watson & Friend, 1969). All four subscales (Fear, Avoidance, Social, and Performance) in our sample had good to excellent internal consistency (Cronbach, 1951); Social (α= .92), Performance (α= .90), Fear (α= .91), and Avoidance (α= .88). These values are consistent with previous research (Baker et al., 2002; Social α= .89, Performance α= .84, Fear α= .91, and Avoidance α= .92).

Results

Participants with more than 20% of missing data (n=2) were not included in the analyses. In the remaining participants, any missing data were missing at random (determined using Little’s MCAR test) and were accounted for using a within-subject mean substitution. Skew values were as follows: Fear=0.71, Avoidance=1.01, Social=0.91, Performance=0.86. Kurtosis values ranged from 0.88 to 1.96. Initial t-tests did not reveal any significant differences on the four LSAS subscales between men and women (F’s 0.14–1.20; p’s .07–.54), between lesbians and gay men (F’s 1.24–2.82; p’s .09–.32), or between White individuals and those of a marginalized racial status (F’s .72–2.95; p’s .06–.17). Of note, individuals with a marginalized racial status scored slightly higher than White counterparts on all subscales. Means, standard deviations, and Cronbach’s alpha values for groups on the four subscales can be found in Table 1.

Table 1.

LSAS Subscales by Sexual Orientation

	n	LSAS Fear M(SD) α	LSAS Avoidance M(SD) α	LSAS Social M(SD) α	LSAS Perform M(SD) α
Lesbian	22	20.02(12.76) .92	20.48(11.73) .87	19.55(12.17) .89	20.94(12.46) .91
Gay	21	16.63(8.96) .92	15.34(7.22) .87	16.01(7.67) .89	16.03(9.35) .91
Heterosexual	54	18.21(12.00) .94	17.84(10.73) .89	18.14(12.31) .93	17.87(10.22) .89
Bisexual	55	26.27(11.85) .88	25.93(12.42) .86	27.52(13.16) .89	24.94(11.92) .90
Indicated a Write-in Response	28	26.29(11.46) .83	23.92(11.94) .81	26.53(13.24) .93	23.33(11.29) .86
Write-in Response subgroups:
Queer	5	26.72(18.86)	27.43(17.98)	33.44(19.58)	20.62(18.53)
Questioning	5	27.12(9.23)	25.18(11.80)	27.50(10.63)	22.33(10.13)
Unspecified	8	23.00(8.91)	16.71(9.86)	19.38(11.24)	21.52(8.33)
Pansexual	2	33.63(7.97)	31.76(6.73)	29.07(5.76)	36.00(8.49)
Additional Identitiesa	8	27.01(9.93)	24.85(8.19)	27.38(10.98)	24.52(8.02)

Note:

^aAdditional identities = Asexual (1), Mostly Women (1), Free Love (1), Continuum (1), Mainly Men (1), Choose Not to Define (1), Open (1), Sexual (1).

LSAS = Liebowitz Social Anxiety Scale Self-Report; α= Cronbach’s alpha value

Unspecified = individuals who selected the write-in response category but did not write in an identity label

We explored correlations of the four subscales of social anxiety within sexual orientation groups. For individuals who identified as bisexual, correlations ranged from .76 to .91. Correlations for the heterosexual group ranged from .79 to .91. Correlations for the gay/lesbian group ranged from .79 to .92. Finally, correlations in the write-in response group ranged from .65 to .91. The lowest correlation across categories (excluding heterosexuals) was between the performance and social subscales.

A series of ANOVAs were run to assess the main hypothesis regarding sexual orientation and scores on the four LSAS subscales in the lesbian/gay, bisexual, heterosexual, and write-in response groups. There was a significant difference among groups on the Fear subscale [F(3, 178) = 7.04, p<.001, η²_p = .11]. A Tukey’s post-hoc test revealed that gay and lesbian respondents did not report significantly different fear ratings than heterosexual participants (lesbian/gay: M=18.49, SD=10.98; heterosexual: M=18.21, SD=12.00; p=.99). Respondents identifying as bisexual or endorsing the write-in response did not differ significantly from each other (bisexual: M=26.27, SD=11.85; write-in response: M=26.29, SD=11.46; p=1.00), but did report significantly more anxiety than individuals identifying as lesbian/gay (bisexual: p<.006; write-in response: p=.03) and heterosexual (bisexual: p=.002; write-in response: p=.02).

Similarly, there was a significant difference among groups on the Avoidance subscale of the LSAS [F(3, 177) = 6.39, p<.001, η²_p = .10]. A Tukey’s post-hoc test of revealed that gay/lesbian respondents did not report significantly different experiences from heterosexual participants (lesbian/gay: M=4.08, SD=1.21; heterosexual: M= 4.05, SD=1.23; p=.99). Respondents identifying as bisexual or endorsing the write-in response did not differ significantly from each other (bisexual: M=4.93, SD=1.29; write-in response: M=4.74, SD=1.24; p=.91). Bisexual individuals reported significantly more anxiety than individuals identifying as lesbian/gay (p=.005) and heterosexual (p=.002). Individuals who endorsed the write-in response category did not report significantly different anxiety than heterosexual (p=.08), or gay/lesbian individuals (p=.13).

On the Social subscale of the LSAS, there was a significant difference among groups [F(3, 177) = 8.51, p<.001, η²_p = .13]. A Tukey’s post-hoc test revealed that gay/lesbian respondents did not report significantly different experiences from heterosexual participants (lesbian/gay: M=4.04, SD=1.30; heterosexual: M=4.03, SD=1.40; p=.1.00). Respondents identifying as bisexual or endorsing the write-in response group did not differ significantly from each other (bisexual: M=5.07, SD=1.36; write-in response: M=5.00, SD=1.27; p=.99). Bisexual individuals reported significantly more anxiety in social situations than individuals identifying as lesbian/gay (p=.001) and heterosexual (p<.001). Individuals who endorsed the write-in response category also reported significantly different anxiety than the heterosexual (p=.01) and lesbian/gay (p=.02) groups.

Finally, on the Performance subscale of the LSAS, there was a significant difference among groups [F(3, 177) = 4.48, p=.005, η²_p = .07]. A Tukey’s post-hoc test revealed that gay/lesbian respondents did not report significantly different experiences from heterosexual participants (lesbian/gay: M=4.10, SD=1.37; heterosexual: M=4.07, SD=1.16; p=.99). Respondents identifying as bisexual and write-in response did not differ significantly from each other (bisexual: M=4.81, SD=1.34; write-in response: M=4.70, SD=1.15; p=.98). Bisexual individuals reported significantly more anxiety in performance situations than individuals identifying as lesbian/gay (p=.03) and heterosexual (p=.01). Individuals who identified as write-in response did not report significantly different anxiety in performance situations than individuals identifying as lesbian/gay (p=.22) or heterosexual (p=.15).

Discussion

Previous research suggests that as a group, LGB individuals report higher levels of social anxiety than heterosexual individuals (Gilman et al., 2001; Potoczniak, Aldea, & DeBlaere, 2007; Safren & Pantalone, 2006). We investigated lesbian/gay, bisexual, and heterosexual groups separately, as well as participants who selected a write-in response on measures of social anxiety. Based on previous research, we hypothesized that all sexual minority individuals would rate higher levels of social anxiety than their heterosexual counterparts. However, in our sample, lesbian/gay, and heterosexual individuals rated their social anxiety at similar levels. Additionally, bisexual and individuals who endorsed a write-in response rated their anxiety significantly higher than lesbian/gay, and heterosexual individuals on multiple social anxiety subscales.

Initial correlations across each of the LSAS subscales within each identity revealed general consistency, indicating that the subscales are measuring a similar construct. Contrary to previous research, we did not find a significant difference among individuals who identified as lesbian/gay, and heterosexual. This may be due to the fact that our sample was recruited in Boston Massachusetts, a liberal city, in the first state to allow gay marriage. Though sexual minority individuals still experience oppression across the United States, specific geographic location plays a role in daily social experiences. State-level prevalence rates reveal that compared to heterosexuals, LGB populations have higher frequencies of psychiatric disorders in states without policies protecting against hate crimes and employment discrimination (Hatzenbuehler, Keys, & Hasin, 2009). Further, living in a liberal city is correlated with greater social support, and in turn lower levels of social anxiety (Potoczniak, Aldea, & DeBlaere 2007; Safren & Pantalone, 2006).

Our findings also revealed that individuals who identified as bisexual reported significantly more social anxiety related fear and avoidance and greater anxiety on the social and performance subscales compared to those who identified as lesbian/gay, or heterosexual. Research on bisexual individuals indicates lower social support compared to lesbian/gay, and heterosexual individuals (Potoczniak, 2007; Weis, 2003). Balsam and Mohr (2007) found that bisexual individuals reported greater identity confusion and lower self-disclosure and community connection than their lesbian and gay peers. Further, discrimination stemming from experiences of biphobia (rejection not only from the heterosexual community, but also the LG communities) is linked to higher psychological distress and lower levels of wellbeing (Brewster et al., 2013; Brewster & Moradi, 2010). Thus, it may be that the bisexual individuals are experiencing higher social anxiety due to lower levels of social support in addition to rejection from the LG communities.

In addition to the distress associated with being of a marginalized sexual identity, we believe individuals whose identity does not fit into one of the choices listed might incur distress from their identities not being offered on demographic forms, which is why offering a write-in response allows everyone to be represented. Individuals who endorsed the write-in response category rated similar levels of social anxiety compared to bisexual individuals on all subscales, and higher overall fear and anxiety in social situations than their heterosexual and lesbian/gay counterparts. Historically, participants have been asked to select their sexual orientation from limited options, usually heterosexual, gay, lesbian, and bisexual. However, there are many other ways people identify (e.g., queer, questioning, pansexual, asexual, demisexual) that likely come along with life experiences distinct from the LGB and heterosexual populations. Not accurately capturing how all participants identify suggests that our current data and research on sexual orientation are likely missing the complexity of experiences that may exist in our society. Indeed, research on multiracial identities suggests that asking people to “fit into a box” that does not accurately define them (i.e. not offering a write-in response option for participants to self-define), causes emotional distress and identity denial (Townsend, Markus, & Bergsieker, 2009). Though level of distress has not yet been studied in sexual minority individuals, option restriction could potentially cause similar distress in individuals who do not identify as LGB or heterosexual. Therefore, we propose offering a write-in response on demographic questions about participants’ identities allowing them to self define.

The frequencies reported in our demographic data highlight the importance of offering participants inclusive demographic variables such as a write-in response option on multiple measures of identity. In our total sample of 1080 participants, 30 participants for sexual orientation, 12 for gender, and 45 for race endorsed the write-in response option. As these frequencies and our analyses with sexual orientation highlight, it is important to offer this category so we can more accurately analyze and report data.

Individuals within the write-in response category might experience even less social support putting them at risk for greater psychological distress. Many of the terms written in by our participants in the write-in response category are identities not often discussed in United States society (such as “pansexual” and “asexual”). Perhaps having to explain one’s identity also adds a level of social anxiety and another barrier to coming out. Indeed, researchers have shown a link between greater self-concealment of personal information across various social settings and higher social anxiety (Endler, Flett, Macrodimitris, Corace, & Kocovski, 2002). Thus, more research is required to assess whether stigma around less represented sexual identities is contributing to evidenced higher levels of social anxiety.

The present results should be considered in light of a few limitations. Primarily, the data were collected as part of a larger study not designed to focus on investigating the unique experiences and contributors of social anxiety in groups of individuals that do and do not identify as LGB or heterosexual. This limitation of the data prevents us from more fully understanding contributing factors to the observed differences in levels of social anxiety. For example, we did not include measures addressing the experiences of one’s sexual identity and level of social and community support. Also, we did not have a large enough sample within the write-in response group to perform subgroup analyses between all identities endorsed, preventing us from examining potential differences among subgroups. A larger, population-based study is needed to more fully explore the write-in response group, including finding factors that contribute to higher levels of social anxiety and exploring potential differences among subgroups. Findings highlight potential differences of social anxiety across sexual minority statuses and the importance of offering as many options for sexual orientation as possible, as well as a write-in response on demographic forms. However, it is also important to consider the problematic nature of using static identity labels, as an increasing literature suggests that sexuality is fluid over time (Diamond, 2008; Stokes, Damon, & Mckirnan, 1997). Future research should explore the interactions of sexual identity, attraction, and the fluidity of both over time related to measures of psychological distress, in all sexual identity groups. While the present study specifically investigates sexual orientation/identity, it would be interesting to see how these results relate to a continuous measure of sexual attraction. Another potential limitation of our study is that we did not use a clinical sample, highlighted by the data showing that most of our LSAS means were subclinical. Though LSAS data suggest that many of our participants do not have a likely diagnosis of SAD, we feel that the variation in social anxiety symptoms does provide valuable information about group variation, especially because higher rates of SAD have been evidenced in sexual minority populations. Finally, the present sample was recruited in Boston, Massachusetts, which for the aforementioned reasons might represent a sample that is different from the modal experience of sexual minority individuals in the United States. With these limitations in mind, our results provide insight into varied experiences with in the LGB”O” community.

Our findings suggest that in a liberal, urban city, individuals who identify as bisexual and those who indicate a write-in response experience higher rates of social anxiety than their gay/lesbian, and heterosexual counterparts. These findings might be due to the exclusion that bisexual and individuals individuals face from the gay, lesbian, and heterosexual communities. Results highlight the importance of looking distinctly at the experiences of each sexual minority group, instead of grouping all sexual minorities together. These findings also support the usage of inclusive demographic forms that offer qualitative responses so that samples can be more accurately described and analyzed. Future research should investigate social anxiety in sexual minority populations in larger samples in multiple geographic locations with varying opinions on gay rights and accompanying laws and policies. Further investigation into the exclusion of bisexual and should also take place, to help to elucidate the phenomena and potentially increase support for these individuals.