Looking heteronormatively good! Combining story completion with Bitstrips to explore understandings of sexuality and appearance

ABSTRACT This study sought to develop knowledge about understandings of sexuality and appearance by using a story completion task combined with an innovative visual methodology. Fifty-four (mainly female) participants were randomly assigned to complete a story about a fictional bisexual, lesbian, or heterosexual character who is going on a date, and then asked to build a Bitstrips online avatar of their character. Our social constructionist thematic analysis identified that looking good was a common feature of the stories and that this was understood in largely heteronormative terms. This imperative was also clear in stories of the date, which relied on traditional notions of gender and heterosexuality. Finally, individuality and authenticity were rewarded in participants’ stories, yet this was simultaneously constrained by (dominant) heteronormative narratives. We conclude that combining story completion with visual methods enabled deeper insights into representations of appearance, dating, and relationships, than a textual method alone.

homosexuality was illegal and homophobia meant lesbians and gay men were pathologized, but shared communities remain potentially safe spaces away from voyeuristic or homophobic gaze (Hutson 2010;Krakauer & Rose 2002). Physical shared spaces, such as the "gay scene," are where appearance norms around sexual identity are most readily recognised and regulated by others. In these spaces, dressing appropriately serves as a signal of shared identity, and appearance becomes an indicator of membership of the group and, therefore, of authenticity and belonging within the shared space (Clarke & Smith 2015;Clarke & Turner 2007;Eves 2004;Hutson 2010;Rothblum 1994).
The most documented and recognised lesbian and gay "looks" have traditionally been masculine/butch lesbians and effeminate gay men (Clarke & Turner 2007;Clarke & Smith 2015). However, within contemporary Western culture there are a diverse range of ever-changing visual identities (Eves 2004;Hutson 2010). These can be created through nuanced clothing choices, piercings, tattoos, and hairstyles (Clarke & Turner 2007;Hutson 2010). Lesbian and gay visual identities can also vary according to intersections of age, ethnicity, and social class (Rothblum 1994;Taylor 2007Taylor , 2008. Most recently, researchers have reflected on whether lesbian and gay appearance norms may be becoming less distinctive than they once were (Clarke & Spence 2013;Clarke & Smith 2015;Huxley, Clarke & Halliwell 2013). This has been attributed to the mainstreaming of tattoos and piercings (Huxley et al. 2013), alongside an increase in (heterosexual) metrosexual men who invest in styling and grooming in ways more commonly associated with gay men, to the extent that these practices may now be somewhat normative (Casanova, Wetzel & Speice 2016;Clarke & Smith 2015). This mainstreaming of lesbian and gay looks could be related to social and legal changes which have improved equality and promoted the acceptance of lesbian, gay, bisexual (LGB+) and trans identities (Huxley et al. 2013). In turn, this may mean lesbian and gay appearance is less identifiable than it once was (Clarke & Spence 2013;Huxley et al. 2013).
Bisexuality is often referred to as invisible and scholars have highlighted that bisexual people have been overlooked or marginalized both within academia and the wider culture (Barker & Langdridge, 2008;Monro 2015). There are seemingly no known distinctive bisexual visual identities and bisexual people may lack a recognisable "look" or looks. Instead, bisexual people may only be able to "borrow from" lesbian and gay appearance norms, or adopt androgynous or "alternative" and "funky" appearances (e.g., Goth, hippy) (Clarke & Spence 2013;Clarke & Turner 2007;Hayfield et al. 2013;Huxley et al. 2013). Therefore, bisexuality seems unlikely to be recognised through dress and appearance, even by other bisexual people .
Minimal research has specifically investigated whether heterosexual people recognise sexuality through dress and appearance. During the 1970s and 1980s, researchers seeking to explore homophobia asked heterosexual participants to assess sexuality from photographs or images. Overall, masculinity was associated with lesbians and femininity with gay men. Images rated as less attractive than others were most likely to be perceived as lesbian or gay, particularly by those with "homophobic attitudes" (see Clarke, Hayfield & Huxley 2012;Clarke & Smith 2015). Indeed, heterosexual recognition of lesbian and gay identities has traditionally been associated with negative consequences, including staring, verbal and/or physical abuse, and discrimination (Eves 2004). Hayfield's (2013) survey research identified that UK students (most of whom were heterosexual) were unable to identify any distinctive bisexual appearance norms (although some described "alternative" bisexual looks). However, they were able to describe heterosexual, lesbian, and gay appearance norms. Lesbians were described as butch and gay men as effeminate. However, participants dismissed their descriptions as stereotypes, perhaps to avoid positioning themselves as subscribing to stereotypes or as prejudiced.

Researching appearance using story completion and visual methods
In this study, we combined two innovative qualitative methods, story completion (Kitzinger & Powell 1995) and Bitstrips online avatars, to investigate contemporary constructions of bisexual, lesbian, and heterosexual looks and looking within the context of a first date. Story completion tasks (SCTs), where participants tell a story rather than presenting their own views (see Clarke et al. 2018), offered a useful tool to explore people's understandings of sexuality and appearance, without participants having to "take ownership" of their response to the scenario . Within the context of the Story Completion Research Group (see Clarke et al. 2017;Clarke et al. 2018), we wanted to explore what using visual methodology could add to the potential for SCTs to generate rich data. Few researchers who have explored sexuality and appearance have drawn on visual methods (for exceptions, see Hayfield 2011;Holliday 1999). This may be because within psychology, written and spoken text has tended to be privileged over visual tools (Frith et al. 2005;Reavey 2011). However, visual methodology logically corresponds with the exploration of visual identities. We anticipated SCTs combined with a visual element would be ideally suited to explore our research question, what are people's understandings of cultural norms of bisexual, lesbian, and heterosexual women's appearance?
Visual approaches can enable researchers to explore verbal data alongside the visual to enrich (and potentially disrupt) participants' accounts (Frith et al. 2005). Some psychologists have used photographs, while others have used scrapbooks, diaries, paintings, and drawings to explore various research topics (e.g., Frith et al. 2005;Hurworth 2004). Contemporary representations of sexuality and identity have also been explored through participants' video diaries and social media to enhance interpretations of how identities are managed and maintained (Goodings & Brown 2011;Pini & Walkerdine 2011). With the increased popularity of online avatar software at the time of the study, we saw an opportunity to combine an innovative digital visual element alongside the written story to see whether, or how, inviting participants to generate a representation of the character in their stories could complement SCTs.
The online platform Bitstrips rapidly gained popularity on the Internet in the early 2010s as a way to digitally create a virtual cartoon character, or avatar (with the option to generate short comic-strip narratives). Virtual worlds have been said to represent new sites in which to represent or perform identity, particularly in more established spaces such as Second Life, where avatars may have a coevolving relationship with their creator/occupier (Meadows 2008). New technological expressions of embodiment are becoming more commonplace (Reavey 2011), and avatars may represent a complex relationship between "real" and "virtual" identities (Serapis & Boellstorff 2008), particularly in how various aspects of identity are played out (Nakamura 2002). Some scholars met this new medium with enthusiasm given the opportunities it presented within educational contexts (Nisa & Al-Hafizh 2014;Wertz 2014), but despite calls for researchers to willingly engage with how people use new technologies (Reavey 2011), Bitstrips was seemingly not taken up by psychologists.
While we were designing this research, Bitstrips was particularly popular on Facebook, and the Bitstrips platform, with its relative ease of creating virtual cartoon characters, seemed ideal for our purpose. We were interested in how sexuality and appearance might be represented, so we asked participants to create an online Bitstrips avatar of their character after completing the SCT. Participants were therefore able to complete the study entirely online; their written response in Qualtrics and the visual element in Bitstrips.

Recruitment and procedure
Ethical approval was granted by the Faculty Research Ethics Committee at the University of the West of England (UWE). Participants were mainly UWE undergraduate students recruited using an online participant pool and awarded credit toward a research assignment. We also advertised the study via our own social networks (e.g., Facebook). When participants expressed an interest, they were directed to Qualtrics online survey software and instructed to read the information sheet that provided details about the researchers, participation, and the study. Participants created an anonymous identifier and provided their fully informed consent online.
We designed an SCT with a story stem involving a date where appearance was clearly a factor, but in which participants were not asked directly about the topic; instead they were asked to create a visual representation of the character's appearance via Bitstrips. Qualtrics was set up to randomly assign participants to a bisexual, lesbian, or heterosexual version of the following stem: Jess is a 21-year-old (bisexual/lesbian/heterosexual) woman. She has recently met someone, and they have arranged to go on a date.
The stem was left deliberately ambiguous (e.g., who the character has met and how; see Clarke et al. 2017), although the implication was that this was a first date. The stem was followed by instructions to tell a story about the run up to the date, how the character prepared for the date, how the date went, and what happened next. Participants were advised to write as much, and as far into the future, as they would like. We then asked participants to create a visual image of the character using Bitstrips and provided detailed instructions, including to copy and paste the image into a Microsoft Word document and email to the researchers. Participants answered demographic questions so that the sample could be situated.

Analysis
We analysed the data within a social constructionist framework (Burr 2003), as is common in SCT studies, particularly when researching gender and sexuality (e.g., Clarke, Braun & Wooles 2015;Frith 2013;Kitzinger & Powell 1995). Most visual methods research has focused on personal narratives within an experiential framework, where the visual is seen as a "window to experience" (Pink 2015). By contrast, our participants wrote stories and created fictional characters in response to a hypothetical scenario. We understood Bitstrips images as produced in accordance with dominant narratives and societal resources available to participants (Pink 2015). We aimed to interrogate how these images reproduced social and cultural understandings and focused on the representations of sexuality and appearance generated through text and images.
Thematic analysis (Braun & Clarke 2006) was chosen to identify patterns in the textual data alongside Bitstrips images. When data collection ended, we organised the data into three separate Microsoft Word documents (bisexual, lesbian, and heterosexual) with the associated Bitstrips for each story. Familiarisation began during data collection when we read responses on Qualtrics and looked at Bitstrips images as we received them. We took a mainly inductive approach during familiarisation and coding, where we started with the data rather than approaching analysis with any particular codes or theories in mind. However, we were inevitably somewhat deductive due to our personal and academic backgrounds and associated awareness of LGB+ appearance norms, which we were able to identity (a lack of) within text and images. We independently read and re-read the data in conjunction with the images before discussing our impressions and moving to coding the data. We identified semantic codes at the surface of the data and latent codes at a deeper level and examined how data fitted together within each sexuality grouping and across the data corpus. We considered which codes were most prevalent and how these could be conceptually organised into themes, either by "promoting" codes to themes or by collecting codes together into groups. We reviewed these themes before defining and naming them; final analytic adjustments were made as we wrote our analysis.

Results
Overall, we found heterosexual appearance norms, characterized by chasing (unattainable) standards of heterosexual beauty dominated textual and visual responses, while lesbian (or bisexual) appearance norms were recognised only to a limited extent. Gendered norms and heterotypical happy endings prevailed in these stories. While narratives of authenticity were sometimes a vehicle by which to challenge normative notions, careful negotiation was required; nonauthenticity was often punished, as were departures too far from (gender and hetero) normative dating scripts.

Looking (heteronormatively) good!
An imperative of "looking good" was a common feature of nearly all the stories, yet this was largely governed by heterosexual appearance norms. Female characters were consistently represented through traditionally (heterosexual) feminine attire, including dresses, long hair, and makeup (Gonzalez & Spencer 2015;Hayfield 2013). This was common across the dataset, whether the character was bisexual, lesbian, or heterosexual. Also, participants provided lengthy accounts of preparation rituals. Jess was described as shopping for clothes, bathing "in the glorious aromas of geranium rose and bergamot oils" (H04, Black British, 41, story/avatar of a After dragging out half her closet she has finally decided what she is going to wear, then the tedious challenge of choosing shoes. . . This brings a whole new dilemma to the situation, to wear heels and possibly be significantly taller than her date-just to achieve longer looking legs or just wear cute flat shoes. (B01, White female, 24, story/avatar of a bisexual character) Jess had a shower, shaved her legs, trimmed her other regions before moisturising everywhere. (L02, White female, 24, story/avatar of a lesbian character) Jess went into her favourite shops, topshop and River Island and eventually found a black dress. She put it with a leather jacket, and some boots. She spent a fair few hours getting ready, doing her hair and makeup. (H02, White female, 18, story/ avatar of a heterosexual character) heterosexual character), shaving, attending to hair and makeup, and making careful and deliberated decisions about what to wear: Jess's (intensive) preparation for her date and, in particular, her related accounts of "angst" suggest managing appearance for others' approval as a (hetero)normative part of dating (Pickens & Braun 2018;Tolman, Davis & Bowman 2016). Managing a (heterosexual) "look" was not always portrayed as pleasurablethe "tedious challenge" and "dilemma" of choosing shoesbut this was seemingly inherently necessary in preparation for a date. This imperative She spent hours on her hair to make sure that not even a strand was out of place and her makeup was inspired by Audrey Hepburn in Breakfast at Tiffany's. She wanted to look pretty and chic. She tried on every dress in her closet and none of them seemed to go well with the look she was aiming so desparately for. (L01, Black female, 18, story/avatar of a lesbian character) She will probably re-change her outfit three or four times to find the perfect one, spray perfume and then leave. (H10, White female, 20, story/avatar of a heterosexual character) She brushed her hair until it fell softly down her back in loose waves and applied the final application of lipstick, checking her appearance in the mirror one last time, not 100% happy with what she saw but it was the best that she was going to get. (B04, White Irish female, 21, story/avatar of a bisexual character) was identified regardless of Jess's sexuality. These appearance norms were also frequently characterised by portrayals of the character striving for perfection: For L01, the "pretty and chic" look, inspired by the stereotypically beautiful "Audrey Hepburn," was "desperately" sought, whereas for H10 the "perfect" outfit could only be decided upon through repeated clothing changes. Elsewhere, friends were consulted to confirm "the best possible choice" (H08, White female, 20, story about a lesbian character). A perfection imperative was also apparent in B04 through a wariness that she hasn't reached the "highest standard," "but it was the best that she was going to get" (B04, White Irish female, 21). The characters' tireless striving for an (unachievable) perfect appearance depicted in these stories links with (unattainable) flawless standards of beauty associated with heterosexuality (Pickens & Braun 2018).
It took several readings of the text to identify any (minimal) evidence of alternative accounts of appearance. Only five images portrayed the character with short hair, an indicator of how (heteronormatively) feminine the characters were (Gonzalez & Spencer 2015). Further, only two pictures clearly subscribed to what could be described as a "lesbian look" with "comfy shoes" and short hair (Clarke & Turner 2007). The written responses accompanying these images also placed somewhat less emphasis on the perfect appearance: Although the "dilemma" of appearance was present in L03's story, Jess's hair was short and combed rather than styled, her final clothing decision Jess's first dilemma was her appearance; Jess combed her short black hair, put on one coat of black mascara to emphasis her piercing blue eyes and smeared on some lip gloss. Then Jess was fretting about what to wear, so chose her old faithful jeans and tight white vest top. (L03, White female, 32, story/avatar of a lesbian character) At least we have agreed on doing an activity that I really enjoy. Cycling round the reservoir and taking in some pubs on the way round is much more my thing. (L20, White female, 44, story/avatar of a lesbian character) guided by reliability ("old faithful jeans") rather than perfection; in L20's story, appearance was not a dilemma: We suggest lesbian appearance norms were being drawn upon here, despite a more prominent discourse of normative heterosexual appearance and (unattainable) appearance standards. However, there were no descriptions of lesbian or bisexual characters actively resisting feminine (heterosexual) appearance practices in the textual responses, despite lesbian and bisexual women potentially being less invested in norms such as body hair removal and cosmetics than heterosexual women (Hayfield, Halliwell & Clarke 2017). Overall these participants provided relatively uniform accounts in which creating the perfect (heteronormatively feminine) appearance was constructed as an important aspect of preparing for a date, regardless of the character's sexuality.

A traditionally (heterosexual) gendered life
The stories of the characters and their dates often relied upon traditionally and heteronormatively gendered appearances, roles, and expectations. If Jess's date was mentioned, and was a woman, then her physical appearance was often described. Jess's date was "looking beautiful in a cute skater dress and heels" (B04, White Irish female, 21, story/avatar of a bisexual character), "much better looking in person compared to their picture" (B06, White female, 22, story/avatar of a bisexual character), and had "long wavy blonde hair glistening in the light" (L03, White female, 32, story/avatar of a lesbian character). When on a date with a man, his appearance was rarely described, illustrative of appearance norms governing women more than men (see Frith & Gleeson 2004 participants were not asked to create an image of Jess's date, a few did, mainly of male dates in the heterosexual scenario: These were sometimes accompanied by a description: H04 described Jess's hyper-masculinised male date in lavish detail. His hair was "short and styled in a way that enhanced his beautiful cheekbones and framed his face so perfectly" (H04, Black British female, 41). These male avatars largely subscribed to contemporary masculine appearance standards, such as beards, a slender physique, and muscles. In this sense, there was some notion of the heterosexual metrosexual man (Clarke & Smith 2015;Casanova et al. 2016) evident in depictions of the characters' dates.
Stories were also traditionally gendered when participants evoked heterosexual roles and expectations (Yancey & Emerson 2016), such as the danger of Jess "wear[ing] heels and possibly be[ing] significantly taller than her date" (B01, White female, 24, story/avatar of a bisexual character) or the protocols of who pays for dinner (Jaramillo-Sierra & Allen 2013; Lever, Frederick & Hertz 2015). In contrast, when Jess was on a date with a woman, she was sometimes said to feel "unsure about the 'rules' of girls dating girls" (B05, White British female, 28, story of a bisexual woman) and was often reported to be new to dating women. This portrayal of different "rules" is interesting, given that LGB+ dating may differ from heterosexual conventions (Lamont 2017). This may be indicative of the cultural availability of traditional heterosexual scripts versus the unimaginability of any alternative same-sex narratives. We also engaged in vertical analysis, which focused on story structure and endings (Braun & Clarke 2013). Traditional dating scripts are reported to largely revolve around the couple going for dinner and ending in the man kissing the woman goodnight (Laner & Ventrone 2000). The heterosexual stories were largely underpinned by gender stereotypes of the man mainly leading the date (Pickens & Braun 2018): Jess and Mike met at a pub [then] went to have some food. Jess had a lovely time; Mike made her laugh a lot, and he was more handsome than she'd remembered. He offered her a lift home after dinner but she refused as she said she'd like the walk. They walked in the same direction for a little while, and when it came to time to part ways, Jess was pleased that he took the opportunity to kiss her. (H05, White British female, 34, story/avatar of a heterosexual character) Although we note that participants did sometimes rely less on a dominant leading character, lesbian and bisexual stories were broadly similar to heterosexual narratives: Jess met her date in a restaurant in the early evening, they both quickly got into deep conversation that was only broken by the waiter taking their order [. . .] By the end of the night they really felt that they've "clicked". Jess and her date walk together to the taxi rank where there is an awkward moment before they kiss goodbye. (L07, White English male, 21, story about a lesbian character, no avatar) Most stories progressed with the characters continuing to date, living together, and having children (although in lesbian and bisexual narratives this was sometimes pets). Stories often clearly ended (heteronormatively) happily, regardless of the sexuality of the character: A few years later they eventually bought a house together and bought another dog. When she was 27 years of age, after being together for roughly six years, they got married and shortly after she had a baby girl. They grew old together and retired and moved to Spain. (H08, White female, 20, story about a lesbian character) Jane and Ella continued to see each other and became closer and closer until it was clear they were in love. Jane sold her London flat and they moved into a cute apartment by the Thames [. . .] One day they decided to take it to the next level. They drove to the nearest animal shelter and adopted a tiny black kitten, whom they named Bambi. A perfect addition to the household, endless entertainment, and their lovechild, as Ella would often joke. (B07, White female, 20, story about a bisexual character) Lesbians and gay men (and perhaps bisexual people) may be critical of, and subvert, heteronormative relationship practices (Lamont 2017). This can include resisting gender norms (partly due to their lack of relevance in many LGB+ relationships), being nonmonogamous, and a range of nonheteronormative living and parenting arrangements (Kitzinger & Coyle 1995;Weeks, Heaphy & Donovan 2001). However, overall these participants presented relatively homogenous stories that most closely mirrored heterosexual notions of dating and long-term relationships.

Imperatives of individuality and authenticity
The happy endings, common across our data, were often underpinned by representations of individual and authentic selves. These stories often contained some sense of individuality being maintained by Jess or her partner: Carl's silverware ridden career continued for years to come, picking up all major honour captaining club (his beloved Arsenal) and country. He [. . .] cherished the normal relationship they shared and they lived many happy years. In retirement Carl went on to manage Arsenal to further success, retiring surprisingly early to spend time with family he had with Jess. (B10, White male, 20, story/avatar/comic strip about a bisexual character) In the most visually striking depiction of our Bitstrips data, Carl's life is literally turned upside down. Yet the story featured Jess embracing his focus on a successful career as manager of Arsenal football club, a commitment to a traditionally (authentically) masculine sport (Hayfield 2013), which culminated in a happy ending of family and early retirement.
When authenticity was exercised, it was often rewarded in a variety of different forms, which sometimes departed from a heterotypical narrative. Although traditional dating scripts were most common across our data, the story endings sometimes (especially about bisexual characters) provided a contrast to dominant heteronormativity through a narrative of a (successful) authentic individual: Before anything else, Jess was an artist. Not in the "brushes & paints" kind of way no, no. Rather more she preferred to make her life her art & fill it with beauty whenever she could [. . .] so, irrespective of the brilliance of that night, with its ginsoaked, smoky kisses & pretend-Parisian style [. . .] Jess knew there would be many more of them, with many more lovers. (B03, White female, 32, story of a bisexual character, no avatar) H14's story challenges gender roles, specifically the protocol of the man paying for dinner (Jaramillo-Sierra & Allen 2013; Lever et al. 2015). Jess was illustrated with an alternative appearance and is rewarded, nonheteronormatively, for her authenticity in following a (feminist) passion to be successful as a YouTube star. B03's story also directly challenged romantic and monogamous dating narratives. In remaining true to her individuality as an artist, Jess's happy ending revolved around not settling down but rather the prospect of "many more lovers," which perhaps draws on (often negative) associations of bisexuality with multiple partners (Hayfield, Clarke & Halliwell 2014).
When it was time to pay the bill, Matt insisted on paying for everything. This made Jess angry as she felt it was demeaning to her as well as chauvinistic [. . .] she decided that men were not worth the trouble after all [. . .] Jess became a successful youtube feminist video maker who made thousands from advertising every day. (H14, Chinese male, 20, story/avatar of heterosexual character) In contrast, when appearance practices were deemed to be inauthentic, the protagonist was often punished: In H06's story, overly heterosexual appearance norms are characterised as inauthentic, her conformity for the date branded as "typical Barbie girl," and the character was ultimately punished for going against her better judgement.
Within the dating scenario we presented to participants, authenticity was typically rewarded, but only within certain parameters. When authenticity existed too far outside the boundaries of a heteronormative framework, there was a risk that the character would be punished or the date would become (heteronormatively) unsuccessful: Jess's friends tend to dress very feminine and like to look very girly, so they suggested Jess should dress like this for her date [. . .] Harry was disappointed. Harry thought Jess would look alternative, a free spirit, he was so shocked to see her look like a typical barbie girl. (H06, White female, 22, story/avatar of a heterosexual character) As much as jess wanted to meet again she realised there was no spark, and the meeting would be just as friends. Jess decided to tell them how she felt, better to be honest then drag them along. Her date, however, was not so gracious, and sent back abuse [. . .]  Her date became very arrogant and didn't like the ice skating or the magical ice kingdom Jess had booked for them [. . .] They argued all the way home and separated going different ways as soon as they were back at the station. (B18, White female, 18, story/avatar of a bisexual character) B06's account was the only story in our sample where Jess proposed friendship rather than a relationship at the end of the date, challenging the traditional (heterosexual) dating scripts (Laner & Ventrone 2000). However, instead of being rewarded for being authentic, in disrupting dating scripts Jess was punished with sustained abuse. Likewise, Jess taking her date to "the magical ice kingdom" arguably represents a date that contains individuality and an "authentic self." However, this nontraditional location exists outside the construct of a normative adult heterosexual date, almost childlike, which resulted in a (heteronormatively speaking) unhappy ending. In general, departures from heteronormative scripts were rarely rewarded, hence a danger of being too authentic or too individual, outside what appear to be fairly tightly regulated heteronormative norms (Pickens & Braun 2018). When stories differed from traditional dating scripts, happy endings became jeopardised and being too authentic represented a threat to heteronormative happy endings.

Discussion and conclusions
A key finding of this study is the minimal evidence of diversity across lesbian, bisexual, and heterosexual characters, in either textual or visual data. Our main conclusion is that the characters' appearance, appearance practices, dating, and relationships were all represented in predominantly heteronormative ways. We anticipated SCTs might free participants from socially desirable responses in relation to appearance (Hayfield 2013). Therefore, it was hoped this study could offer additional insight into contemporary understandings of cultural norms around bisexual, heterosexual, and lesbian appearance. However, stories and images were overwhelmingly dominated by the notion that all women (regardless of sexuality) invest in traditional (unattainable) feminine beauty ideals (Pickens & Braun 2018). Only a few participants represented Bitstrips characters in ways which resonated with classic lesbian appearance norms such as short hair, comfortable shoes, and "butch" looks (Clarke & Turner 2007). This indicates these (mainly heterosexual) participants may not have had the cultural capital to reproduce "lesbian looks." The lack of diversity in bisexual appearance was unsurprising given the cultural invisibility of bisexuality, and hence there are few culturally recognised images of bisexual appearance for (heterosexual) participants to draw upon (Hayfield 2013;Hayfield et al. 2013;Hayfield et al. 2014).
Overall, we conclude that heterosexual appearance norms dominated understandings of women's appearance. Our findings also link with the normalisation of lesbian and gay (and perhaps bisexual and queer) identities within mainstream culture (see below), which could mean that specific looks associated with particular sexualities are diminishing and/or becoming less culturally accessible to those who are not "in the know" (Clarke & Spence 2013, p. 25;Huxley et al. 2013). In our SCT, the setting was implied to be a formal first date between two people, so these understandings about the (re) production of appearance norms are context specific. SCT designs could be used with lesbian, gay, bisexual, trans, and queer (LGBTQ+) participants who may have different cultural knowledge, and the story could be set within specifically LGBTQ+ locations (e.g., stories revolving around a night out on the "scene"), where appearance norms relating to sexuality have historically been most relevant, recognised, and regulated (e.g., Eves 2004; Taylor 2007Taylor , 2008. Heteronormativity also dominated participants' accounts of dating and relationship practices. Participants' stories were notably similar regardless of the sexuality of the character and relied heavily on traditional (heterosexual) scripts of dating and relationships (Laner & Ventrone 2000). It was unusual for participants to provide any stories which disrupted these or indicated they had access to either the idea or reality of LGBTQ+ people "doing" dating and relationships any differently from heterosexual people (Kitzinger & Coyle 1995;Lamont 2017;Weeks et al. 2001). The uniformity across different versions of the story and the punishment of characters who deviated from heterosexual norms suggest participants' location within a culture of "homonormalisation" (Croce 2015;Roseneil et al. 2013). The term homonormalisation refers to lesbian and gay (and perhaps bisexual) identities becoming (to some extent) assimilated within mainstream heterosexual society. Scholars have been critical of homonormalisation, in part because assimilation limits the radical potential of LGBTQ+ identities to challenge traditional values and norms. Instead, particular forms of LGBTQ+ lives and relationships become tolerated, provided they are compatible with heterosexual norms (Croce 2015;Roseneil et al. 2013). Empirical research could further develop our knowledge of the implications of homonormalisation on understandings and lived experiences of LGBTQ+ appearances and identities.

Reflections on combining story completion and Bitstrips images
The few differences accorded to (mainly lesbian) sexuality were sometimes more readily apparent in images than in stories, indicating that our visual method added depth and richness to our data and analysis. Bitstrips fitted well with story completion because the focus was on hypothetical characters and hypothetical scenarios. Writing a story and designing an image may have made the study novel and appealing. Story-writing and cartoon images are potentially fun and creative, and this was sometimes reflected in the responses. Bitstrips was a relatively resource-lite visual methodology and creating an online character meant there were few ethical difficulties (e.g., confidentiality and anonymity), unlike other forms of visual data (Frith et al. 2005). There may be value in communicating research findings via visual methods, and Bitstrips characters provided a novel and engaging way to present our results.
However, there were practical and methodological limitations. Nearly one-third of participants did not complete an avatar. This may link to the detailed step-by-step guidance that participants were required to follow so they could use Bitstrips without registering. We had to instruct participants on how to create, edit, and save a Bitstrips character. They then had to capture a screenshot of their image, paste it into Microsoft Word, and email the researchers. Participants may have found this cumbersome or challenging, especially for those using mobile devices or Mac computers. Future researchers engaging with online technologies need to consider these factors.
Similar to SCTs (Clarke et al. 2018), there are questions around what visual data represents and there remains a lack of guidance around visual analysis (Frith et al. 2005). It is difficult to consider Bitstrips as anything other than supplementary to the text, which highlights the dilemmas researchers continue to face in the use of visual methods, regardless of what form they take (Gleeson 2011). There has been also scepticism around the generic nature of avatars, in particular the lack of creativity Bitstrips offered users (Cimarusti 2013). Participants had some flexibility and could choose from a range of clothes, shoes, hair (length and styles), facial hair, and accessories (e.g., scarves, necklaces, glasses). However, Bitstrips did not allow participants to portray subtle aspects of appearance (e.g., piercings, tattoos, or jewellery such as pride symbols), precisely the nuanced features that can convey sexuality within the wider culture (Clarke & Turner 2007;Hutson 2010;Huxley et al. 2013). The processes of avatar building are constantly evolving, however. Since our research, Bitstrips has become Bitmoji (www. bitmoji.com), which provides more advanced albeit limited options. The purpose of Bitmoji has evolved in scope from fantasy contexts to focus on expressions of self-identity, particularly through new communication mediums such as Snapchat (www.snapchat.com), which is an area warranting further exploration.
Our research captured a particular form of cartoon imaging that was popular at that time but only lasted for a short period. However, in being mindful of how these technologies are evolving and considering the cultural purposes around these new forms of digital visual expression, these mediums provide research opportunities. We argue that engaging with online visual methods alongside novel forms of data collection offers opportunities for both methodological innovation and increased knowledge and understanding of phenomena. Our overall finding was that heterosexuality dominated, and this was evident in stories and Bitstrips images. Therefore, this study successfully combined SCTs with innovative visual methodology to further develop our understandings of (heteronormative) cultural norms of sexuality and appearance.