Insider and Outsider Perspectives: Reflections on Researcher Identities in Research with Lesbian and Bisexual Women

In this article, we reflect on the concept of the insider and the outsider in qualitative research. We draw on our different experiences of conducting research with lesbian and bisexual women, using our PhD research projects as case studies to consider our similarities to and differences from our research participants. We highlight the impact that insider/outsider status can have at each stage of the research process, from deciding on a research topic, the design of materials, communicating with and recruiting participants through to data collection and analysis. We discuss the advantages and disadvantages of both insider and outsider positions and reflect on our own experiences. We conclude that, in reality, insider/outsider boundaries may be more blurred than the terms imply and highlight some of the ethical considerations that need to be taken into consideration during qualitative research.


Introduction
An 'insider' is a researcher who personally belongs to the group to which their participants also belong (based on characteristics such as ethnicity, sexual identity and gender), while an 'outsider' is not a member of that group (Gai 2012). The need to reflect on insider/outsider researcher positions is important (LaSala 2003;Watts 2006). Whether a researcher is an insider or an outsider is an epistemological matter because the researcher's position in relation to his or her participant has a direct impact on the knowledge that is co-created between them (Griffiths 1998). It is also important to consider the implications of being an insider/outsider because researchers have an active role in the description and presentation of marginalized voices, such as those of lesbians and bisexual women (Tang 2007). This article makes a unique contribution to the body of literature by providing two different and contrasting perspectives of insider/outsider positions. We consider how this impacted on our negotiations of the researcher/researched boundary and on our research. The aim of this collaborative article is to explicitly show our reflection processes and demonstrate the importance of reflecting on research after it has been completed. We also see this article 92 N. Hayfield and C. Huxley making an important contribution through its focus on bisexuality (an often neglected topic of focus) in relation to insider/outsider positions. Finally, we add to the existing literature on the consideration of ethical and practical issues when conducting qualitative social psychological research.
We first set the scene by outlining debates about insider/outsider research. Then we describe our research with lesbian and bisexual women, and introduce ourselves as researchers, before briefly providing an overview of the history of lesbian, gay and bisexual research. We then discuss and reflect upon our different positions as an insider and an outsider in interview research with lesbian and bisexual women.

Insider/Outsider Research
There has been much discussion as to the value and significance of both insider and outsider researcher positions, and we briefly describe some of these debates here to provide background to our own reflections. Since the 1990s, researchers have argued that insiders hold a privileged research position when conducting qualitative research, particularly when they disclose this to their participants (Perry, Thurston & Green 2004). It has been argued that the familiarity of insider status is advantageous when developing research questions, designing interview schedules, accessing and recruiting participants, and during data collection, analysis, and dissemination. It has been suggested that insiders are more aware of the lives of their participants than outsiders and are therefore in a strong position to conduct ethical research which keeps (often marginalized) participants at the top of the research agenda and represents their voices (e.g., Bridges 2001;Corbin-Dwyer & Buckle 2009;Gair 2012;Griffith 1998;Kanuha 2000;LaSala 2003;Labaree 2002;Miller & Glassner 2004;Platzer & James 1997).
However, a number of challenges that the insider researcher may face have also been identified. Participants may have high expectations of insider researchers due to their shared positions, which places responsibility on the researcher to treat their data and the knowledge that it generates in particular ways (Kanuha 2000;Kitzinger & Wilkinson 1997). Further, the breaking down of the researcher/researched boundaries may cause ethical difficulties if participants treat the researcher as a friend or counsellor, and therefore disclose more than they are comfortable with (Birch & Miller 2000;Watts 2006). While data may be richer and deeper due to the shared context of researcher/researched, assumptions of shared understandings can be problematic when collecting data (Kanuha 2000). Additionally, during data analysis there is potentially a risk of the insider overlooking parts of the data if they take-for-granted its content (LaSala 2003;Perry et al. 2004). Further, a degree of commonality does not guarantee that an insider will understand participants' perspectives any more than an outsider, especially if their lives are as different as they are similar through other personal, social, and situational characteristics which outweigh what is shared (Bridges 2001).
Some authors have suggested that the position of outsider researcher may hold some advantages. Outsiders may be able to make observations and draw conclusions that insiders could not, for example, by asking 'naive questions' to explore topics in depth, gaining valuable insight precisely because of their outsider perspective, and noticing features of the data that an insider may overlook (Hellawell 2006;LaSala 2003;Morrow 2005;Tang 2007). However, arguments have also been presented against outsider research, with the most commonly cited being that outsiders will be unable to understand or accurately represent the experiences of their participants. This is a particularly salient topic when research is conducted with those who are othered, oppressed, or socially marginalized (Bridges 2001;Pitman 2002). Therefore, the outsider must address the psychological and social distance between themselves and their participants in an ethical way to ensure that their research is culturally sensitive and that it enhances the lives of participants, and the understanding of communities and the public (Bridges 2001;Sixsmith, Boneham & Goldring 2003). We turn now to considering our own positions as insider/outsider in relation to our research.

Our Research and Researcher Positions
Dr. Nikki Hayfield's (NH) PhD research explored bisexual women's appearance and visual identities (Hayfield 2011). As part of this research, she interviewed 20 self-identified bisexual women about how they manage their bodies and appearance in relation to their (bi)sexuality (Hayfield 2011;Hayfield et al. 2013). NH considered herself primarily an insider because both her and her participants self-identified as bisexual, with many of them also being like her in age group and class category, and identifying as able-bodied and white. While some participants identified as trans, or had reflected upon their gender identities as a result of their bisexuality or their involvement in bisexual, trans, and lesbian and gay communities, most broadly considered themselves to be cisgendered. In addition some of her participants were similar to her in other ways, such as identifying as feminists, being atheist/non-religious, and so on. Therefore, her position was an emic one; in many ways she was 'one of them' and therefore likely to share similar viewpoints with many of her participants (Kanuha 2000;LaSala 2003).
Dr. Caroline Huxley's (CH) mixed-method PhD research focused on the sociocultural influences that shape lesbian and bisexual women's body image (Huxley 2010). For the qualitative component of her research, she interviewed eleven lesbian and four bisexual women about their feelings towards their body and appearance, exploring the influence of different social environments and personal relationships (see Huxley, Clarke & Halliwell 2011, 2014a, 2014b. CH identifies as a heterosexual woman, and therefore her research was primarily from an etic or outsider position. She embarked on the research without any insider knowledge or membership of her participants' communities (Kanuha 2000;Hellawell 2006). However, CH did share other characteristics with her participants; all but one identified as cisgendered, and the majority were white, able-bodied women from a similar age and class bracket. We highlight these diverse aspects of ourselves in order to demonstrate the complexity and multiplicity of identities and the ways in which they intersect (Fish 2008). Hence, to consider 'insider' and 'outsider' within a dichotomous framework is to over simplify the position of the researcher in relation to their participants because researchers are rarely one or the other (Gair 2012;Griffiths 1998).

History of Lesbian, Gay and Bisexual Research
Before reflecting further on our insider/outsider perspectives, it is first important to contextualize this discussion by specifically considering the topic of our research. Here, we briefly introduce the history of lesbian and gay research and researchers to highlight the changing focus and explicitly draw attention to the power relations that exist between the 'researcher' and the 'researched.' Traditionally psychological research on 'homosexuality' was located within a clinically-focused and positivistic framework where the researcher, who was presumed to be heterosexual, was considered the 'expert' on the topic of lesbians and gay men. Much research focused on 'causes' of homosexuality and followed a disease discourse with descriptions of pathologies that classified and stigmatized the lesbians and gay men who were the 'subjects' of these studies (Kitzinger 1999;Kitzinger et al. 1998;Kong, Mahoney & Plummer 2002). Many of those who participated in early homosexuality studies were psychiatric patients and were often depicted as sick or perverted, while heterosexuality was seen as the 'norm' (Kong et al. 2002;Tang 2007). This early research then, followed a medicalized discourse where lesbians and gay men were othered for their difference, and their nonheterosexuality was positioned as a sickness or disease. Accordingly, lesbians and gay men were seen to be in need of treatment and rehabilitation (or even imprisonment) as a normalizing process to 'cure' their homosexuality and make them heterosexual (Kitzinger 1999;Kong et al. 2002).
During the 1970s, the rise of the 'lesbian and gay movement' triggered new understandings of homosexuality, new types of research, and the removal of homosexuality from the American Psychological Association's (APA) Diagnostic and Statistical Manual of Mental Disorders (DSM) in 1973 (Kitzinger et al. 1998). The latter was largely based on an argument for tolerance toward homosexuality following Hooker's (1975) empirical evidence that homosexual men could not be distinguished from heterosexual men by expert clinicians on a range of projective tests. Based on this research, Hooker and gay and lesbian activists argued that homosexuals should not be pathologized as mentally ill (Clarke et al. 2010;Kimmel & Garnets 2003). Following the removal of homosexuality from the DSM researchers who identified as lesbian and gay themselves began to conduct affirmative homosexuality research which disrupted the 'old split between subject/researcher and the object/researched' (Kong et al. 2002, p. 243). This researcher/researched distinction was also the focus of early 1980s feminist researchers who had begun to discuss reflexivity and the need to reflect upon the relationship between 'us' as researchers and 'them' as participants (Oakley 1981). Subsequently, reflections upon the researcher/researched relationship have been of particular interest to those who belong to marginalized groups such as lesbians and gay men (Kanuha 2000;Kong et al. 2002). What has emerged from this reflexivity is the recognition of the 'native, indigenous, or insider researcher' (Kanuha 2000, p. 440). Recent research tends to be critical and politically orientated and is often conducted by researchers who are 'out' and open about their sexual identity (Clarke et al. 2010;Tang 2007).
Researchers have an active role in the description and presentation of marginalized voices (Tang 2007), and with this history of heterosexuals conducting pathologizing research, the politics of who is entitled to conduct research with lesbian and bisexual women and how to go about it are particularly sensitive (e.g., Barker et al. 2012). The issue of insider/outsider researchers is particularly salient within sexuality research, although it is by no means the only area of the social sciences where it is important. For example, Shakespeare (2006) discusses the debates that have arisen regarding who should conduct research with disabled people. Some nondisabled (outsider) researchers have been criticized for their oppressive research, which has led to suspicion of all nondisabled researchers. Subsequently, nondisabled researchers are encouraged to follow the research agendas of disabled people, to consider their own skills and knowledge carefully, and to be sensitive when conducting and reporting research. Attempts have also been made to generate research conducted by disabled people (insiders). However, debates that mirror those of insider/outsider research within social psychological sexuality studies have also taken place, around whether being similar to our participants on one axis of identity necessarily means that we automatically have greater insight into their lives. Further, those who do not occupy an insider position may nonetheless have existing friendships with those they are researching, or may share similar experiences of marginalization in another aspect of their identity (such as their gender). Finally, it is worth noting that the researcher/researched boundary will always remain, and that academic research is always constrained by the demands of the university (or setting) in which it is conducted (Shakespeare 2006).

Insider/Outsider Perspectives at Different Stages of Research
In the following section, we draw on our own experiences to consider how insider/outsider perspectives can shape the decisions that researchers make at different stages of the research process. In developing this article, we drew on the transcripts of the interviews we each conducted for our PhD research and a reflexive conversation between the two authors which we recorded during the latter stages of our PhDs. The purpose of the reflexive conversation was to share our experiences of the different phases of our research, and reflect on how our personal identities shaped our design, data collection, and analysis. As we have discussed, some authors have argued that insider researchers hold an advantaged position particularly when they disclose their position as an insider (Labaree 2002;LaSala 2003;Perry et al. 2004) while others have argued that an outsider perspective brings benefits (Bridges 2001). We reflect on our own experiences in relation to this debate, specifically considering the issue of disclosure of insider/outsider status; sometimes our insider status was apparent (e.g., our ethnicities were visible to our participants), while other identities required specific declaration (e.g., our sexual identities).

Research Design
Researchers' interests in particular topics and their reasons for conducting research are often related to their personal experiences (Tang 2007). For NH, much of the reason for researching bisexuality stemmed from her personal interest, hence she considered herself to be deeply immersed in the research through her own subjective position (Kanuha 2000). In contrast, CH was drawn to her PhD because of an interest in body image research. She was interested in exploring how lesbian and bisexual women felt about their bodies as this was an overlooked topic in the body image field, and unlike NH, she did not have a personal interest in lesbian and bisexual research. This is not to say that people who are not lesbian or bisexual should not be concerned with lesbian and bisexual issues, rather that the nature of heteronormativity and heterosexual privilege (Ingraham 1994(Ingraham , 2002 means that heterosexual researchers may not realize the importance of sexual identities until they have their attention drawn to them. CH's attention was initially drawn to lesbian and bisexual issues through a more personal interest in body image. Hence her research offered her an opportunity to move beyond a position of heteronormativity. This also raises the point that researchers will never only research issues relating to (lack of) personal privilege because of the complexities and multiplicity of identity. While some researchers will be marginalized or othered based on one aspect of their identity, they will often hold privileged positions in relation to other aspects of their identity (gender, age, bodies, ethnicities and so on). Hence, it is important to reiterate the point that the intersections of different aspects of identity points to a 'space between' that of insider and outsider. Further, in this space, researchers will never be entirely 'insider' due to the irremovable boundary between researcher and researched (Corbin-Dwyer & Buckle 2009, p. 61).
In terms of designing research questions and research materials, all researchers should refer to previous literature for guidance. However, it has been suggested that insiders hold an advantage over outsiders in terms of developing nuanced and meaningful research questions. LaSala (2003) argued that insiders are familiar with the types of issues that affect their participants' lives and therefore may be aware of pertinent questions to ask that might not occur to an outsider when developing an interview schedule (also see Labaree 2002). NH believed that her own experiences and knowledge of bisexuality allowed her to easily construct an interview schedule. In contrast, CH did not have a personal knowledge-base to inform her interview questions. However, both of us engaged in an iterative process, drawing upon previous research, asking others' opinions, reflecting upon the questions we asked, and revising our interview guide during the research. We would argue that while an insider may have an initial advantage in formulating questions, it is important that all researchers follow guidelines for designing qualitative research questions and materials (e.g., Braun & Clarke 2013) and for conducting research with their populations (e.g., see Barker et al. 2012 for guidelines for working with bisexual populations and Hale 2009 for writing about trans issues).

Communicating Insider/Outsider Status to Participants
When conducting qualitative research, researchers need to consider how they present themselves and their research to participants; providing specific information can influence participants' willingness to participate and affect how they feel about and behave toward the researcher (Richards & Emslie 2000). If researchers do not explicitly provide this type of information participants are likely to wonder about it. For example, there may commonly be an assumption that researchers who study lesbian, gay, bisexual and transgender (LGBT) topics are themselves LGBT.
Both CH and NH made their own sexual identities explicit in the information given to potential participants when they first expressed an interest in the research. As an outsider to the social identities of 'lesbian' and 'bisexual,' CH had to carefully consider what information potential participants might want to know about her as a researcher. Heterosexual researchers (and heterosexual people more broadly) are not often required to disclose their sexual identity as western society tends to be heteronormative, viewing heterosexuality as 'common' and 'the norm' and assuming that everyone is heterosexual until they are informed otherwise (Schwartz 2007). However, CH knew she would meet participants faceto-face for their interviews, so she decided to be open about her heterosexuality early on to avoid participants needing to ask, or having suspicion or wariness about her sexual identity and her motivations during the interview. This is also an ethical consideration because there was no necessity or justification for her to be covert in this research and it was therefore ethically important that she was open about her sexual identity. For these reasons, she was upfront and candid about her interest in, and motivations for studying the topic, which meant that potential participants could make a fully informed choice about whether to participate. Like previous research conducted by a heterosexual woman with lesbian participants (Asher & Asher 1999), CH found that her candor encouraged openness and trust between her participants and herself and helped to assuage any suspicions about her motives. Disclosing information about herself also helped build rapport between herself and her interviewees, which is vital during data collection (Miller & Glassner 2004).
NH believed that it was important to disclose her bisexuality and deliberated far less than CH about her decision; as an insider she simply thought it made sense to acknowledge the identity she shared with her participants. As we described earlier, some authors have suggested that being open about an insider position brings benefits. Bridges (2001) argued that an insider is more likely to be able to understand or represent participants' experiences, which can be particularly important in research with groups that have been marginalized or oppressed. For example, Pitman (2002, p. 285), an open lesbian feminist researcher, described her insider position in her research with lesbian participants: "It was more than just being 'the same'; it was that we shared an understanding of oppression." However, because of the qualitative research process, participants often gain knowledge of an insider's identity prior to participating in the research. Some of the bisexual women knew of NH's sexual identity before explicitly receiving information on her research because her recruitment techniques included snowball and opportunistic sampling via friendship networks and bisexual community events.
Many of CH's participants expressed further interest about her motivations for researching lesbian and bisexual women during their interviews (and she answered such queries honestly describing her interest in the topic). This curiosity might simply reflect assumptions that researchers conducting research with LGBT populations are LGBT themselves. Alternatively this curiosity might represent an understanding of heterosexual privilege, and the assumption that heterosexual researchers are not concerned with the experiences of LGBT people. In contrast, NH did not have to explain her interest in the topic to participants. It seemed to her that the bisexual women who took part in her study were keen to talk to someone who could perhaps be assumed to intrinsically understand why the topic was interesting and important.

Participant Recruitment
Both NH and CH used snowball and opportunistic sampling via friendship networks, which are techniques commonly used to recruit for qualitative research (Braun & Clarke 2013).
A key argument made in the literature is that insiders will find the recruitment process fairly straightforward because a person's sexual (and other) identities often affect who they socialize with (Browne 2005). Therefore, insiders may know where to recruit and have access to participants through friendship/social networks and subsequent snowball sampling (Platzer & James 1997;Labaree 2002;LaSala 2003). Accordingly, potential participants can easily 'check out' the insider researcher and their credentials. This approach was also appropriate for CH. Recommendations from existing participants promote trust in the researcher and enable potential participants to 'check out' the outsider researcher through an intermediary. Positive recommendations from existing participants also provide a researcher with social capital within the community, giving them a degree of status and credibility (Sixsmith et al. 2003).
LGBT people may consider researchers to be an intrusion unless that researcher is a member of their community, or shares their identity, and is therefore more likely to be considered trustworthy in their motives (Bridges 2001;Labaree 2002;Perry et al. 2004). In LaSala's (2003, p. 18) U.S. research with gay men, participants explicitly told him that they had taken part precisely because he was a gay man who could therefore 'be trusted to accurately portray their lives' (see also Kitzinger's research with lesbian women, 1987). CH was not a member of any lesbian or bisexual communities, so she was dependent on others to advertise her research and highlight her as an 'ethical' researcher. This ultimately influenced her recruitment process. However, not all marginalized groups have clear communities and not all researchers are active within community spaces. As an insider, NH was already aware that there were no bisexual groups in her locality at the time of recruitment, so she still had to find alternative ways (other than her group membership) to reach a hard-to-find bisexual population (Hartman 2011). She did so in ways similar to an outsider, partly through local advertising and by approaching non-local bisexual communities as a stranger. Possibly because of the lack of local bisexual groups, some of NH's participants had never met another bisexual woman and stated that they were keen to take part specifically in order to meet her.
CH had some enquiries from women who subsequently did not take part in the research. She did not get any feedback from these women, so the actual reasons that they decided not to take part are unknown. It is possible that they lacked interest or thought that it was too much of a commitment. However, it is also possible that they declined to participate because she had 'outed' herself as heterosexual. This is in contrast to NH's recruitment process where a greater proportion of those who expressed an interest subsequently participated. Insider and outsider statuses, then, can have a significant impact on the recruitment process and participants' motivations for taking part. Being an outsider has distinct disadvantages in terms of accessing potentially hard-to-reach populations, while an insider benefits from additional knowledge and an implied credibility, especially because their understanding may enable them to be more aware of ethical matters.

Data Collection and Analysis
Insider/outsider status is also relevant during data collection (Labaree 2002;Corbin-Dwyer & Buckle 2009). Insider researchers' shared identities with participants mean they are likely to be aware of (and share) interests and concerns, so participants' priorities can be kept at the top of the research agenda (Bridges 2001;LaSala 2003). Participants may believe that an insider researcher wants to improve perceptions and understandings of the group or 'rectify social misconceptions,' and therefore be more willing to share their experiences (LaSala 2003, p. 18).
Although CH was an outsider in terms of her sexual identity, both of us were insiders in that we and our participants were all women (although not all our participants were cisgendered). This is important, first because it highlights one of the ways in which the categories insider and outsider are not necessarily clear cut. Second, some feminist researchers have discussed the idea that there are often unequal power relations between 'researcher' and 'researched,' with participants answering questions, being analyzed and presented in ways often largely determined by the researcher (Oakley 1981). When women interview women, rapport can be built and these hierarchies can (to some extent) be broken down due to their shared identities (Labaree 2002). The more 'like' your participants you are, the more opportunity there is for empathy, implicit understanding, and shared experience (Hellawell 2006). This can make it easier to gain participants' trust and build rapport; therefore they may become more fully engaged with the research resulting in rich and authentic accounts (LaSala 2003; Perry et al. 2004). Despite explicit differences between herself and her participants, CH empathized with them through their shared identity as women.
Some argue that just as the insider researcher is able to generate rich data with their participants so too they are in a strong position to understand and make sense of participants' worlds (Bridges 2001;Labaree 2002). Their existing experiential knowledge, deeper understanding, and culturally specific interpretations can be utilized by the insider both during the data collection and analysis stages of the research in order to produce deep and authentic findings (Perry et al. 2004;Corbin-Dwyer & Buckle 2009). NH's data reflected her insider position; participants acknowledged their shared identities during the interview, for example, by using the term 'we' rather than 'I' (e.g., when Blue said 'why should we [bisexual women] be negated?'), and divulged their personal views and opinions willingly. Participants also asked her questions about her opinions during or after the interview, and many of the women led the interviews, allowing them to voice what was of importance to them and break down the researcher/researched boundary. This meant that NH developed her interview schedule to include other interesting and important areas (such as bisexual identity and 'biphobia'/bisexual marginalization) that these women specifically wanted to talk about, but that were not initially part of the research question (see Hayfield, Clarke & Halliwell 2014).
Despite such advantages, there is the risk that participants may respond to a researcher who is 'on the inside' by making the assumption that shared knowledge is intrinsically understood. This assumption can be problematic, because the participant's narrative may be misconstrued or misunderstood by the researcher. Assumptions of shared understandings may mean that participants leave unfinished sentences, which do not contribute to the data (Kanuha 2000). Here, Emily assumed that NH (referred to as Nikki in the dialogue) would have an awareness of the way in which some lesbians carry themselves: Emily: (I)t's the, the lesbian being butch walk, it's the like confident, butch lesbian and that's how they walk and do their swagger (laughs) [. . .] Nikki: Mm. You mentioned the swagger earlier as well (laughter). Can you describe it for me? Emily: (Laughs). Nikki: I mean, I think I know what you mean, but can you to describe it for me?
Emily then went on to elaborate on the 'lesbian swagger' (e.g., Esterberg 1996), which without NH's prompt she may not have done. This quotation provides an example of how NH, having read about the potential pitfalls of insider research, sought to clarify participants' narratives rather than assuming that they had a shared understanding. In addition to assumed understandings, it is also possible that insiders overlook interesting aspects of the data, as they may unintentionally neglect topics which fall outside of their own experiences (LaSala 2003;Perry et al. 2004). Furthermore, insiders may not adequately explore an issue with a participant, because they assume commonality in how they view the phenomenon (LaSala 2003).
In light of these criticisms, some authors have argued that outsiders may actually be in a better position to be able to see what an insider may take for granted (LaSala 2003;Perry et al. 2004). This is partly because the outsider position can make it easy for the researcher to ask 'naive' questions for clarification, whereas a participant might assume that an insider would be knowledgeable (Tang 2007). Bridges (2001) argued that culturally sensitive outsider research can enhance the understanding of the researcher, the community being researched, and the wider public. CH did not feel that her outsider position negatively impacted on her interaction with her participants or created difficulties understanding or empathizing with them. Her candidness about her own sexual identity helped to encourage openness and honesty within each interview, and because she had openly disclosed information about herself, her participants seemed willing to share their personal experiences with her. While the aim of CH's research was to gain the perspectives of her participants, she also found that they were willing to explain things that she may not have heard of, and voluntarily 'educated' her about their lives. However, it has been argued that part of the oppression of non-heterosexual people is the expectation that they are responsible for the education of heterosexual people about their sexual identities (Kitzinger 1990). Therefore outsider researchers should be aware of this ethical point and ensure that they do not expect or indeed request that their participants are responsible for educating them. Instead, researchers must be open to learning about the topics they are researching through published literature, community web pages, or by other means as well as through listening to their participants if they volunteer to share their experiences in an educative way.

Ethical Research
There may be challenges associated with being an insider for both participant and researcher (Bridges 2001;Labaree 2002). The researcher may find that responsibility is inadvertently placed on their shoulders by participants, who expect that the researcher will produce knowledge that improves the lives of their group (Kanuha 2000). Further, researchers may need to consider how to respond if participants treat them as a counsellor (Birch & Miller 2000) or if they overdisclose and feel uncomfortable after the event (Watts 2006). There is also the necessity to be especially aware of issues of confidentiality, particularly when there is the possibility of seeing research participants in circumstances where others are present (LaSala 2003). NH certainly encountered some of these challenges, for example, when a research participant whom she had previously known seemed distant towards her when they next encountered each other through shared social networks.
Over a year later, the participant commented that in her interview she had shared details of her life that she had not revealed to anyone else, which may explain why NH perceived some distance.
NH continues to feel a strong responsibility to ensure that her research represents the voices of her participants. However, authors have highlighted how it may be problematic when researchers are keen to faithfully report and validate their participants' experiences (Kitzinger & Wilkinson 1997;Watts 2006). During analysis it is not necessarily realistic or appropriate to simply be a conduit for participants' stories, and at times it may be necessary to challenge or criticize their accounts rather than simply validate and legitimate them (Ktizinger & Wilkinson 1997, p. 566). For example, NH wanted to challenge the participant who stated that cutting off her long hair shortly after 'coming out' and beginning a relationship with a lesbian women was unrelated to these life events and instead was purely a matter of practicality, rather than identity. It is also possible that the participant sensed this because she did later go on to reflect herself that she had begun to 'look a bit more gay' after she first 'came out.' This issue is one of epistemology and also applies to outsider researchers, who may feel responsibilities toward their participants to produce meaningful research that benefits their everyday lives. CH also had to be conscious of not simply replicating participants' stories but to look critically at the data. This can be difficult when researchers feel an affinity with their participants; through shared experiences of being women, CH felt a sense of kinship with several participants. Therefore, issues relating to personal feelings and a sense of responsibility toward participants can affect both insider and outsider researchers.
One aspect of CH's heterosexuality that could have influenced the dynamics within her interviews was the risk of reproducing heterosexist ideas or language. Some researchers have reflected on their own replication of heterosexism, particularly in focus groups and interviews (Allen 2006;Braun 2000). Braun (2000) highlighted how even people who are 'accepting' and 'liberal' or who attempt to challenge heterosexism may collude with it in subtle ways, such as: talk about generic 'women' becoming talk about 'heterosexual women,' and leaving heterosexist talk unchallenged. When participants were talking about other women, CH sometimes had to clarify that they were referring to nonheterosexual women, and the majority of the time they were. CH was aware of the potential problem of reproducing heterosexist language, however this need for clarification may represent heterosexist assumptions about the women her participants were referring to. Braun (2000) argued that researchers must confront the heterosexism within their research and take responsibility for it; therefore it is important to note that this is one aspect of research on which outsiders particularly need to reflect.
Given her outsider status as a heterosexual researcher, CH was keen to ascertain whether participants had felt comfortable talking to her during their interviews. Informal post interview discussions indicated that they had enjoyed the fact that a heterosexual researcher was interested in their lives. McDermott (2004) described how research interviews can provide lesbians and bisexual women with space and time to be listened to and, to a certain degree, validated. Five of CH's participants suggested that they appreciated being given the time to talk about issues that were important to them, and have their stories heard and listened to (e.g., at the end of her interview Helen commented that it felt 'like one of my therapy sessions'). This indicates that insiders and outsiders can both be appreciated recipients of participants' stories, and may face similar challenges in relation to participants viewing interviews as counselling sessions (e.g., Birch & Miller 2000).
While evidence suggests that insiders may make more nuanced and empathetic interpretations of participants' experiences, everyone's understandings are unique, and a degree of commonality does not guarantee that an insider's data and analysis will be any more or less meaningful than that of an outsider (Bridges 2001). Regardless of how these challenges are approached, it is of the utmost importance for both insiders and outsiders to ensure that participants are consulted and kept informed about the analysis and results. We offered participants the opportunity to receive a summary of the findings and the option to comment on the analysis so that they remained involved throughout the research. We also reported our findings in places other than academic sources, such as bisexual communities and associated publications (e.g., Bisexual Community News) and at publically accessible LGBT events (e.g., LGBT Health Summit; National LGB&T Partnership Lesbian & Bisexual Women's Event). Doing this ensures that the results of the study are more likely to be accessible to participants and LGBT communities, rather than just the academy.

Intersections of Insider/Outsider
During our reflections on our interviews we concluded that we were sometimes as different from our participants as we were similar. There are many subtle ways in which a researcher can be an outsider or an insider (Hellawell 2006), for example, through the intersections of sexual identity, race, class, health status, and age (Fish 2008;Tang 2007). Consequently, a researcher can simultaneously be both an insider and an outsider, and degrees of alienation and empathy can be useful qualities in research. Alienation enables researchers to critically gauge a situation, while empathy with participants enhances rapport and communication between researcher and participant (Hellawell 2006). Although CH's sexual identity made her an outsider, she was also an insider in terms of being a white, able-bodied woman like most of her participants. Watts (2006) described how an outsider researcher can build on commonalities between themselves and their participants to establish a level of insiderness in order to build rapport and aid data collection. As suggested, CH's position of both insider and outsider afforded advantages. As a woman she had shared experiences with her participants, specifically regarding mainstream social appearance pressures. In contrast, participants discussed their sexual identity in more depth than they may have done if they had assumed any commonality.
Conversely, an insider's difference may produce as many boundaries as their commonalties overcome. NH was often aware of the characteristics which sometimes (but not always) positioned her as an outsider. Her participants included non-white women from a range of socioeconomic backgrounds whose relationship statuses sometimes differed from hers in terms of longevity, legal status, partner's gender, and whether their relationships were polyamorous or monogamous.
Further, because our interviews focused on appearance, dress and body size, we reflected on our own bodies and appearances. Not only is this something which cannot be hidden from view, it is also an important element of what Burns terms 'embodied reflexivity' (Burns 2003, p. 229). She argued that while researchers have often reflected on their own contribution to knowledge production, they have neglected to consider the impact of the physical body on the researcher/researched relationship and on our interactions with participants. We carefully considered and discussed our own appearance and the identity we wished to project when recruiting at LGBT events and attending interviews. Additionally, NH became accustomed to audience comments on her dress and appearance when presenting at conferences and events. We were both more consciously aware of how/whether we embodied and expressed our own sexual identities through our dress and appearance than we had been before undertaking our research. This was particularly salient for CH as heterosexual privilege and heteronormativity means that heterosexuals are less likely to be aware that they communicate their sexual identity through their appearance, even though they may do so (see Hayfield 2013).
It is clear that participants also noticed our appearance, and whether participants read us as insiders will likely have impacted on our interactions and the data we subsequently co-created with them (Burns 2003). In NH's interviews, size was brought into the conversation by some participants, one of whom emphasized that NH's body was smaller than hers. In doing so, she positioned the researcher as different from her-an outsider. This was despite the same participant's acknowledgement of NH's insider status elsewhere in the interview; she drew attention to their shared sexual identity and following their meeting commented that she could potentially read NH as bisexual, through her dress, appearance, and piercings. These types of interactions demonstrate the complexity of insider/outsider status and indicate the importance of reflecting on how aspects of our appearance that can (e.g., clothing) or cannot (e.g., body size) be changed affect the dynamics when conducting face-to-face data collection. While none of CH's participants directed any comments to her when discussing body size, it is possible that her body affected their comfort in expressing their views. During her interviews, one participant described how she felt that female romantic/sexual partners who had the same body size as herself could understand her body image concerns more than thinner partners (Huxley et al. 2011). It is possible that such understandings translate into the research environment: if a researcher is smaller or larger than a participant, the participant may feel that the researcher is an outsider and cannot empathize with their experiences and feelings about their bodies. Therefore, we agree with Burns (2003) that embodied reflexivity is an important element to qualitative research, even for those researchers who are not specifically researching matters of physical appearance.
Intersections of our other researcher characteristics were more noticeable or applicable than others, and there were also some 'grey areas.' For example, NH resided in an urban location, but grew up in a rural area. One participant, with no awareness of NH's rural background, drew on her rural location as relevant to aspects of her and her partner's 'outness,' without realizing that NH had some insight into rural life. All of these aspects and others may be relevant in our interviews with these lesbian and bisexual women; they entered into the interviews and fed into the complexity of the insider/outsider and researcher/researched relationship.

Reflections
Part of the reflexive process is to consider how insider/outsider are not simplistic or mutually exclusive categories that exist in isolation (Hellawell 2006; see also discussions of intersectionality such as Fish 2008). Therefore, to argue for the privileged position of the insider researcher is overly simplistic (Corbin-Dwyer & Buckle 2009), and instead a more nuanced reflection on 'the complex nature of relationships between the researcher and the informant' is necessary (Labaree 2002, p. 103) and we have demonstrated this in our own reflections. The intricacies of being both an insider and an outsider mean it is critical that researchers reflect on the multiple positions and identities they hold, and consider their degree of involvement with, as well as detachment from, the research topic (Perry et al. 2004).
NH reflected upon her part in the production of knowledge by constantly considering her position as an insider during data collection and analysis. Her aims were to recognize how her insider status could enable a deeper understanding of the issues that were important to the women to whom she spoke. However, she was keen not to overlook interesting findings in the data or to let her own knowledge and personal experiences dominate participants' stories (Tolman & Szalacha 2004). CH followed McClennen's (2003) recommendation that heterosexual outsider researchers should reflect on their position and adhere to guidelines for good practice in order to conduct culturally sensitive research with LGBT populations (see Barker et al. 2012;Hale 2009). Such guidelines include avoiding research questions that stereotype and stigmatize LGBT people or that assume behaviors can be explained solely by their sexual identity; avoiding heterosexism; and understanding difference as difference, not as being indicative of deficiency (Clarke et al. 2010).
When reflecting on research it can also be useful (with the appropriate ethical permissions) to show others your data and to discuss it with them, as they may read your transcripts from a different perspective and offer additional insights, as our supervisors did during our PhD supervision. Researchers may also find their own reflexive tools. For example, as we mentioned above, part way through our individual research projects, we conducted a reflexive conversation where we interviewed each other about our own contrasting insider/outsider status. This discussion was an effective way to see our own research from a different perspective, and to reflect on how our position affected our research. It has also proved a useful resource to draw on since, for example it inspired and informed the writing of this paper. Furthermore, both insider and outsider researchers should not only reflect on their motives for conducting the research but also communicate effectively with their participants (without being alienating or patronizing) and ideally produce research that can lead to meaningful outcomes for people (Clarke et al. 2010).

Conclusion
Both NH and CH felt strongly that their position of insider or outsider brought a number of benefits to their research. However, we conclude that to see oneself as purely an insider or an outsider is to oversimplify the complexities of researchers' relationships with their participants. Boundaries between researcher and participants are often more nuanced than they may first seem. Further, while both participants and researcher contribute to data production we always remain researcher/researched and this boundary cannot be removed even when researchers are members of the same (stigmatized) groups as participants (Corbin-Dwyer & Buckle 2009). The nature of conducting research versus participating in research will always remain (Richards & Emslie 2000).
In this article, we have explicitly shown our reflection processes, and demonstrated the importance of reflecting on research after it has been completed. This uniquely contributes to the literature by providing two different perspectives of insider/outsider positions and how this impacted on our negotiations of the researcher/researched boundary. We have also contrasted our different researcher positions and examined how these impacted every stage of our research. In conclusion, our reflections on our research have demonstrated the complexity of insider/outsider researchers, and the importance of considering how researcher/researched status shape the production of knowledge throughout the research process.