Go-along interviewing with LGBTQ youth in Canada and the United States

Abstract

Go-along interviews, which are interviews conducted whilst being in and moving within participant selected spaces, were conducted with 66 LGBTQ adolescents (14-19 years old) in their self-identified communities to explore perceived community attributes, including safe spaces, resources, and supports; this paper highlights methodological lessons learned. Successful recruitment in two countries and varied community settings required partnership with local LGBTQ supporting agencies, including school-based Gay Straight Alliances. Youth chose to walk, use public transportation, and drive to community locations, identifying numerous formal and informal resources in their communities. Participant reflections highlighted that go-along interviews can be conducted in safe ways that encourage LGBTQ youth to express themselves about communities in which they live, study, work, play, and relax.

Introduction

Lesbian, gay, bisexual, transgender, and queer (LGBTQ) adolescents experience typical developmental milestones that can be easier to traverse with the presence of protective factors such as supportive environments and social structures. Unfortunately, not all LGBTQ youth have adequate positive supports when compared to their heterosexual counterparts. LGBTQ youth experience health and social disparities, including fewer social supports (e.g., caring adults or teachers) (Eisenberg & Resnick, 2006; Saewyc et al., 2009), higher rates of risk behaviors, and poorer health outcomes (e.g., cigarette smoking (Corliss et al., 2010), alcohol and drug use (Corliss et al., 2010; Herrick, Marshal, Smith, Sucato & Stall, 2011; Marshal et al., 2008; Newcomb, Birkett, Corliss & Mustanski, 2014; Saewyc, 2011), sexual risk behaviors (Saewyc, 2011), depression (Haas et al., 2010), suicidality (Haas et al., 2010; Marshal et al., 2011; Saewyc, 2011). Experiences of stigma and maltreatment likely contribute to these disparities (Almeida, Johnson, Corliss, Molnar, & Azrael, 2009; Bontempo & D’Augelli, 2002) and have been linked to a larger societal environment that lacks actual or perceived supportive qualities (Saewyc, 2011; Saewyc, Poon, Homma, & Skay, 2008; Saewyc et al., 2006). Indeed, societal values and policies have macro-level influences on a young person’s health and well-being, as do influences including school climate, peer relationships, or family connectedness (Hatzenbuehler, 2011; Hatzenbuehler, Wieringa, & Keyes, 2011; Kosciw, Greytak, Bartkiewicz, Boesen, & Palmer, 2011; Kosciw, Greytak, & Diaz, 2009; Needham & Austin, 2010; Sallis, Owen, & Fisher, 2008; Toomey, Ryan, Diaz, & Russell, 2011).

School climate for LGBTQ youth (e.g., Gay Straight Alliances (GSAs), anti-bullying policies, supportive school staff) has been researched, with clear relationships established between unsupportive climates and bullying, academic performance, depression, and substance use (Goodenow, Szalacha, & Westheimer, 2006; Heck, Flentje, & Cochran, 2011; Russell, Kosciw, Horn, & Saewyc, 2010; Russell, Ryan, Toomey, Diaz, & Sanchez, 2011; Toomey et al., 2011). For example, Hatzenbeuhler (2011) found that LGB youth were at significantly lower risk of attempting suicide if they lived in an area with more GSAs, protective school policies, and a more progressive political environment than those living in less supportive environments, even after adjusting for other known risk factors. Similarly, Saewyc, Konishi, Rose, and Homma (2014) documented that the benefits of a supportive school environment might also spill over to heterosexual students; this group was found to experience lower odds of anti-gay discrimination, suicidal thoughts and suicide attempts in schools with GSAs or supportive school policies compared to schools without these supports.

Research on how school environments influence LGBTQ health has been more common than studies focusing on community-level environments (Russell, Seif, & Truong, 2001). One of these few studies documents that sexual minority youth in rural areas report similar, but also different, concerns about safety and health than their urban counterparts (Poon & Saewyc, 2009). This indicates a need to better understand important community-level supports and risks experienced by sexual minority youth residing in rural areas and small towns. Rural locations are typically characterized by greater political conservatism, and fewer and more diffuse resources (of any type); this combination of factors is likely to contribute to substantial differences in the life experiences of LGBTQ youth in urban versus rural contexts. The social environment is best understood by hearing from those who are residing in and experiencing the environment, namely, the youth themselves. Some researchers have relied on close-ended, multiple choice online and school-based survey techniques to assess school acceptance and climate; while these studies have yielded important findings about the demographics, experiences, and feelings of youth (for example, see Taylor et al, 2011; Kosciw, Greytak, Palmer, & Boesen, 2014), there are additional insights to be gained using complementary data collection techniques, such as go-along semi-structured interviews, particularly with respect to understanding and strengthening community-level supports in the lives of LGBTQ youth.

Novel qualitative methods that access social spaces have the potential to build on existing knowledge and to provide additional contextualized insights that inform strategies, programs, or policies that encourage community support and minimize risks. Studies that use open-ended data collection techniques to capture LGBTQ youths’ voices about their environments, including community-level supports or risks are particularly rare. Furthermore, common qualitative data collection methods (e.g., focus group, stationary interview) are not necessarily versatile enough to answer specific research questions about community environments, or in particular, to explore contextualized community-based experiences of vulnerable populations. For example, stationary interviews, conducted in a private setting, inherently miss visual cues because the participant is not prompted by cues that might present when they move through the environment (i.e., neighborhood) (Garcia et al, 2012). Focus groups are commonly used to elicit in-depth insights, but for some youth, particularly those who have experienced stigmatization, a group discussion format might not be preferable, especially for LGBTQ youth who are working through processes of coming out about their sexual orientation and/or gender identity. Focus groups pose additional challenges to moving around (some researchers have attempted this), and are most often conducted stationary. Certainly, there are important qualitative studies recently conducted with LGBTQ youth that have explored social influences using stationary oral/written storytelling narrative techniques (Bond & Loewenstern, 2014; Harper, Brodsky, & Bruce, 2012; Hillier, Mitchell, & Ybarra, 2012; Olive, 2012; Trocki, Michalak, & Drabble, 2013). Go-along interview methods present opportunities to overcome the constraints of stationary qualitative techniques and expand the utility of open-ended interviewing, as well as increases youth ownership and direction over the data collection experience.

This go-along method facilitates contextualized understanding by asking interview questions whilst being in and moving within participant selected spaces (Carpiano, 2009; Kusenbach, 2003; Thompson, Cummins, Brown, & Kyle, 2013). Unique insights can be drawn from observing and experiencing the spaces and resources that LGBTQ youth are describing in response to interview questions (Carpiano, 2009). Go-along interviews have been conducted to physically and virtually explore perceptions of physical (DyckFehderau, Holt, Ball, & Willows, 2013; Frankova, Woodcock, & Dunham, 2013; Pawlowski, Tjørnhøj-Thomsen, Schipperijn, & Troelsen, 2014; Thompson et al., 2013) social (Niland, Lyons, Goodwin, & Hutton, 2014), and environmental factors including barriers, risks, and assets (Neary, Egan, Keenan, Lawson, & Bond, 2013). For example, our previous research used a go-along approach to understand college students' perceptions of sexual health resources on campus, including condom distribution programs, sexual violence prevention and support resources, and health clinic services. Findings demonstrated the utility of this method for obtaining place-related information from young people, even on sensitive and personal topics (Eisenberg, Garcia, Frerich, Lechner, & Lust, 2012; Garcia, Eisenberg, Frerich, Lechner, & Lust, 2012). Specifically, college students moving around their campuses were triggered by visual cues to discuss atypical sexual health resources, such as a personal trainer in the gym (Eisenberg et al., 2012). Conducted on foot, by car, or using public transportation, go-along interviews facilitate firsthand observations and promote participant recall as they navigate through spaces whilst responding to interview questions (Bergeron, Paquette, & Poullaouec-Gonidec, 2014; Garcia, Eisenberg, Frerich, Lechner, & Lust, 2012; Oliver et al., 2011; Sunderland, Bristed, Gudes, Boddy, & Da Silva, 2012).

The purpose of this study was to determine the utility and appropriateness of go-along methods for the qualitative study of LGBTQ youth populations; we describe methodological lessons learned from a bi-national, multi-site study designed to systematically collect information from LGBTQ youth in their self-identified neighborhoods. Specific issues include IRB approvals, recruitment strategies, enrollment processes and data collection options, with attention to issues of working across locations and in different types of communities. Detailed findings regarding LGBTQ youths’ perspectives on safe and positive environments will be the focus of other publications.

Theoretical Underpinnings

To contextualize this project, it is important to describe the disciplinary and theoretical underpinnings of the research team. Our team represents diversity in disciplines including nursing, public health, sociology, anthropology, psychology, social work, and family social science. As such, we collectively adhere to ways of knowing that are constructivist, asset-based, and grounded in experiential appreciation of meaning as ascribed by the individual, community, or society (e.g., symbolic interactionism). Further, we are informed by critical and feminist theories, within the context of a socio-ecological model that organizes understanding influences on health, including risk and protective factors according to proximity (e.g., individual-, micro-, meso-, and macro-level influences). Our team collectively aims to advance understanding of multi-layered, complex influences on health that translates into evidence-based, relevant interventions, programs, and policies.

IRB Approval

Mustanksi (2011) describes unique ethical issues for conducting research with LGBTQ youth populations, and notes the challenges involved in obtaining IRB approval. In particular, studies with this population typically request a waiver of parental consent in order to minimize the risks of harm for young people who have not disclosed their sexual orientation to their parents. Because research activities for the present study presented minimal risk, a parental consent waiver was requested from each of our institutions (University of Minnesota, University of British Columbia, San Diego State University) to protect adolescent participants and minimize risks potentially associated with parental awareness of their involvement in a study recruiting LGBTQ adolescents.

Each participating academic institution’s IRB separately reviewed the study protocol. Although similar human subjects protection procedures were initially proposed, there were notable differences across the respective IRBs in what was required to provide assurance of participant protection, specifically in terms of possibly moving around the community in a vehicle. One institution approved a partial waiver of parental consent with two specific provisions. First, interviewers were required to ask if participants under age 18 were “comfortable” asking for parental consent to participate, and parental consent was required in cases where the participant was comfortable. If a participant was not comfortable seeking parental consent, a pre-determined “youth advocate” (e.g. the GSA advisor or community youth group leader) was to be offered as a resource, so that youth could confer with an adult regarding whether participation was in their best interest, as recommended by Mustanski (2011). Second, driving or riding in a car with study staff was deemed “greater than minimal risk” to participants, and parental consent was required for all interviews involving driving as the method of moving through the community. Of the 17 minor participants in this location, 16 were comfortable seeking parental consent, and this consent was obtained. Safety precautions were put into place to address the potential risks associated with the interviewer and participant being alone in a private space and/or vehicle; these were similar to any interview research, whether stationary or mobile, and included notifying a study team member of scheduled interviews and locations, and communicating via text as soon as an interview was completed. The IRBs concurred with the safety precautions that were put in place.

Recruitment

Our recruitment strategy was informed and aided by organizational partners working closely with the LGBTQ youth-serving community throughout each state or province. The three methods of recruitment we used employed direct and snowball sampling via: (a) community organizations, (b) schools, and (c) community events for LGBTQ youth. School principals, sponsor teachers of GSAs, and key contacts in LGBTQ youth-serving community organizations shared information about the study verbally and in fliers. Most, but not all, of the recruitment challenges were similar to those experienced in any research study, including difficulty recruiting during the holidays/winter, research fatigue in urban areas near research universities, and school district hesitancy to support the study because it involved engaging with youth off the school premises.

Two challenges were specific to using the go-along method with LGBTQ youth. First, adult contacts shared that youth in rural settings were extremely vulnerable to hostility in the community, and participating in a study of this type - even being seen in public talking with an unknown adult - could jeopardize their safety. In some settings, support group meetings were held at secret locations with leaders who were not publicly identified, making recruitment impossible. A second challenge was that some GSA leaders were particularly protective of their LGBTQ students, perceiving them as vulnerable, even in more supportive urban areas. In spite of the support of our community partners, some school-based adults were unwilling to share our study information, thereby limiting access to their adolescent participants.

Key community partners and stakeholders provided feedback that youth would feel more comfortable meeting the interviewers in advance of the interview. Our most successful recruitment involved some kind of face-to-face interaction with research staff, all of whom were relatively young, female identifying graduate students. It is possible that seeing who might be conducting the interview fostered interest by reducing the number of unknowns related to participation. Additionally, study staff members could provide more specific information about the study, answer questions, and speak enthusiastically about the project in ways that community partners could not.

Participants

Interested participants completed an initial intake form in person at the place of recruitment or over the phone with research staff, to confirm eligibility based on sexual orientation/gender identity and age, as well as to collect information such as race/ethnicity, and preferred pronoun(s). Research staff presented details about the study (e.g., purpose, process, risks and benefits) and answered any questions from potential participants. All youth were given similar information upon enrollment, and at the onset of the interview during the assent/consent process. All youth were encouraged to think about places in their community they would like to highlight during the phone screen conversation; the extent to which youth preplanned the experience was personal preference of each youth.

We successfully recruited 14 to 19-year-old adolescents (mean age = 16.6, N=66) from urban (n=19), suburban (n=22), smaller city (n=9) and rural locations (n=16) within the three sites. Effort was made to recruit and enroll participants reflecting diversity in age, race/ethnicity, gender, and sexual orientation characteristics.

We successfully recruited participants with a variety of sexual orientations and gender identities. Youth were asked to provide their own descriptors of these characteristics, which are summarized in Table 1. Furthermore, participants were diverse with regards to demographic and personal characteristics. Slightly more than half of the participants identified European ancestry (only), about a quarter reported a mixed ethnic background, and other participants identified as Latino, Aboriginal/Native American, African/African American, Asian or Middle Eastern.

Table 1.

Self-descriptors of sexual orientation and gender identity (n’s)

	Female	Male	Trans and additional labels^	TOTAL	Mean age (rows)
Gay or Lesbian	8	13	3	24	16.5
Bisexual	8	10	3	21	16.4
Queer and additional labels ~	5	1	13	19	16.8
Straight and other *	0	0	2	2	17.5
TOTAL	21	24	21	66
Mean age (columns)	16.3	16.4	17.1

^{^}"trans" included n=11 whose self-descriptor included "trans" (e.g. "trans-female," "non-binary trans person"; additional descriptors included n=10 who provided various labels, e.g. "genderqueer," "fluid," "non-binary" or "neutral"

^~additional descriptors included n=7 "pansexual," n=1 "asexual," and n=1 "panromantic asexual"

^*n=1 "straight" and n=1 "other"

Data Collection

Interviewers met participants at a location the youth identified during the screening process. These usually included public places such malls, schools, fast food locations, bus stops or at the participant’s home. Participants were asked an initial question of where they would go in their community to get something to eat, to help demonstrate the nature of the go-along interview. From there, participants who had a list of places they wanted to show to the research staff would make a plan as to the best way to visit those places, and participants without a plan ahead of time would be asked where they would recommend another LGBTQ youth go for fun.

Because the go-along methodology is designed to encourage comments inspired by visual cues, youth were prompted to reflect on spaces that were passed or to take detours to explore additional ideas that came to mind during the interview, even if the route had been pre-scripted by the youth. To encourage youth ownership of the experience, research staff accommodated to the youth’s style; more prescriptive youth made plans ahead of time, whereas others organically wandered around their neighborhood. For example, one interviewer shared, “I think I only had maybe four youth who were really motivated to show me places. The majority would need time to think about it before the interview, would change their mind during the interview, and overall, they often didn't seem really committed to showing me anything in particular.”

The youth individually determined how they wanted to undertake the “go-along interview”; this was likely the result of personality-based and comfort-level preferences. For example, during our data analysis process, we recognized that the eight participants who chose stationary interviews included two distinct groups of youth, namely, well supported youth who chose not to access or visit supports because they felt they did not need them and youth who felt generally unsupported and unsafe in their communities. This finding demonstrates one example of the diversity of need, support, and experiences among a sub-group of LGBTQ youth in our study, and presents another area of need for further research.

Youth were asked about accessible resources and prompted to discuss a range of services such as drop-ins, clinics, community centers, and mental health practitioners. Questions also elicited descriptions of feelings and memories of particular experiences they or their friends had in certain areas. Youth were also asked to describe their participation in their school and community activities (e.g., Gay Straight Alliance if available). To conclude the interview, youth were asked to identify what they felt was missing and needed to support LGBTQ in their communities. They also ranked their most important supports, and identified what support meant to them.

In responding to these questions, youth visited a variety of formal and informal resources during their go-along interviews. Participants and interviewers entered some locations; however, some spaces were closed at the time of the interview or participants did not want to enter the space with the interviewer. Formal resources included LGBTQ youth-serving organizations, other youth centers, health care providers, schools and other organizations. Informal resources were also commonly visited, such as coffee shops, fast food outlets, shopping centers, public parks, and queer friendly neighborhoods. The distance traveled during interviews ranged considerably, based in part on the mode of transportation used (described below). In British Columbia, most interviews took place within one mile of the initial meeting place; in contrast, Minnesota interviews covered as much as 31 miles.

Participants were given the option of walking, driving or taking public transit to their desired locations (see Table 2). Choices across locations reflected typical transportation patterns in that area, density of resources in a given location and season/weather considerations. All six interviewers from the three sites noted in their post-interview reflections that walking or driving made it easier to build rapport with interview participants. One interviewer shared, “I do find that it sometimes helps to ‘walk and talk’. It definitely makes the interview feel less formal and helps the participant relax some.” Another similarly noted, “it [walking/driving] did make the interviews feel more relaxed and natural.” Additionally, unlike face-to-face methods of data collection, walking or diving allows participants to be side by side with interviewers and more naturally avoid direct eye contact, which seemed to also put youth at ease. Every interviewer who conducted go-along interviews while driving, however, acknowledged that it was difficult to focus on driving and engaging in the interview at the same time. Allowing youth to decide the location and distance travelled during the interview helped establish rapport with the youth and maintain their sense of safety.

Table 2.

Go-along interviews by study site and modality of movement

	MN	MA	BC	Total
Included Driving	17	5	0	22
Included Walking and/or public transit	4	13	19	36
Stationary	3	1	4	8
Total	24	19	23	66

Youth displayed varying interest in the go-along process. Five of the participants chose to spend most of the interview in a single location such as a coffee shop or park, rather than visiting multiple locations because of time constraints, a lack of knowledge about resources, or they believed the places they could show were not very interesting. Of those who chose to move around the community, 25 took researchers to multiple specific locations, 14 wanted to visit a specific area (e.g. a neighborhood), 18 chose to wander during their interview with no specific locations in mind, and 8 chose to have a stationary interview. In future studies employing a go-along methodology, researchers who really want to be shown environmental attributes, or require exploring specific places, should consider making this expectation clear at the time of enrollment, perhaps asking youth to identify a certain number of concrete places to visit, for example.

Several youth volunteered feedback on the interview experience itself. Participants generally expressed satisfaction with the interview, with some reporting very positive experiences. A Minnesota youth shared in response to a question of how the interview went, “It was very good. I felt like I was in control of it, which is I think what you were going for. Yeah, I felt open to discuss…I think you’re going to be able to help a lot of kids express what they weren’t able to with other people.” Similarly, an adolescent male in British Columbia shared intention to post the following statement to Facebook after the interview: “Wow, this interview has been really empowering.” No participants verbalized dissatisfaction or appeared distressed during or at the conclusion of their interview.

Key Lessons Learned

Importantly, we found that, whenever possible, it was useful to recruit multiple adolescents from the same geographic areas to gain confidence that the depth and variability of experiences with community resources was captured across participants. There were some adolescents in an area who were unaware of existing resources that had been identified by other participating youth; if we had not recruited enough youth the credibility and transferability of the go-along interview data could be questionable (Golafshani, 2003; Shenton, 2004). It is a unique interview focus and process on environmental triggers that requires consideration of how data saturation is assessed and reached. Similar to traditional interview techniques, we sought participants until we were sure we were hearing about the same kinds of resources repeatedly. Without adequate recruitment numbers, one runs the risk of not representing adequately the strengths and risks in a physical community.

Technically, the go-along interviews require attention to logistics such as wearing a lapel microphone connected to the audio recording device. This could be bothersome for some participants; our study demonstrated that all of the adolescents were comfortable wearing the recording device and moving through their communities. For the participants who might be uncomfortable there could be alternative strategies used to encourage participation and maintain comfort; for example, an interview could be conducted via skype or google hangout with visual capabilities that allow the participant to control what is shared and showed and what is not. This was not tested in our study but is another opportunity for future research.

Strengths of Go-along Interviews for Assessing Socio-cultural Environments with LGBTQ Youth

The go-along interview structure appeared to be empowering for many of the adolescents who participated; this is an important strength to the methodology, particularly for an LGBTQ population that has typically been disempowered. In general, adolescents appreciate developmentally appropriate tasks that yield a sense of empowerment (Cargo, Grams, Ottoson, Ward, & Green, 2003; Nation, Vieno, Perkins, & Santinello, 2008); the adolescents in our study were in control of what they chose to show the interviewer. While participants in any type of study are necessarily in control of what they choose to share or not to share, the go-along interview method might afford additional opportunity to experience and perceive that empowerment and control, because the participant controls more pieces of the process including not just what is said, but also where it is said. It is possible, however, that the eight youth who participated and chose to remain stationary or visited few locations also felt a sense of empowerment in the interview process or instead, experienced greater comfort in being stationary; questions to explore these feelings were not specifically asked but could be a valuable aspect of future research employing go-along methods.

An additional related strength is the increased capability of participants to respond to visual triggers that might not be described in a traditional interview format confined to a concrete space. Instead, participants who visited community spaces were able to point out physical items, such as posters or park green space, that might not have come to their minds in the absence of walking or driving by those entities. In particular, participants pointed out many resources that were not necessarily LGBTQ-specific – and may not have been brought to mind by our interview questions – but were noted as safe, unsafe, or relevant to their experience as an LGBTQ adolescent in some other way. What resulted was broad, rich insight into community experienced by these adolescents. Our study did not conduct traditional interviews so direct comparison of these methods is not possible but is recommended as an area of further methodological research.

Finally, the go-along methodology expanded not only what was observed and identified but also the discussion of those resources, assets, or risks. For example, many adolescents could identify the important presence of LGBTQ visual cues such as symbols (e.g., rainbow flag, pink triangle). Beyond these, however, the adolescents discussed representations of safety as they moved through their communities, including people (e.g., peers, adults, GSA advisors). Sometimes these discussions were directly triggered by seeing people moving about a specific space. This is important to understand and explore because, for example, the simple presence of a rainbow flag in a window might not be adequate to assure safety to an adolescent if the staff are unkind, unapproachable, or unwelcoming to the adolescent (Wolowic, Heston, Saewyc, Porta, & Eisenberg, 2016). Exploring these complex aspects of socio-cultural environments was enhanced by employing the go-along interview methodology.

Conclusions

This study yielded insights that could be useful to researchers interested in adding the go-along interview method to their research methods toolkit, particularly for understanding the influence of environmental factors on the well-being of vulnerable individuals, families, and communities. Although not without challenges, our recruitment efforts yielded a sample that was diverse with regards to self-described sexual orientation, gender identity, race/ethnicity and age. Participants made use of diverse transportation options and visited a wide variety of sites, including coffee shops, bookstores, and natural park spaces. Participant reactions to the experience were generally positive, and interviews yielded extensive data regarding community supports for LGBTQ youth (Singer, Mehus, Porta, Wolowic, Saewyc & Eisenberg, 2016; Wolowic et al, 2016; Wolowic, Heston, Saewyc, Porta, & Eisenberg, 2016). Collaboration with community partners was fundamental to the success of this work. Established, trusting community-university relationships (including with school-based and community-based organizations serving LGBTQ adolescents) were a critical first step towards accessing participants in a safe, comfortable and mutually beneficial project.

Go-along interview methods have potential to garner elaborate insights from LGBTQ youth; our study focused on youth describing environmental strengths and challenges. Beyond descriptive projects, go-along interview methods could be valuable tools in evaluation research, such as examining the benefits or outcomes of programmatic efforts, policy changes, or intervention strategies. Their use should not be overlooked in research and practice initiatives that aim to advance the health equity and well-being of LGBTQ adolescents.