Ethnography and homelessness research

Papers dedicated solely to research methods are rare within housing studies. This is despite the importance of ensuring methodological rigour within studies that may go on to influence further research, practice and policy. Likewise ethnographic housing research is relatively uncommon even though this approach captures data at a greater level of depth than other methods. Building upon the work of prominent housing researchers who advocate the use of ethnography, this paper highlights the benefits of this approach for housing and homelessness studies, as well as discussing some of its challenges. An overview of ethnography is given to outline its origins; theoretical influences; benefits and challenges in relation to homelessness research. A study that focused on the drug and alcohol behaviours of young, homeless people in Scotland is then introduced. Drawing on this study, a reflexive account of researcher bias and self-disclosure is presented along with consideration of the processes involved in building trust and obtaining informed consent in ethnography. The paper argues the case for this method within housing research and, further, considers the value of ethnography as an approach in the context of growing interest in the lived experiences of people in relation to housing.


Introduction
Housing has a history of being approached from a wide range of disciplines, yet papers that focus on research methods are relatively uncommon in housing journals. Where such papers have been published, they have usually concerned quantitative methods and statistical models (Dodgson & Topham, 1990;Furbey & Goodchild, 1986;Galster & Hedman, 2013;Kramer, Kronbichler, & Van Welie, 2011;Mullins, 2006), although there are some notable exceptions (Jacobs, 2001;Goodchild, O'Flaherty, & Ambrose, 2014). Method is at the core of research since it is the mechanism which transforms ideas, arguments, questions and hypotheses into tangible data. While quantitative methods were historically preferred by policy makers; the value of qualitative research has become increasingly recognised over the past thirty years as it is useful for exploring and understanding policy issues (Ritchie & Spencer, 2002).
One implication of the minimal space given to method in housing research is that the benefits and drawbacks of using a particular method over others can be overlooked. Authors do often address the limitations of their work, typically at the end of a paper; however this also tends to be brief and written in the context of the particular findings. Therefore, these discussions of limitations are limited in themselves given that space restrictions in journals lead authors to highlight only the most pertinent problems while omitting others. It is quite right that these limitations are highlighted, but they may not fully explore the challenges and potential pitfalls associated with each method. Furthermore, the benefits of using a particular method can be understated or assumed. Consequently, the rigour of a study may be called into question if it is unclear as to why specific methodological decisions were made.
In light of these arguments, this paper examines the approach of ethnography and its application to one strand of housing research, namely homelessness. Twenty five years ago, Franklin (1990) made a similar endeavour in relation to homeownership. In his article, the author provided detail of the theories underpinning ethnographic research and he drew upon his own investigations of how people become homeowners to support his argument that there is great value in using this methodology. Over two decades later, building on Franklin's argument, Ronald (2011) made the case for using ethnography in comparative housing research as it is recognised that housing meanings and policies differ across countries and are shaped by specific cultures and practices.
Despite these prominent researchers advocating the use of ethnography in housing research, such studies remain relatively uncommon. Ethnography is not suitable for every project but its capacity to capture processes; to connect what people say with what they do; and to explore everyday lived experiences, makes it a valuable tool. One way in which this approach can inform housing policy is in relation to the increasing focus on the constructed meanings of home in contrast to the economic and material aspects of housing (Easthope, 2004;. Having a home can mean to have ontological security, privacy, constancy, routine, control and a secure base to construct identities (Dupuis & Thorns, 1996;Mallett, 2004;Parsell, 2012). By focusing on people's everyday lived experiences of home, messages can be derived as to the role of policy in facilitating the positive, and minimising the negative, aspects of home. For example, qualitative research has revealed the difficulties in creating a sense of home when living in private rented accommodation which has policy implications for the ways in which tenancies are structured (Easthope, 2014;Smith, Albanese, & Truder, 2014;McKee & Hoolachan, 2015). However, since home is conceived as both a feeling and a process which is enacted (Blunt and Dowling, 2006) conventional qualitative interviewing cannot capture the 'doing' of home; whereas combining interviews with participant-observation in an ethnographic study enables a more holistic picture of home to emerge. Studies of home which have used ethnography have explored how home is gendered (Gurney, 1997) and how domestic tasks are divided (Hochschild, 1989in Blunt & Dowling, 2006. Therefore, for housing research, ethnography provides a means of exploring, in detail, people's actions and narratives and how these connect to broader social structures. The aim of this paper is to draw attention to the benefits and challenges of ethnography with a particular focus on another aspect of housing, namely homelessness. It should be noted that while Franklin (1990) provided substantial detail about the theoretical underpinnings of ethnography and Ronald (2011) highlighted how ethnography has been used within a broad range of housing-related projects, such detail in this paper has been scaled back in favour of exploring more of the practical issues of using this approach. The paper will begin with a short review of ethnography that includes discussion of its origins and links to theory. This is followed by considering some of the challenges and benefits which are illustrated by drawing upon examples of existing homelessness ethnographies. The second half of the paper turns to my doctoral research which involved an ethnographic study of substance use amongst young homeless people in Scotland. While specific findings from this study are not presented in this paper, extracts from the field notes and research diary are used reflexively to illustrate some of the practical and ethical issues that have not been given substantial attention in the literature. Although this study was culturally specific, the messages that can be taken from it in relation to the value of ethnography are relevant to researchers internationally who may be considering using this approach in their own work.

Overview of ethnography
While ethnography usually involves conducting a study that employs two or more qualitative methods (typically participant-observation along with interviews and document analysis) to collect data in a particular context; it extends beyond this simple description as it also captures a philosophical and theoretical stance. Indeed, rather than being termed a research 'method' which is defined as a set of techniques used to collect data, ethnography is more accurately viewed as a methodological approach or strategy which has its own epistemological and ontological positions (Skeggs, 2001).
Developed by early twentieth century anthropologists, ethnography traditionally involves the researcher travelling to foreign places to live among unfamiliar populations and immerse themselves in new cultures. Studying people in their 'natural' environments, rather than in artificial conditions created by the researcher, is a cornerstone of the philosophical framework of ethnography (Hammersley and Atkinson, 2007;Lambert, Glacken, & McCarron, 2011). Every culture has a world view which defines and orders reality and it is the ethnographer's job to describe the value system(s) and universe of the 'natives' (Soloway & Walters, 1977). These 'world views' are embedded within an interpretivist framework as ethnographers are concerned with how people perceive, construct and interact within their social and economic environment, which is why it is necessary to collect data from within these environments (James, 1977). Therefore ethnographers are concerned with what individuals do as well as what they say; they develop ideas and theories in an inductive manner; and they provide thick description in order to create a sense of 'you are there' (James, 1977).
Ethnography is compatible with symbolic interactionist theories that focus on the construction and transformation of shared meanings that come about through relationships. Interactionists such as Mead (1934), Goffman (1959) and Blumer (1969) theorised that society does not exist independently of people and therefore social issues can be studied through the examination of people's relationships with each other and with the physical environment. Ethnography can therefore facilitate the collection of data concerning the social worlds and daily lived experiences of individuals (Prus, 1996). Furthermore, it is well-suited to examining processes and capturing changes and transformations over time (Rock, 2001).
Throughout the twentieth century, ethnography was adapted to incorporate research that focused on geographical locations that were in closer proximity to western researchers. Researchers from the 'Chicago School' are credited as the instigators of this newer trend (Apter et al., 2009) with William F. Whyte's (1943) seminal Street Corner Society epitomising ethnography in a 'local' context. Whyte's study, which focused on a group of Italian immigrants living near to Harvard University, provided detailed description of the group's values, rituals, behaviours and histories thus demonstrating that 'local' ethnography could produce findings just as informative and sociologically significant as traditional ethnography. The argument for conducting 'local' ethnography diverges from Ronald's (2011) position of highlighting the need for ethnographic housing research in a comparative context. However this is not to say that these different types of research are at odds with each other since they can both address housing issues from different cultural perspectives in a complementary manner.
The capacity for ethnography to derive detailed insights into people's everyday lives has great implications for housing policy. In a commentary between prominent Chicago School sociologists, Herbert Gans argued that there are 'weak links' in society and the job of ethnographers is to help policy makers to find these weak links so they can be addressed (Becker, Gans, Newman, & Vaughan, 2004: 266). The use of ethnography, he believed, would enable the voices of the marginalised to be heard: And I think ethnography is particularly useful here because we come closest to the people who are being studied…Ethnography has always studied the underdog…we can communicate for the underdogs, for the victims to À and in opposition to À the perpetrators as I think of them. (p. 265) Currently in the UK, as in many other countries, austerity measures contextualise the decisions of policy makers and practitioners. Within housing, low-income households are being significantly disadvantaged by cuts to social security benefits and services. This is compounded for those living in private rented accommodation and for those who are employed on a 'zero-hours contract' basis (McKee & Hoolachan, 2015). As a result, researchers have turned their attention to the impact of these economic and housing policies on the daily lives of individuals (Smith et al., 2014). While much of this research has utilised qualitative interviews, ethnography would offer an even greater depth of understanding. Not only would the approach capture how people feel about their situations, it would also allow the researcher to witness first-hand, how these contexts are enacted in daily practices. It is argued that ethnography enables researchers to understand how the complexities of an individual's situation are enacted, felt, negotiated and controlled. Thus, policies can be developed that are more sensitive to the minutiae of everyday life.

Benefits and challenges of ethnography: examples from homelessness literature
The majority of housing-related ethnographic studies concern the topic of homelessness. This is perhaps because the approach is particularly well suited for working with vulnerable populations since it incorporates time for trusting relationships to develop (Cloke, May, & Johnsen, 2010) and it can facilitate access to hard-toreach groups (Hammersley & Atkinson, 2007). The sections that follow will explore these benefits further, as well as challenges, by drawing on existing ethnographic homelessness literature.

Lengthy fieldwork period
One of the significant features of ethnographic research that sets it apart from other methods is the length of time spent in the field collecting data. According to LeCompte & Goetz (1982), fieldwork typically lasts anywhere from six months up to three years, although others have reported even longer data collection phases (Bridgman, 2002;Gibson, 2011). Hammersley and Atkinson (2007) have documented some of the stresses and strains involved in ethnography and these can be exacerbated by such lengthy fieldwork periods. Since homeless populations are a largely mobile group who often do not stay in one location for long periods of time, keeping track of research participants can be particularly challenging. This was noted by Parsell (2011) whose participant-observation included approximately one hundred homeless individuals although he clarified that he could not provide an exact figure as many of his observations were made from a distance. This raises the challenge of trying to capture a comprehensive documentation of the participants' lives as well as obtaining informed consent from everyone observed. Some have addressed these challenges by focusing on a smaller number of individuals (e.g. Ilan, 2013) or on one particular location (e.g. Perry, 2013).
An additional strain that results from lengthy fieldwork periods is exhaustion. Immersing oneself in a different culture for an extensive amount of time can lead to fatigue as the researcher always needs to be alert to the research aims and adopt a marginal position to ensure they never feel fully 'at home' (Hammersley and Atkinson, 2007, p. 90). The ethnographer approaches the research from the position of an outsider (referred to as the 'etic' position) with the aim of acquiring an insider's (emic) perspective (Harris, 1990) and then latterly linking this emic perspective to etic theories and literature, all of which is mentally and physically demanding. This is compounded when members of the participant group engage in activities that some ethnographers may feel uncomfortable with. For example, in his study of homeless youth, Barker (2013) described an encounter with a young male attempting to break into a car to steal a cigarette. Barker soon found out that this young man was on probation and quickly offered to buy him cigarettes as he was concerned about the consequences for this individual. If Barker was to fully take on a role of an insider then he could have assisted with the break in, but rather chose to offer an alternative solution. Therefore, the researcher's position lay between the emic and etic which he had to sustain for the duration of his study.
Despite these challenges, the benefits of lengthy fieldwork are numerous. As will be discussed further, it enables the researcher to develop rapport and strong relationships with his/her participants. The hope is that in doing so, participants will feel more inclined to allow the researcher to learn about aspects of their social worlds that they may not have revealed to a relative stranger. This was evident in Hall's (2003) ethnographic study of a homeless hostel in London. After renting a bedsit close to the hostel and spending a great deal of time in the company of his young participant group, Hall found that they began discussing all sorts of personal information with him and one even stayed in the researcher's bedsit for a few days to avoid his friends with whom he had been arguing. Thus, a strong level of trust can develop between the researcher and participants and it has been noted that ethnography is useful when conducting research with vulnerable populations who are prone to distrust (Cloke et al., 2010).
In addition, the lengthier the fieldwork, the more the researcher is able to observe changes that occur over time thus adding an additional, longitudinal dimension to the analysis (Herbert, 2000). This was noted in a Canadian study by Bridgman (2002) who sought to learn about a 'safe haven' model which was an intensive service for homeless, mentally ill women that provided long-term accommodation coupled with high-level support. Over a four and a half year period, Bridgman collected a wealth of data from the staff and service users to evaluate the model. The prolonged nature of the fieldwork allowed her to capture the small, slow improvements in the wellbeing of the homeless women which would otherwise have gone undocumented.

Generalisability, validity and reliability
Compromising generalisability, validity and reliability is a long-standing criticism of ethnography and qualitative research more broadly. These criteria are rooted in positivist research that seeks to uncover facts and truths in a scientific manner and since qualitative research rests within a different epistemological and ontological philosophy, it has been argued that these measures cannot be applied (Lincoln & Guba, 1985). However, a growing effort has emerged to adapt these standards and some argue that failure to do this can damage the credibility of a qualitative study (LeCompte & Goetz, 1982). Transparency is key to meeting these standards, according to LeCompte and Goetz (1982), although they point out that the objectivist measure of reliability can never fully be achieved because qualitative research is concerned with the non-static experiences of human life which can be impossible to replicate.
Reflexivity is one tool used by ethnographers to enhance the rigour and validity of a study. In contrast to objectivist research which seeks to neutralise the influence of the researcher; reflexivity refers to the recognition that ethnographers are integral to the social worlds that they study. The researchers' social characteristics such as gender, age and ethnicity as well as issues of power differentials, emotions and personal experiences all have an influence on the research process (Lumsden, 2009;Mauthner & Doucet, 2003;Punch, 2012). Reflexivity involves recognising that the researcher becomes a part of the participants' lives and this interaction influences the data that are collected. For example, in an American ethnographic study about 'street kids' in New York, Gibson (2011, p. 17) reflected: As a relatively affluent, white female researcher, I felt I had very little in common with homeless teenagers, and so I would not be able to 'connect' with them. By channeling [sic] the majority of my fieldwork through street outreach, I increased my access to street kids but altered my position as a researcher.
By the end of her fieldwork, Gibson described herself as an outreach worker and since many of the street kids had come to know her in this position it likely affected the ways in which they communicated with her. While quantitative proponents would view this researcher influence as problematic, ethnographers regard it as inevitable. Reflecting personal biases enhances the transparency of the research process which allows readers to place the findings in the context within which they were generated. Therefore, reflexivity is argued to contribute to the validity of an ethnographic study (LeCompte & Goetz, 1982).

An ethnographic study of the substance use of homeless youth in Scotland
This paper now turns to doctoral research which used ethnography to explore the meanings and contexts of substance use amongst homeless youth in Scotland.
Whilst ethnography can and should be used beyond studies of marginalised groups (Ronald, 2011), the insights from this study highlight why it is such a useful means of engaging with vulnerable young people. When developing the proposal for this study, it became apparent that simply asking young homeless people to talk about their alcohol and drug use was problematic. Not only were they unlikely to fully disclose such information, there was also an ethical concern about collecting such personal details and then walking away. Ethnography provided the means of developing relationships with participants as well as offering a much more in-depth understanding about how their substance use intersected with their daily lives in temporary accommodation.
Specific findings from the study are not presented here. Instead, what follows is a reflexive account of my experiences of conducting ethnography with a group of young homeless people. Just as methods papers are rare within housing studies, so too are reflexive accounts. Yet reflecting on the research process can add to the validity of a study and can reveal further insights as to the difficulties and value of engaging in such work. Such a reflexive account adds a different dimension to Franklin (1990) and Ronald's (2011) contributions to the literature of ethnography within housing studies.

The study
The site of this study was a temporary accommodation hostel in central Scotland which housed young people (referred to as 'residents') aged 16À21 years old. This hostel was given the pseudonym of Kelldale and, although its policy indicated that a young person would be accommodated for a maximum of 10 weeks, in reality, the majority of residents were there for much longer due to a shortage of suitable social housing.
Participant-observation was the primary method of data collection and was conducted over a seven month period between May and November 2013. The resulting data were supplemented with four interviews involving six young people and one focus group consisting of another six. Each of these methods were executed in an overt manner as it was felt that not only would covert research have been unnecessary and difficult to achieve; it would also have been unethical to write about personal and sensitive information without the participants' knowledge. Kelldale accommodated 14 young people at any one time. During the seven months of fieldwork, 32 young people passed through the hostel, 22 of whom participated in the research. Sixteen young men and 6 young women participated and of these, 19 were white/Scottish, 1 was Asian/Scottish and 2 were asylum seekers. In addition, 30 members of staff provided consent for their involvement in the study. No data involving the staff are presented here and the residents' names have been anonymised through the use of pseudonyms.
Kelldale was a purpose-built hostel which implemented more 'house rules' compared to other types of homeless accommodation. Some of these rules included prohibiting the use of alcohol and drugs in the building, enforcing a midnight curfew, and not allowing visitors to enter the premises. The residents could come and go as they pleased during the day but when they were inside Kelldale, their movements were monitored by Closed Circuit Television. The service aimed to provide a substantial level of practical and emotional support to its residents, which was the reasoning behind the intensive involvement in their behaviours and whereabouts.
In general, demand for a place in Kelldale outweighed supply as there were always a number of young people waiting to move in when a resident moved out. Therefore, the managers had to make decisions as to which individual on the waiting list should take priority over others. This involved a scoring system based on seven broad categories of need, four of which included: offending; substance use; mental health; and physical health. Individuals experiencing vulnerability in relation to these criteria were given priority. The remaining three criteria were age, current housing status and support network. Younger individuals were prioritised for a place as were those who were roofless or sofa surfing and who had minimal social support. The staff pointed out, however, that these criteria were guidelines only and it was just as important for them to consider the current mix of residents when giving priority. Overall, though, in addition to being homeless, the Kelldale residents consisted of a group of young people who were highly vulnerable in at least one aspect of their lives.
Ethical issues will be discussed further in the proceeding sections however a few points of clarity regarding my own safety are warranted. Approval for the study was granted by the University of Stirling ethics committee. The aim of the research was not to directly witness the consumption of drugs and alcohol but to understand how the residents conceptualised their substance use and how it interacted with their homelessness. Therefore, I was never knowingly present when substances were being consumed and the majority of my observations involved chatting or walking around with the young people. I was however frequently in the company of intoxicated residents and during these occasions it was particularly important to be sensitive to my own safety. This involved ensuring I was never alone with an intoxicated resident and I always placed myself close to the door to ensure I could leave without obstruction. Arguably my position as a young female heightened my vulnerability but it likewise assisted in minimising power differentials. Ethnography is often a riskier approach compared to other methods and substantial consideration of researcher safety is of paramount importance.

Reflections on my position in Kelldale
In the spirit of exercising reflexivity in ethnography, Lumsden (2009) pointed to the need to reflect on the negative as well as the positive aspects of fieldwork. In particular, she acknowledged that often negative experiences and feelings are omitted from research papers for fear that the ethnographer is viewed as incompetent or ungrateful to his/her participants for permitting access to their social worlds. However, reflexivity and transparency are required in all aspects of the research process if the social construction of knowledge is to be explored and validity achieved (Punch, 2012). Furthermore, novice ethnographers may not be fully prepared for the potential reality of their fieldwork experience if authors only reflect on the positive aspects of the approach. Based on these arguments, I wish to reflect on two occasions in Kelldale that were uncomfortable and that revealed different aspects of how those involved were perceived by an outsider. These have been selected on the basis that they highlight researcher bias and the benefits of reflecting on this throughout the research process.
The first experience occurred prior to properly beginning the fieldwork. During a meeting with the manager I saw three residents through the window of the room I was sitting in. This was the first time I had seen any of the young people and my reaction was captured in the following research diary extract: … to be perfectly honest, they looked like the type of people I would try to avoid if I saw them standing on the street. One guy was wearing a tracksuit and baseball cap and was covered in tattoos…I feel guilty for thinking this way…these are the young people I want to talk to and understand, yet I am instantly judging them based on how they look.
Given that I approached my research with a view of exploring and giving voice to a marginalised group, I was shocked and ashamed of myself for judging the residents in this way. However such reflection was beneficial as it made me more aware of personal biases which could be addressed prior to engaging with these young people. Furthermore, as time went on I was able to compare my initial judgement with my knowledge of the residents' personalities and marginality which enabled a deeper understanding of how their constructed identities and appearances often clashed with wider middle-class structures and norms. For example, one young resident called Danielle explained how she had been unable to attend job interviews in the past due to not being able to afford appropriate interview clothing. She owned relatively little clothing and therefore, even if she wanted to, she could not easily change her appearance to appease those who may have made similar judgements as I had initially done.
The second uncomfortable experience involved a young man called Garry, who had several mental health problems and was a poly-drug user. He was almost always intoxicated. On every occasion I interacted with Garry he would talk in graphic detail about sex or violence and he enjoyed asking me personal questions about my own sex life. Despite my refusals to respond to these questions, Garry persisted to the point where it was difficult to encourage him to talk about anything else. I persevered with Garry and believed that his behaviours were attempts at 'testing' me to explore how far he could push the boundaries and that over time he would stop. However, after three-weeks of these conversations, I found myself trying to avoid Garry as much as possible as he increasingly made me feel uncomfortable. This therefore, altered the data collected and introduced an additional aspect of bias since I began giving other residents more attention at Garry's expense.
As well as reflecting on how I viewed the residents, exploring how they viewed me is helpful for illustrating some of the difficulties involved in negotiating the emic-etic position of an ethnographic researcher. Everyone connected to Kelldale could be divided into two groups À residents and staff À and consequently there was a 'them and us' mentality amongst the young people. Since it was impractical for me to live in the hostel, I operated a shift pattern very much like the staff and my relative affluence and stable family background set me apart from the residents. Therefore, it was clear from the beginning that it would be challenging to be accepted as an insider. This was exacerbated by the decision to use the staff office as a base to position myself in the hostel since it was the only site that made me visible to the young people without intruding on their private spaces. These elements combined meant that it was difficult to portray a non-staff image. Indeed, despite informing them daily that I was not a staff member, I was often asked by a resident to perform staff duties such as giving them access to certain parts of the building and checking their telephone messages.
I attempted to distance myself from this staff identity through various means. For example, as will be discussed further in the next section, the residents began to learn that I would not enforce the rules and occasionally would be a witness to them breaking the rules which the staff strived to implement. I copied the way the residents spoke such as using particular slang words, swearing and poking fun at the staff. More significantly, I disclosed a degree of personal information to the residents which is something that the staff seldom did. Self-disclosure is an issue that is often recognised by ethnographers but rarely discussed in great detail. Hammersley and Atkinson (2007) note that it is unfair to expect participants to be frank and open about their social worlds without the researcher reciprocating with details of his/her own life. However, what, and how much, one should disclose is tricky to negotiate. On the one hand, self-disclosure can strengthen the relationships between both parties. For example, one resident named Ryan frequently wished to discuss his relationship with his girlfriend. Ryan learned that I was in a long-term relationship and we found ourselves discussing what 'good' and 'bad' relationships consist of which inevitably involved disclosing further personal details. However, these conversations led Ryan to increasingly invite me into his bedsit when he was hanging out with other residents which was indicative of a trusting relationship. On the other hand, self-disclosure can distort the power differentials so that the participants may feel as though they are inferior (Tang, 2002). This was evident when Garry discovered that I used to attend a certain school that he regarded as 'posh.' However, in my experience, this worked both ways. By disclosing certain information, the residents were able to work out areas in which they had greater expertise. For example, Nathan had experience as a mechanic and when he discovered what type of car I owned, he took great pleasure in pointing out why it was a 'pile of shit'! Developing trust with research participants Trust is tied to ethical issues of working with vulnerable groups. In a paper investigating the ethics of conducting research with homeless young people, Ensign (2003, p. 46) states 'as most homeless youths have had a series of harmful past and present experiences with various adults and institutions in their young lives, they are highly distrustful of adults and institutions'. Such distrust could understandably result in homeless individuals being hesitant to engage with researchers, particularly if they feel they would have nothing to gain by discussing their lives with a stranger (Cloke et al., 2010). Furthermore, the connection that homelessness has with substance use and other antisocial behaviours (Fitzpatrick, Johnsen, & White, 2011) suggests that individuals involved in such activities may feel even less inclined to discuss these aspects of their lives for fear of negative repercussions.
In my own research, the residents of Kelldale vocalised their distrust towards me as indicated in the following field note extracts: Garry told Stephanie that I wanted to know all about the drugs she takes and Stephanie was silent. Garry said 'aye, she's here to grass 1 us all up.' I'm no' gonna tell you all my secrets when I've just met you. (Ryan) Overcoming this distrust required consistency on my part in terms of spending concentrated periods of time at the hostel over several months as well as consistency in how I approached and behaved towards the young residents. As I got to know them slowly over time they gradually opened up and divulged personal information and details of their drug using behaviours which they had previously kept hidden. For example, Tom later explained that on an earlier occasion when I had spent time with some of the residents in one of their private bedsits, Tom, along with a few others, had taken it in turns to go into the en-suite bathroom to smoke cannabis. I had been sitting in the bedsit and was completely unaware of this, much to the amusement of those involved.
Closely related to this trust was the need to demonstrate that I would adhere to the confidentiality agreement I had made with the young residents. This involved not reporting their behaviours to the hostel staff, even when this involved breaking the rules or the law, unless I was concerned for an individual's safety. Opportunities to prove myself in this respect were spontaneous and therefore largely dependent on being present at the hostel as much as possible.
Tom and Ryan started talking in low voices and when I asked what they were doing Ryan said they were stealing yoghurts from the fridge: 'don't tell anyone.' Danielle said she didn't want to do a bucket 2 because I was there. I explained that I wouldn't tell the staff and if they were actually going to do a bucket then I would leave…Danielle seemed worried that the staff would figure out what they were up to by the fact I was going back to the office. I said they don't tend to ask me why I've left someone's room but if they did I would just tell them it's because I'm going home soon. Danielle nodded with approval.
Investing time and maintaining consistency in this respect not only increased the trust that the residents had in me, it also meant that when new residents moved into the hostel, the existing residents endorsed my presence.
Ryan looked at me, paused in thought momentarily, grinned and said 'you coming up [to his bedsit]?' I said yes and as we were walking up Matt expressed concern that I was with them. Ryan reminded Matt about confidentiality 'nah she cannae say anything to the staff remember.' Building trust and rapport with research participants so that they feel comfortable enough to invite the researcher further into their social worlds requires a great deal of time and effort. However the benefit of this investment is that it facilitates greater depth of data. This type of relationship building is often not possible when conducting non-ethnographic research as researchers typically do not spend a great deal of time getting to know their participants before commencing an interview or distributing a survey. Therefore it is likely that the data collected by these latter methods will have a narrower focus and understanding of the fuller context may be limited.

Informed consent
Within research ethics generally, it is a challenge to ensure that individuals have a sufficient understanding of what it is they are consenting to, since signing a form is not necessarily an indication that the participant is fully informed (Murphy & Dingwall, 2001). Ethnography poses additional challenges to the informed consent process. Notably, it is often not possible to obtain consent from every person that enters the research field, particularly those whose presence is brief and tangential (Hammersley & Atkinson, 2007). In Kelldale, this issue mostly concerned professionals who visited the hostel once or twice, such as social workers or police officers. However, it was also difficult to obtain formal, written consent from some of the residents. Unlike other methods, obtaining informed consent in ethnography is viewed as a process that continues for the duration of the fieldwork (Ensign, 2003).
Given the lack of trust by the residents, it was felt that informed consent would take some time to achieve and therefore I did not ask an individual to sign a consent form when I was initially introduced to them. Instead, I interacted with residents for a week or two to enable them to gain a gradual knowledge of the research and myself before asking them to formally consent by signing a form. In the majority of cases this was deemed successful, but difficulties arose when a young person moved out of the hostel before I had asked them for formal consent. Each of these cases was assessed individually based upon the amount of time I had spent with the resident and the level of understanding about the research they had expressed verbally. In most of these cases, the data collected about these individuals were discarded. However, if a field note extract described a general conversation or scenario involving one of these residents that did not reveal any personal information, then it was retained. For example, Stacy moved out of the hostel two weeks after I began my fieldwork. I had not had the opportunity to obtain a signed consent form from her but we had verbally discussed the research. Care was taken to omit data including Stacy where possible, but her presence during group activities was retained as demonstrated in the following field note extract: I re-joined the group for a game of bingo organised by the staff. This is where I got to meet Cara and Stacy although Stacy went back to her room before the game started as she wasn't feeling well.
Despite these challenges of informed consent, the lengthy duration of ethnographic fieldwork also creates opportunities. For example, having an extensive amount of time to inform participants about the research means there is greater scope to assess their level of understanding throughout the fieldwork. In the context of Kelldale, it took a substantial period of time for some of the residents to understand why I was at the hostel and that I was not a member of staff despite initially agreeing to consent and confirming that this consent was informed: Liam asked me to check his mail for him and I explained once again that I'm not a member of staff. Liam remembered that I'm at university and that has something to do with why I'm there, he said he would catch up with me later.
This conversation with Liam revealed that he had not fully understood the nature of the research when I had gotten him to sign a consent form and consequently it was necessary to repeatedly explain it to him over a period of time.
In addition to ensuring consent is informed, those who are initially distrustful and uninterested in participating in a project may change their minds as they become more familiar with the researcher and the study being conducted. Therefore ethnographers are in a position to obtain consent at a later date from those who would initially have refused to participate. In Kelldale, only one resident at first refused to give her consent on the basis that she did not see why she should divulge personal information to a stranger. Respecting this decision it had been my intention not to include any data involving her in any written publications. However over the next few visits to the hostel I came to spend a great deal of time with her and when she began to question me about my research I revisited the possibility of consent: Amanda said it was fine to include her in the 'book': 'aye I was only bammin' 3 ye up'

Discussion and conclusion
The aims of this paper have been twofold: to highlight the usefulness of engaging in discussion about method in housing and homelessness research; and to argue that ethnography is a valuable methodology that should be considered when planning a housing-related research project. As with all research methods, ethnography has its own unique benefits, challenges and drawbacks. Taking a reflexive approach can contribute to highlighting these nuanced experiences. The lengthy period of time spent collecting data can place great stress and fatigue on the researcher and increases the likelihood of difficulties occurring such as navigating the lines between emic-etic positioning. Furthermore, it is not possible to achieve objectivity in this type of research which makes it susceptible to criticisms concerning generalisability, validity and reliability. Bias can be introduced into the fieldwork through the social characteristics and experiences of the researcher; self-disclosure; and avoiding individuals who are challenging. Finally, the ethnographer's lack of control over who enters the field, and for how long, can make obtaining informed consent from everyone rather difficult.
However, ethnographers have developed techniques for responding to these challenges such as incorporating reflexivity and transparency into the research process. Although it is demanding, the extensive fieldwork creates substantial opportunities for collecting in-depth data as the dimension of 'change over time' can be incorporated into the analysis and there is scope to be flexible and adapt to changing circumstances where appropriate. Most significantly, it has been argued that ethnography is particularly well-suited to research with vulnerable groups such as those who are homeless. This is due to its facilitation of building trusting relationships which can potentially lead to participants becoming increasingly revealing to the researcher, resulting in the collection of data at a deeper level. In addition, the lengthy period of fieldwork can increase the likelihood of ensuring the participants are fully informed about the research and may lead to the inclusion of individuals who would otherwise not have given consent.
Social science research is used to influence further research, practice and policy and therefore it is important to understand and critique the methods employed to generate research findings. In relation to the area of homelessness, Gachet (2010) contends that it is necessary to develop a common approach towards an issue: a quantified language that makes use of measurable outcomes for tackling issues of practice and policy. Despite its subjective and interpretivist positioning, ethnography can complement this common approach by giving a voice to individuals who are at the core of these measurable outcomes. This could be in terms of highlighting a particular social problem that requires a policy response or it could involve investigating how a particular policy impacts on individuals.
Homelessness represents the largest section of housing studies in which ethnographic research has been conducted. It is important to note that while some of the benefits and challenges of ethnography apply across studies, there are potential differences between 'homeless ethnography' and 'housing ethnography'. Homeless populations typically (although certainly not always) reside within (semi-)public places. Most ethnographic studies are carried out either with rough sleepers or those engaged with services. This provides a level of protection for the researcher since visibility means that there are people on hand to provide help should it be required. On the other hand, 'housing ethnography' may involve the researcher spending time inside people's private homes away from the public gaze. Therefore, as Blunt and Dowling (2006) note, it is crucial that an ethnographer only enters a home if he/she feels safe and that they inform someone of their whereabouts at all times. In addition, the process of accessing homeless and housed populations is likely to differ. Whilst accessing homeless people usually requires additional gatekeeper and ethical approval; accessing housed individuals may prove problematic in that a researcher could be seen as an intruder in someone's home which may act as a deterrent for participating. One way of overcoming these challenges is to initiate an ethnographic study in a public place which may lead to invitations into the home and a level of trust to develop; although the risks of entering a home will still persist.
Notwithstanding these differences, other aspects of housing research could gain considerably from implementing an ethnographic approach. While certainly not limited to vulnerable groups, the arguments presented in this paper suggest that ethnography may be a valuable methodology for researching the housing experiences of other marginalised populations such as disabled, migrant, homo-/bi-sexual or transgender individuals. Alternatively researchers interested in the concept of home could utilise ethnography to investigate how meanings of home are constructed and experienced with regards to different groups and different living situations. Otherwise, this approach could be used to investigate the impact of policies such as the 'Bedroom Tax' or welfare reforms introduced in the UK in recent years or, following in the footsteps of Ronald (2011), to explore housing issues on a comparative basis.
Regardless of how it is used, ethnography is a methodology which promises the collection of detailed in-depth data that have the potential to highlight new issues that may require policy responses. Despite being used as a research approach for decades in anthropology and sociology, it is still relatively under-used in housing studies and like Franklin (1990) and Ronald (2011), this paper echoes the call to consider the benefits of conducting ethnographic research and collecting data which would sit alongside existing knowledge to provide a richer context. This is particularly timely as housing researchers are increasingly becoming interested in the dayto-day, lived experiences of people: an area in which ethnography excels.