Exploring the experiences of non- statutory homeless individuals who engaged in an adventure therapy intervention

Johnstone et al. (2016, p.421) identify homelessness as “a significant and worldwide problem”. Whilst homelessness is often associated with “sleeping rough” (Ministry of Housing, Communities, & Local Government, 2018), UK legislation (Housing Act, 1996) identifies individuals as homeless if they do not have the legal right to occupy habitable accommodation (nonstatutory homelessness). The current paper refers to “individuals experiencing homelessness” (IEH) in relation to “nonstatutory” homelessness and individuals who do not meet “priority need” legislative criteria and therefore are not legally entitled to housing. Received: 27 August 2020 | Revised: 15 February 2021 | Accepted: 16 February 2021 DOI: 10.1002/capr.12399

IEH have higher rates and complexity of mental and physical health issues than non-homeless individuals, leading to increased mortality (Elwell-Sutton et al., 2016;Fazel et al., 2014), and frequent contact with mental health services (Russell et al., 2020). They also report experiences of abuse, stigmatisation and social exclusion (Boydell et al., 2000;Hwang & Burns, 2014), which have a negative impact at the interpersonal and intrapersonal level, leading to exacerbation of mental health issues and social isolation (Williams & Stickley, 2011). Poor access to healthcare services intensifies increased morbidity and mortality rates (Elwell-Sutton et al., 2016;Fazel et al., 2014;Woollcott, 2008). Direct discrimination (e.g. loss of autonomy, feeling patronised) and indirect discrimination (e.g. registration issues, use of the medical model to "treat" homelessness) from healthcare services increase health inequalities and experiences of social exclusion (Lyon-Callo, 2000;Rae & Rees, 2015;Williams & Stickley, 2011;Woollcott, 2008). Furthermore, it has been identified that generic services struggle with providing sufficient flexibility for IEH (Spence, 2009;Williams & Stickley, 2011;Woollcott, 2008).
Although housing support is suggested to be the most important service provision for IEH, Hwang and Burns (2014) stipulate that increased mental health service provision is needed, with prioritisation of engagement. Recommendations include development of positive relationships between staff and IEH; flexibility and adaptation to IEH needs; opportunities for informal, non-therapeutic interactions; and peer support (Barker & Maguire, 2017;Barker et al., 2018;Bentley, 1997;Flanagan & Hancock, 2010;Hwang & Burns, 2014;Johnstone et al., 2016;Williams & Stickley, 2011;Woollcott, 2008). Community reintegration is also recommended due to the negative impact of sustained social exclusion (Thomas et al., 2011). Nevertheless, there appears to be limited research in this field.

| Adventure therapy
Adventure therapy (AT) is a mental healthcare intervention, using adventure activities within a therapeutic framework, which promotes interpersonal relationships and reflection (Alvarez & Stauffer, 2001;Bowen & Neill, 2013;Hanna, 2012). Activities include perceived risk and/or high level of challenge and physical activity, and can occur outdoors, indoors or in an unfamiliar environment (Bowen & Neill, 2013;Crisp, 1998;Itin, 2001;Russell, 2001). Intentional use of activities for therapeutic purposes is based on client goals, which are usually developed with a mental health professional (Bowen & Neill, 2013;Crisp, 1998;Russell, 2001). Trusting relationships with AT practitioners are essential in facilitating physical and psychological safety (Gass et al., 2012).
Adoption of a challenge-by-choice model facilitates autonomy development by enabling participants to choose their level of engagement (Schoel & Maizell, 2002). When conducted in groups, processes such as peer support and modelling, shared experience, and interpersonal and intrapersonal adaptation are facilitated.
Nature connection can facilitate emotional well-being, stressreduction, and improvements in mental and physical health issues via provision of a healing environment (Summers & Vivian, 2018;Taylor et al., 2010). Therapeutic use of nature is beneficial to individuals with negative early relational experiences (Jordan, 2016), such as IEH (Fitzpatrick et al., 2013;McNaughton, 2008), indicating suitability of outdoor AT for the needs of IEH.
Research into AT interventions report improvements in mental health symptoms and well-being, intrapersonal and/or interpersonal relating, physical activity, and social support (Bidell, 2010;Bowen & Neill, 2013;Carless et al., 2013;Gillis & Speelman, 2008;Kyriakopoulos, 2011;Wilson et al., 2010Wilson et al., , 2011. Furthermore, socially excluded groups report improvements in psychosocial wellbeing and better social and psychological adjustment (Bidell, 2010;Carless et al., 2013;Marchand et al., 2018;Wilson et al., 2011). In a recent review of ecotherapy interventions, Summers and Vivian (2018) have found that interaction with nature and nature-based interventions positively affects well-being and health throughout life. Similarly, Picton et al.'s (2020) systematic review of 18 qualitative studies identified that outdoor-based therapeutic programmes make a positive contribution to mental health with increased levels of physical activity, greater self-esteem and enhanced sense of identity being some of the perceived positive changes.

| Current study
Despite the emerging number of studies exploring the relationship between mental health and AT, there do not appear to be F I G U R E 1 Adaptation of Kolb's (1984) experiential learning theory to a rock-climbing activity within an adventure therapy intervention. This is a non-exhaustive example and does not include all learning opportunities that could occur. Experiential learning is a cyclical process, beginning at "Concrete Experience" any qualitative phenomenological studies, which have specifically explored the experiences of IEH participating in AT interventions.
Studies so far seem to have focused on mental health problems for community or psychiatric populations. Furthermore, of the phenomenological studies conducted so far, none has used a repertory grid as a method of data collection. Hence, the aim of the present research was to address this gap in the current UK homelessness literature and explore the experiences of IEH, who were living in supported hostel accommodation and had voluntarily engaged in an outdoor AT intervention programme. A repertory gird was used to increase verbal expression.

| Intervention
The AT intervention was developed and facilitated by an AT provider independent of the research team. The main aim of the programme was to improve mental well-being and level of physical activity. The intervention was promoted with leaflets and visits to the hostels by the AT practitioners. Interested service users were allocated to two separate groups. Both groups had the same 9week structure; 6 days over 6 consecutive weeks (1 day per week) at the AT provider's private, 4-acre Meadow; an overnight camp at the Meadow; and then a 3-day trip to Wales. Activities at the Meadow included archery, canoeing, arts and crafts, and team games. Service users were responsible for set-up and de-camping during all overnight trips. The two Wales trips had slightly different activities. One involved a 6.5-mile hike and underground caving session, and the other involved a mountain summit and gorge walking. A final session at the Meadow involved a traditional "sports day" (e.g. egg and spoon races) and closing ceremony, where participants were awarded certificates and personalised commendations, and speeches were made by AT practitioners and service users. Voluntary sessions were scheduled for 2 months after the intervention end, where participants would assist with maintenance of the Meadow and AT equipment. Group discussions and reflections were encouraged throughout the intervention and were facilitated by the practitioners.

| Methodology
The methodology is informed by the theoretical orientation of interpretivism. The interpretative phenomenological approach focuses on the lived experience of the participants as they engaged in the AT programme whilst living in supported accommodation.
Interpretative phenomenology focuses on the way individuals make sense of events and accepts that the way people talk about their experiences is unique in that it is influenced by their personal cognitive constructs and emotions (Smith & Osborn, 2015).

| Data collection
Following ethical approval, the primary researcher conducted individual, semi-structured interviews, two weeks after the end of the AT intervention. These occurred in communal spaces in participants' hostels of residence, which were booked for the interview duration (average 50 min; range 31-64). For all participants, exploration and clarification of their AT experiences was facilitated with prompts such as "How do you mean…?"; "In what way…?"; or "Can you tell me more about…". The final questions asked whether there was anything additional they wanted to express in relation to/about their overall experience, thereby facilitating the opportunity to speak about all aspects of their experiences.
Due to potential reduced verbal expression of marginalised populations (Odusanya, 2018;Samuel, 2004), including IEH (Gelberg & Siecke, 1997;Padgett et al., 2013), a repertory grid was used as an aid to the semi-structured interview. Repertory grids are used to explore participants' internal world by accessing lower levels of cognitive awareness (Burr et al., 2014;Fransella et al., 2003;Jankowicz, 2004) and in this case aimed to increase verbal expression and minimise researcher bias.
Participants were introduced to the repertory grid, followed by collaborative discussion to identify up to 10 elements that described their AT experience; these were concrete rather than abstract. Participantelicited elements were chosen to facilitate greater insight into their internal worlds (Jankowicz, 2004). Once a maximum of 10 elements were identified, in-depth exploration of participants' AT experiences began. For three participants, this included partial continuation with the repertory grid; for three, the discussion of identifying elements encouraged them to speak expressively about their experiences; and for the final participant, a full list of elements was not required as they had a high level of verbal expression from the interview start. Continuation with the repertory grid involved comparison of random, triadic groups of elements. The researcher wrote down each triad on a piece of paper and asked how two of them were alike, but different from a third, in relation to the AT experiences. Exploration of this was facilitated to gain insight into participants' internal worlds.

| Participants
Seven participants were recruited by purposive, homogenous sampling; all were classed as having a "non-statutory" homeless status and volunteered to participate in the AT intervention between May and August 2017. This included five male and two female participants (M age = 44.1 years, range 22 to 53 years; one participant chose not to disclose their age, but indicated they were within this range). All participants were living in supported hostel accommodation (n = 3 hostels; M period in hostel = 12 months, range 3.5 months to 24 months). Participants self-reported a range of mental health disorders including depression, anxiety, and substance abuse.
Participant characteristics are reported in Table 1.

| Analysis
Verbatim transcripts of the audio-recorded interviews were typed and analysed using Smith et al.'s principles (2009) of interpretative phenomenological analysis. This involved case-by-case reading and analysis of the transcripts. Aspects that appeared to be important to the participants were noted in the left margin. With further reading, emergent themes were noted in the right margin. Themes were then chronologically ordered, with supporting verbatim quotes and location in the transcript. From this, clusters of superordinate themes were developed into a table with supporting quotes. This process was repeated for all transcripts, and a table of master themes relating to all participants was developed.
Engagement with the hermeneutic circle involved an iterative and inductive process, which maintained IPA's idiographic focus, whilst interpreting this within a wider context, as appropriate (Smith et al., 2009;Smith & Osborn, 2015).

| FINDING S
Three major themes were identified from the data (illustrated in Table 2). These are outlined with supporting verbatim quotes; pseudonyms are provided after each quote. Square parentheses indicate explanatory information that was not verbalised, and missing text is indicated with three dots.

| Respite
Participants described this respite as having a positive impact on their mental health and well-being as it enabled them to temporarily forget about the negative experiences of homelessness. Participants felt that the AT intervention provided respite from anxieties relating to hostel-life and socioeconomic stressors; and feelings of social isolation, emptiness, monotony and social exclusion. This respite was reinforced through the development of a sense of community and belonging, mediated by the positive, egalitarian relationships with practitioners and peers, which contrasted with experiences of stigmatisation.
Most people in this house have got drug problems, alcohol, and you can't escape from it, being in these places. But that's how-we got to, we got to escape from it, going to the Meadow. And get away from the drugs, get away from everything. That was what I enjoyed, was getting away from what's going on at home. 'Cause living here's not easy. (Jenny) This respite was further enhanced by the physical relocation from the "the hustle and bustle" (Rachel) of the city to a natural environment, described as relaxing, peaceful and tranquil. Their opportunity to explore the AT site also gave participants a sense of freedom, autonomy and connection to the location. Participants who chose to explore used their time alone in nature for focused thought and relaxation, or to clear their heads. This was of particular relevance to participants who described an absence of autonomy in relation to the hostels and the wider socioeconomic system. Achievement and mastery 7 Changed worldview Self 6 Others 5 desire to "give something back" (Ben), indicative of initial steps towards community reintegration via a desire to contribute to the AT community.
That'll be good for us physically, mentally, and we're giving something back. (Colin)

| Empowerment
In addition to respite, participants expressed feeling empowered by the intervention. Subthemes were development of a "sense of purpose" and "achievement and mastery".

| Sense of purpose
The contact with nature also facilitated achievement and mastery by inspiring participants to increase their level of engagement or persevere beyond their expectations.

| Changed worldview
The AT experience also challenged participants' negative beliefs, which led to changed worldviews about themselves and others. Six partici-  and their identity becoming associated with negative traits rather than more socially accepted work or family roles (Bentley, 1997;Williams & Stickley, 2011). Attachment of negative stereotypes to a "homeless identity" is suggested to negatively influence selfperceptions of IEH (McCarthy, 2013). Experiencing positive relationships with AT practitioners, peers and nature facilitated development of a sense of community, which provided respite from participants' negative experiences and set the foundation for developing a more positive identity. This community acted as a holding environment, a suitable caregiving environment where a person is physically and psychologically held (Winnicott, 1960;Winnicott, 2005), which facilitated physical and psychological exploration and motivation for change.

| Implications for practice and policy
Relationships with nature also seemed to contribute to the holding environment, supporting previous research outlining nature's capacity to provide a healing environment, where improved emotional well-being, physical health and mental health issues, and reductions in stress, are facilitated (Beringer, 2004;Jordan, 2016;Summers & Vivian, 2018;Taylor et al., 2010). It further appears that contact with nature seemed to provide a space for reflection for participants, hence highlighting the importance of the natural environment when designing psychological interventions for IEH. This also reinforces Revell et al.'s (2014) findings that the opportunity to reflect was considered a helpful aspect of outdoor therapy by participants.
Relationships with peers further facilitated development of a positive identity via peer support, peer modelling, and shared experiences. This fits with previous homelessness and AT research linking peer relationships with service engagement, community F I G U R E 2 Conceptual model showing the relationship between themes and subthemes development and reintegration, and improved interpersonal relating (Barker & Maguire, 2017;Barker et al., 2018;Hanna, 2012).
Furthermore, participants spoke about interpersonal resources that were sustained beyond the intervention, which were previously identified as predictive of housing stability (Aubry et al., 2016), which is recommended as a priority goal for homeless services (Hwang & Burns, 2014).
Participants felt secure in the company of like-minded, homeless individuals, compared with negative experiences of the general public or other residents and staff in the hostels. This could be linked to Baumeister and Leary's (1995) Belongingness Hypothesis, which suggests individuals are motivated by a need for regular social contact with individuals they feel a social connection to, thus, supporting Johnstone et al.'s (2016) findings that the formation of multiple group memberships enhance the well-being of those experiencing and exiting homelessness. Further adding to Barker et al.'s (2018) research, this study highlights that interactions with peers led to development of a sense of purpose and increased desire to engage with the wider community. It also shows that engaging in meaningful activities and having a sense of group identity appears to be a significant motivator for engaging in peer support. Furthermore, informal, non-therapeutic interactions, which can counteract negative experiences of homelessness (Bentley, 1997;Hwang & Burns, 2014;Williams & Stickley, 2011), appeared to be facilitated by the intervention, which may have increased service user engagement. This is supported in AT literature (Gass et al., 2012), which has previously linked engagement with meaningful activity and development of a sense of purpose (Carless et al., 2013;Wilson et al., 2010) and is relevant to IEH due to service engagement difficulties (Hoffman & Coffey, 2008;Rae & Rees, 2015) and fits with recommendations for incremental progression towards community reintegration for IEH; beginning with meaningful activity before leading to stable employment (Parsons & Palmer, 2004).
In the development of a PIE model for IEH, Johnson and Haigh (2010) indicated, "There is no prior assumption of a need for 'therapy', nor is a community focus necessarily central. If a sense of belonging, security and affinity, even of common purpose is developed we may well see this as a good thing; but it need not and usually should not be segregating and exclusive" (p. 33). This appears to be a suitable description of the current AT intervention.
The participants reported belonging, security, affinity (to peers and practitioners) and a common purpose, which appear to have been the main sources of empowerment that they experienced, ultimately leading to changed worldviews. Furthermore, there was no requirement for (talking) therapy to occur; however, the informal nature of the AT could accommodate it if desired, also fitting with homeless service recommendations of flexibility, adaptability and facilitation of therapy in non-traditional formats (Chaturvedi, 2016;Timms & Taylor, 2015).
From their AT experiences, participants indicated increased desire to reintegrate with society, in the form of volunteering, employment and independent living. This indicated the potential for the AT intervention to provide a preliminary step towards this, due to its reported capacity to counteract experiences of stigmatisation and social exclusion. With further development, it could complement a PIE model, which is usually applied to hostel settings (Keats et al., 2012), or be used as a stand-alone model for IEH living in supported hostel accommodation, as in the current study. Based on these findings, a suitable model for further investigation could involve enrolment of AT groups for a programme of AT using an incremental model of difficulty.
Finally, engagement with meaningful activity within the AT community appeared to support participants in the current study to become motivated and to feel worthy enough to reintegrate with wider society. Future developments could expand this meaningful activity beyond the AT community and into wider society. This could include fundraising activities for future AT programmes, which could include selling crafts they produced during the AT intervention at local events or designing an event specifically for this purpose. An alternative development could be in the form of offering links to training opportunities in relation to some of the activities provided during the AT, with the aim of longer-term community integration through professional training and employment opportunities.

| Methodological considerations
An innovative methodological aspect of this study was the use of a repertory grid during a phenomenological interview. Repertory grids are used "merely as a tool" (Kelly, 2017as cited in Ashworth, 2015 16), which helps focus on the subjective experience of participants and should be discontinued if they are not meaningful to participants (Fransella et al., 2003;Jankowicz, 2004;Kelly, 2017). Use of repertory grids appeared to be beneficial for at least six participants, as development of the elements at the start of the interviews encouraged verbal expression and ensured interviews were anchored within their internal world. The repertory grid also appeared to be useful for participants who were less verbally expressive, as it increased the focus of the empathic lens and facilitated exploration of their AT experience, fitting with PCP and IPA, as well as recommendations for working with marginalised populations (Ashworth, 2015;Burr et al., 2014;Fransella et al., 2003;Jankowicz, 2004;Larkin et al., 2006;Odusanya, 2018;Smith et al., 2009). Furthermore, initial use of the repertory grid may have reduced the intensity of the research interview. This was particularly relevant to the current participants due to their experiences of homelessness and the potential for this to have a negative impact on their verbal expression (Gelberg & Siecke, 1997;Padgett et al., 2013). However, it should be noted that most grids were not completed and were only partially used at the beginning of the interview process to facilitate discussions. As the interviews went on, participants stopped using the grid and talked more openly about their experiences. Also, one participant did not find the use of the grid helpful at the start of the interview and preferred the open-ended questions.
Findings were based on qualitative data from a small (N = 7), homogenous sample. Homogeneity was in relation to their "single homeless" status; residence in hostels provided by the same organisation, in the same city in South East England; and engagement with the same AT provider. Due to this, generalisability to other IEH is limited.
Furthermore, the participants were living in supported hostel accommodation and showed elements of housing stability (as identified by a mean duration of 12 months' residence in the hostels). Housing stability and access to financial resources indicated basic physiological and safety needs were being met, which may have made the AT intervention more accessible than it would be to IEH without the means to meet basic needs. Heterogeneity of the participants in relation to gender, age, mental health difficulties and early experiences further limits generalisability even within IEH engaged in the current AT intervention.
The participants also volunteered for the present study. Therefore, experiences of IEH who were engaged in the AT intervention but did not volunteer to be involved may not be reflected in the findings presented.
A further limitation was the time available for interviews, which was approximately one hour. Although it seemed that interviews did reach a natural conclusion at the hour mark, the time limitation might have restricted the level of richness that could have been generated from an interview without a time constraint.
The data and analysis were a result of the interaction between participants and researchers. Therefore, alternative findings may have been developed by different researchers. Attempts were made to reduce this bias and ensure quality assurance with bracketing and reflexivity (Burr et al., 2014;Fransella et al., 2003;Smith et al., 2009).
Member-checking is also recommended for quality assurance purposes within qualitative research (Smith et al., 2009;Yardley, 2015); however, due to geographical constraints and the nature of contact (via the AT provider) this was not possible. However, triangulation of the findings was conducted with the research supervisor, who is experienced in IPA and AT research. This is identified as an appropriate method for ensuring quality assurance, and therefore, the current findings are considered of a suitable quality to support the conclusions drawn (Smith et al., 2009;Yardley, 2015). A final potential limitation was lack of follow-up interviews, which have been recommended for qualitative AT research due to long-term processing of an AT intervention (Bidell, 2010). Previous AT research has indicated continued processing may occur for six months post-intervention (Bidell, 2010), which may reflect participants' continued movement through the experiential learning cycle (Kolb, 1984). Follow-up interviews would have increased the likelihood of capturing their movement through the experiential learning cycle, thereby giving insight into longer-term influences of the AT intervention. Future qualitative studies should, hence, include follow-up data collection and member-checking.

| Recommendations for future research
Additionally, participants expressed the significance of relational development with AT practitioners. Further research into the specific processes involved in AT programme delivery has previously been recommended (Carless et al., 2013); this could include explicit exploration of service users' and AT practitioners' experiences of relational development within AT interventions. Furthermore, exploring the moderating and mediating effect of each aspect of AT would be beneficial in future research.
Larger scale efficacy studies could be conducted with larger samples, thereby improving generalisability and contributing to programme design for wider implementation. This could include a pre-to post-(adventure) therapy, with follow-up design utilising outcome measures to investigate, for example, perceived self-efficacy or mental well-being.
In relation to methodological recommendations, it appears that the use of a repertory grid as an exploratory interview technique aided verbal expression for participants in the current study, which facilitated their reflection on their experiences. Based on this, further exploration of use of repertory grids as an exploratory tool within IPA research, with marginalised groups, or other populations that may experience difficulties with verbal exploration, is recommended.

| CON CLUS ION
Participants' accounts seem to be consistent with existing litera- relating, including self-acceptance, self-value and an increased desire to reintegrate with society. This is characteristic of AT interventions, which encourages self-exploration, empowerment and re-conceptualisations of the self, via a process of reflection and processing (Itin & Mitten, 2009).
The small, homogenous sample limits generalisability; however, findings are indicative of the potential for an outdoor AT intervention to meet the needs of IEH living in supported hostel accommodation.
Future qualitative and quantitative research is recommended to expand on these findings and work towards developing a suitable model of AT for use with IEH. Nevertheless, the findings indicate the potential for AT to meet the service recommendations for IEH; that is, a mental healthcare intervention, focused on engagement, that is flexible and adaptable and facilitates positive relational development (Hwang & Burns, 2014;Spence, 2009;Williams & Stickley, 2011;Woollcott, 2008).

ACK N OWLED G EM ENTS
The authors would like to thank the participants and the AT provider for their participation and openness throughout the research process. The authors would also like to thank Dr Jessica McCarrick for her initial contributions to the project.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflicts of interest.