The project stemmed from an existing youth music mentoring programme, Young Warriors, that focused on building young people’s music skills in school settings. The providers of this programme, the Australian Music Association (AMA), were keen to develop a branch of the programme that would support young people in mental health contexts, and partnered with the researcher to pilot a more therapeutically-oriented design of the Young Warriors model. Young Warriors (used hereon to refer to the current research project) was set up as a group music mentoring programme to offer young people receiving mental health support opportunities to meet and collaborate with others to form bands in a facilitated environment. The programme aimed to foster young people’s social and musical capacities for everyday forms of social participation in the future, for continued recovery, and wellbeing. Approval to run Young Warriors as a pilot research project was obtained through the mental health service’s ethics review board (LNR/16/MH/94). The project was funded by a university research grant and the Australian Music Association.

Young Warriors was run through Headspace: The National Youth Mental Health Foundation that provides individual and group services for young people aged 12-25 experiencing a range of mental health concerns including depression, anxiety, emerging signs of psychosis, personality disorders, as well as issues resulting from gender identity, sexual orientation, homelessness, substance misuse, or disability. Clinicians at Headspace include general practitioners, psychiatrists, psychologists, occupational therapists, and social workers. Services are funded by the national government and are provided free of charge to young people. Young Warriors was open to all young people accessing one of three different Headspace centers in metropolitan Melbourne.

The programme was introduced to staff through several in-services facilitated by the research coordinator (researcher) and advertised to young people through flyers. Young people were referred via clinicians of the service as well as through administration staff and self-referral. The researcher used general clinical reasoning in determining the appropriateness of referrals for the programme, considering group composition, age, and reasons for attending, as well as existing musical skills and interests. Any concerns about the appropriate fit of a young person to the group was approached by talking with the young person and supporting them to come to a decision about involvement. Any young people who subsequently chose not to engage were supported to connect with other music or group therapy opportunities in the service or community if they wished, however this was only relevant for one young person.

The programme was designed to run across eight weekly, 2-hour sessions that were facilitated by a music therapist and a professional music mentor. The exact focus and content of each group was determined collaboratively between the music therapist, music mentor, and the young people, with consideration of what might best support young people’s goals. This approach was informed by recovery-based mental health philosophy that emphasizes collaboration between clients and service providers in working towards individualised health outcomes (Slade, 2009a). However, a basic template was followed for each weekly session, based on theory in adolescent group work (Delucia-Waack, 2006) and the researcher’s clinical experience in the field. This involved a weekly warm up that aimed to bring the group together and prepare group members for the focus of the session. The session then moved into the main activity for the day and closed with the debrief. The programme progressed from earlier sessions focusing on building group cohesion, towards the group’s identified purpose and goal, and involved processes of closure towards the conclusion.

The researcher (and author) acted as the music therapist facilitator for one group and a research assistant was hired as the music therapist facilitator for the other two groups. The music therapists were responsible for obtaining referrals, collaboratively assessing and setting goals with young people, as well as facilitating the group dynamics during sessions and ensuring the mental health needs of young people were adequately supported. The music mentors were responsible for the set up and care of music equipment, the musical skill building of each group, and the music content of the sessions. Facilitators collaborated to review and debrief from each session and plan for the following week.

Collaborative goal setting. Given the practice-based focus of this project, data was obtained from processes that were fundamental to the clinical programme rather than as additional research-specific tasks. The music therapists were responsible for collecting data from their corresponding groups. The primary form of data was collected from young people’s goals for attending the programme. In line with recovery philosophy and theory in adolescent group work (Malekoff, 2007; Slade, 2009a), the groups were formed on young people’s self-identified needs and interests rather than diagnoses. Goal setting was used as a way to gain insight into each young person’s needs and reasons for attending. This process also served to raise young people’s consciousness about what they wanted to get out of their attendance and open up dialogue for how the therapist could best support each young person to achieve their goals and work through any barriers.

To obtain this data, each young person who attended the programme was engaged in a processes of collaborative goal setting, based on a model implemented in recovery-based mental health services (adapted from Clarke, Oades, Crowe, & Deane, 2006) and used in one other study with young people accessing a mental health service in Australia (Schell, Cotton, & Luxmoore, 2012). The music therapist met with each young person individually or via phone to discuss why they were interested in the programme and what they hoped to get out of it. Together, the music therapist and young person created three primary goals for their attendance. Young people were encouraged to generate their goals without suggestions from the music therapist. Where young people needed prompting, the music therapist enquired what they envisaged they could be doing by the end of the programme. Then, the young person was supported to identify what they needed to accomplish in order to achieve this vision. Although most young people easily supplied answers to these questions, goals set by other peers were anonymously provided as examples where needed. Each young person was then asked to rate their competency on each goal pre-and post-involvement in the programme. Rating scales were simple and included options of 0% none, 25% some, 50% keep going, 75% success, and 100% awesome. These ratings were used to prompt therapeutic discussion with the young person about their growth and challenges. Thus, whilst the goals were used as data for the research, their primary function was seen as supporting young people’s mental health outcomes.

Group debriefs. The second form of data was the weekly group debriefs at the end of each session. Debriefs serve an important therapeutic function in groups and are often used at the end of sessions to prompt discussion about material and help people become aware of and process experiences (Delucia-Waack, 2006). Each debrief lasted approximately 10 minutes and was facilitated by the music therapist. During this time, young people were supported to engage in a group conversation about the session’s activities including any challenges, successes, or key moments. The music therapist began with open questions to the group about what stood out that day, as well as make specific reflections that they felt would assist young people in formulating interpretations of their experience in relation to their goals. Young people were also encouraged to raise any other reflections they felt were important, and as the group progressed, members were invited to provide feedback about the content, focus, and structure of the group in ways that could inform evaluation. Debriefs from groups 1 and 2 were audio recorded and used as data. One member of group 3 did not consent to this process being recorded but was keen to participate in the programme. Recruitment for this site was low and so the researchers chose to proceed without including this group in the audio analysis.

The researcher conducted the analysis and consulted with the research assistant weekly throughout the project to discuss emerging findings and any implications for the running of the groups. A senior researcher was also consulted at regular times throughout the project and provided feedback and guidance for the project and analysis as needed.

Research questions 1 and 2 were investigated using an inductive content analysis. This method is often applied as a descriptive approach for both qualitative and quantitative data (Kippendorff, 2004), seemed appropriate to the aim of conveying young people’s reasons for attending the programme, and their perceived experiences of any growth in these areas. Research question 1 What capacities do young people express as most important to their social music engagement? was addressed by analyzing young people’s goals for attendance, to see what emerged as important to young people’s participation in the programme, and whether these related to the programme’s focus of fostering social music engagement. This involved grouping any goals that appeared similar and giving each group of goals a title that reflected the category. Any sub-categories were separated out and given their own title to form more nuanced categories. To identify common reasons for attending, the number of young people with goals in each category was also noted.

Research question 2 To what extent do young people feel a group music therapy programme is able to support the building of their capacities for social music engagement? was addressed through an analysis of young people’s pre-post ratings of their competency on each goal. During the rating process, it was frequently noted that young people struggled to determine the exact percentages in these ratings, but had an overall sense of whether or not they had improved at the goal. The researchers felt that the exact percentage ratings was not meaningful and decided to convert this data to a simplified system where 0 = no improvement, and 1 = improvement. The number of young people who improved in all goals, some goals, or no goals were calculated to assess the degree to which the programme supported young people to improve in their chosen areas.

Research question 3 What group processes appear most relevant to supporting young people’s social music engagement? was explored through an inductive thematic analysis of the audio data. Where the content analysis used for questions 1 and 2 involved grouping and describing the content of the data, this thematic approach allowed for greater interpretation of the material and a focus on the meaning in the data. This analysis involved generating initial codes through in-depth repeated listening to each week’s audio recording, where the researcher initially noted key statements in the audio and any of her own responses to the data. Rather than coding all data, the researcher interpreted which material appeared relevant to the research question, noting the frequency with which it came up across the weeks of the programme, or its apparent significance to the group members. Initial codes were then grouped into larger abstract themes. The researcher’s experience with grounded theory methodology as well as the research question’s emphasis on identifying processes meant that she listened with a focus on what seemed to be happening in the audio and created themes from the codes that focused on central processes. Where the researcher was not the music therapist facilitator, analysis was discussed with the facilitator and their interpretation was used to add depth or alter the analytic ideas as needed. Each week the emerging analytic ideas were presented informally to the young people as part of the debrief. This process was intended to promote feedback from young people about the emerging findings and to provide opportunities for them to respond to or alter the direction of the findings if desired.

At the conclusion of the analysis, results were presented to the supervising researcher and considered with regard to what would be the most useful way to conceptualise the overall findings. It was important to the researcher to focus on what would best promote service development for the young people involved and many others like them. The existing framework of recovery theory provided a natural backdrop for critiquing the findings, while the concept of scaffolding (Wood, Bruner, & Ross, 1976) emerged as particularly useful in conceptualising how services could be developed in this context. These will be presented in the discussion section.

Throughout the project, several strategies were implemented to assist assumptions and biases to be explored in an open and reflexive manner. For instance, the development of the project in response to some young people’s requests for more community-based music options following periods of music therapy in mental health care meant that the researcher found herself assuming young people in this study would desire ongoing forms of music participation in the community. Implementing regular discussion with young people about what they wanted from music programs allowed the researcher to remain open to the present needs of young people in this context.

In wanting the project to ideally receive funding from Headspace services to be replicated in the future, the researcher had to be careful not to unwittingly shape the findings in ways that promoted success of the programme. Regular supervision with an experienced researcher who was not invested in any element of funding or service management helped in remaining open to what the data presented and allowed the researcher to engage in critical discussions about the findings.

In addition to research supervision, weekly peer supervision with the research assistant (the music therapist facilitating the other groups) allowed space for reflection on the clinical aspects of the project as well as open discussion of the dual researcher-clinician roles held by both these women. Being clinician required investment in supporting the outcomes of participants, whereas being the researcher necessitated a grounded and slightly more detached view of what young people gained from the project. The result was a more critical reflection on the outcomes, to address the limitations of the project in a way that resembled the ethos of participatory research (Israel et al., 2008).

The dual researcher-clinician role also raised considerations in terms of asymmetric power relations between the researcher and young people. In recruiting young people to the study, the option to participate in the programme with or without the research component was used as a way to minimize coercion. However, young people were enthusiastic to contribute to the research and no non-research groups were required for this reason. The researcher was previously unknown to the young people, and so the risk of a pre-existing clinical relationship effecting coercion into the study was not seen as a problem. The researcher and research assistant openly discussed the possibility that young people would offer feedback which aimed to please their facilitator and this was talked about with young people at various stages. The data was also critically appraised during analysis to explore whether feedback from young people was genuine.

The age-related power imbalances between the researcher and young people was approached sensitively and consciously. In part, this was done by employing a young music mentor who was similar in age to participants and a young research assistant to run the other group so that each group had at least one young facilitator involved. This strategy aimed to diffuse some of the authority of the older researcher and promote more collaborative dialogue with young people.

Several procedures were in place to assist young participants to understand the process of the research prior to engagement. The governing ethics review board required that participants under 16 years of age had a parent provide consent for their participation in addition to the young person themselves. This was approached by providing both the parent and young person with take-home information about the study and following up with a call. Any interested families were invited back to obtain consent in person. Headspace supported young people to understand their rights in research processes through discussion with a third party prior to any contact from the researchers. The service also had a Youth Advisory Board who provided feedback on all research projects during development and who offered participating young people a point of contact if they had questions or concerns during their involvement.

Another measure of addressing the inherent researcher-participant power imbalance was attempted through a process of collaborative analysis (Hense & McFerran, 2016), whereby the researcher openly discussed what she had interpreted from the previous week’s data in the following session. This process aimed to reduce the disconnect between data and the young people and allowed young people to respond to and challenge emerging analytic ideas. In reality this was challenging to implement and successful only to a limited degree, with young people tending to accept what they heard in the group context. This process may have been more successful in groups where the researchers had existing relationships with participants, a longer period of time to build up relationships, or through individual discussion with young people (Hense, 2017). The 8-week model was simply insufficient to establish this level of trust among the group in order to openly dialogue about such issues. Further attempts were made to combat power imbalances by striving for a more “mutually beneficial” research agenda that aligned with community-based participatory research (Minkler & Wallerstein, 2008). This was done by addressing ongoing service development and access for participating young people after the project. Although it was not possible to provide an ongoing service to young people, the project was greatly extended for one group who requested it, and another group were actively linked in with appropriate music teachers.

Four groups emerged from the project. Group 1: Eight young people engaged for the duration of the programme at this site and the group was divided into two (labeled 1A and 1B) to accommodate the wide age range, variance in musical skill, and goals for attending. Four young people aged 12-15 formed a beginner guitar group (1A) and met weekly for a shared learning experience across the 8 weeks. In order to include one member who was away, this group chose to delay the start of their programme and ran outside the research time frame. Their data was not included in the analysis. This group chose to conclude their programme with a performance for their families. At the conclusion of the programme, all four young people expressed an interest in ongoing guitar lessons and were supported to link in with appropriate local guitar teachers who worked in music studios with group performances and other shared learning opportunities.

The other four young people from this site were aged 19-23 and had existing musical skills. This group formed a band and chose to work on cover songs. At the conclusion of the 8 weeks, this group requested to continue rehearsing with the support of the facilitators but to move to a community-based rehearsal location. This group continued to meet for a further 10 weeks and concluded with a small performance at the end of the year. The group members remained in contact but did not continue music rehearsals past this time.

Group 2: Eight young people chose to engage in the programme at the second site with seven members engaged for the duration of the programme. This group also chose to focus on playing covers but incorporated improvisation warm ups that progressed musically and socially each week. Towards the end of the programme, it was decided that the group would finish as planned and some individuals were supported to link in with ongoing music opportunities in the community. Two young women re-engaged in their school music programmes, and one young man bought a bass and began lessons. Some of the others appeared to need group-based opportunities that were not available in their community.

Group 3: The final group went ahead despite some challenges in timing and recruitment. Six young people were recruited to this group, but only two remained engaged throughout. The only available time for the group was a late Friday afternoon, which proved to be challenging for many young people to attend. Given the inconsistent attendance, this group worked through different music activities weekly and there was no sense that the group would continue to meet at the end of the programme. After concluding the programme, the two engaged group members were supported to connect with local music-based services, however, one chose not to pursue this at the time. The facilitators reflected on the difference in demographics in the area where this group took place, noting the chaotic life circumstances of most of the young people, with transient housing and very limited family support, which impacted their ability to attend the sessions and possibly their views of what community participation could entail.

Research question 1: What capacities do young people express as most important to their social music engagement?

Thirteen young people completed the individual goal setting process yielding a total of 46 goals. Five young people did not complete the goal follow up at the conclusion of the programme, due to no longer being involved, and one young woman completed the programme but did not complete the follow up.

Inductive content analysis of young people’s goals resulted in seven goal categories: being social, building social confidence and skills, building musical skills, reconnecting with music, building modes of self-expression, positive time use, and trying new things (see table 1). The two most prominent goal categories related to the two social and two musical capacities with 19 and 21 goals respectively. The other three goal categories had a combined total of six goals listed.

Table 1.

Goal Categories.

Research question 2: To what extent do young people feel a group music therapy programme is able to support the building of their capacities for social music engagement?

Nine out of the ten young people who completed the goal follow up reported improvement in at least one of their selected goals. Six of these reported improvement in all of their goals and three reported improvement in two goals (see table 2).

The total number of improvements per goal category was also calculated (see table 3). Five out of seven young people reported improvement in the category being social, six out of eight reported improvement in building social confidence and skills, eight out of nine reported improvement in building musical skills and the one person with the goal reconnecting with music reported improvement in this area. Goals building modes of self-expression and trying new things were created by one person each who reported improvement in these areas.

Table 3

Total Improvements per Goal Category.

Research Qn 3: What group processes appear most relevant to supporting young people’s social music engagement?

Two central processes emerged and these were supported by a central theme that influenced young people’s experience of the group and thus enabled the processes to occur. This theme will first be explained, followed by each of the two processes.

Young people frequently described the group as a safe and supportive space (see table 4). This was highlighted in many weeks of the programme by both groups and also referred to as being the reason that other processes occurred. For example, in response to the question, “why did you feel able to try that new song?” a young person replied “because it’s safe to do what you want here.” This safety and security afforded the two other themes that were conceptualised as central processes identified in the groups – making it possible for young people to feel comfortable to express and explore their identity, and to come together in playing music with others.

Table 4.

Quotes Illustrating the Group as a Safe Space. *Quotes are not an exhaustive list of reference to this theme but highlight key phrases that were used for analysis.

The two following emergent processes coming together and constructing identity contain several sub-themes that describe how young people engaged with these tasks. These will now be described. Sub-themes will be presented in italics.

Coming together seemed to occur as a pre-cursor to more familial experiences of belonging or social connection (see table 5). Young people’s descriptions of the closest moments came in music making, where the music itself afforded moments of togetherness and musical cohesion not otherwise experienced in the group. Some young people articulated how these musical moments also supported and shaped relationships within the group, while others noted how getting to know one another outside of their music also contributed to musical connection. Young people appeared to need supported experiences in which they could safely explore modes of musical and verbal communication with others. At a musical level, such gentle experiences afforded the building of musical confidence, and at a social level, appeared necessary for those overcoming anxiety. Although not many young people articulated their experiences of overcoming anxiety, this theme was often acknowledged by agreeing with those who were brave enough to identify themselves as anxious.

Table 5.

Sub-Themes and Quotes Illustrating Coming Together. *Quotes are not an exhaustive list of reference to this theme but highlight key phrases that were used for analysis.

The group provided a safe social space in which young people felt free to express and explore identity, as well as make interpretations of the self and others in the process of constructing a social identity (see table 6). Many young people described coming out of musical isolation having not played with others for a long time, if ever. The group offered a new way to experience the musical identity as a social identity. Young people frequently referred to themselves and others using well known social roles such as those of family members or instrumentalists. One young man articulated how the group environment helped him build self-insight by seeing how his emotions and experiences can be reflected on another and what this looks like as a social identity in the group. This statement stood out among the group, with many members agreeing.

Table 6.

Sub-Themes and Quotes Illustrating Constructing the Social Identity. *Quotes are not an exhaustive list of reference to this theme but highlight key phrases that were used for analysis.

This style of support can be also explored using the concept of scaffolding that is used to portray the temporary but essential nature of the facilitator’s assistance in constantly adjusting the level of support in response to the learner’s competency. One of the challenges reported by educators in applying this approach is that the theory was developed in the more intimate context of carer-infant dyads that allow for flexibility and adaptability in the task compared to the restrictions imposed by group educational settings (Maybin, Mercer, & Stierer, 1992). This feature highlights the suitability of scaffolding to the therapeutic approach where clinicians have the opportunity and skills to adapt and respond to client needs dynamically.

When applied to the Young Warriors project, the theory of scaffolding highlights the need to provide supported therapeutic opportunities across both the mental health service and community contexts in order to foster sustained community-based music engagement. This perspective challenges traditional models of music therapy that tend to be provided in one setting, where a change of context also means a change of service. Although no examples of this model have been documented in music therapy within youth mental health, several authors described elements of this approach in their work in related fields. Fouche and Stevens (2018) described a youth-based programme with South African communities that aims to build resilience among marginalised communities through active music experiences. By bringing together music therapists, local musicians, and school teachers, the programme aimed to build on and strengthen existing resources to promote sustained engagement. Fouche and Stevens demonstrated how collaborative efforts between therapists and community members can help foster an inclusive environment in which personal and social resources and be optimised. Tuastad (Tuastad & Stige, 2015) detailed his work forming a band with in-mates accessing music therapy in prison, to address their goal for an ongoing and independent form of social recreation and income post prison release. Tuastad described working with the band long after their release from prison and illustrated a natural progression of his role from the more conventional music therapist to band member. Although he did not explicitly reflect on this process in the article, Tuastad illustrated the organic transition of his role in response to the band members’ evolving needs. Wood and Atkinson (2004) also described supporting people across service contexts in a project with adults recovering from neurological trauma. Although each individual in their programme was linked into suitable community-based music options at the conclusion of therapy, the music therapist’s role did not involve transitioning with participants beyond the service. In the mental health context, Procter (2004) described challenging the traditional music therapist role in a non-medical community-based service, with an excursion to an art exhibition and impromptu jams with service users outside of the sessions. Although his intention was not to transition people into community group engagement, his approach demonstrated the responsiveness needed by music therapists to support often tricky navigations of the sense of self in a new place.

The combined findings from this study highlight the need for music therapy programmes that not only span both the mental health service and community contexts but engage music therapists in dynamic roles that remain responsive to the evolving needs of young people as they move across these domains. This scaffolding approach should include providing early experiences that foster a sense of safety and allow for foundational social processes of coming together with others and constructing a social identity. These processes may be best supported in the familiar mental health service before extending young people into community settings. Later opportunities should include community-based group work where the music therapist adapts their role in response to young people’s decreasing need for clinical mental health support and increased focus on grappling with the social dynamics of everyday music participation.

This project was envisaged as a collaboration between community musicians and music therapists who could each contribute their expertise to the programme. In line with general best-practice in group therapy (Yalom & Leszcz, 2005), it was planned that this collaboration would be enacted through co-facilitation of the groups by the music therapist and a community musician acting as the music mentor. However, the findings from the audio analysis illustrating young people’s need for safety in the group, highlighted a challenge in the facilitator roles that was subsequently considered in more depth.

The mental health service context naturally drew in a group of young people who were more vulnerable to experiences of failure or exclusion as a result of group music participation. Although the music therapists had experience and training to meet this need for safety within the group, the music mentors were pushed outside their comfort zone when social dynamics became challenging or mental health symptoms interfered with young people’s participation. Although the mentors approached these challenges positively, it became the responsibility of the music therapists to “train” the mentors in how to facilitate these groups and required extensive explanation about the need to prioritise the mental health and social needs of young people over the quality of the musical outcome - a process that altered the dynamic between the two facilitators. Wright (2003) has challenged the assumption that co-facilitation provides the optimal environment for group members, and proposed that although co-facilitation can indeed increase the quality of the experience for members, not all facilitators are suited to provide this level of experience. A study by Okech and Kline (2006) also found that ill-fitted facilitator combinations can detract from the quality of the programme. In Young Warriors, it was felt that the music mentors were restricted in utilising their expertise because the programme design focused on the early stage of therapy when young people were most focused on the social and mental challenges of the group environment rather than the music itself. Further to this, the programme did not ultimately link into the community-based music engagement in which it was envisaged that the mentors’ expertise would be championed.

Altering the model of facilitation to better meet the needs of young people across the trajectory of the programme could optimise their expertise and improve the outcome for group members. Selecting to co-facilitate with Headspace clinicians with mental health specific training in the earlier weeks could help to embed the programme in the mental health service and support the mental health needs of young people as they adjust to the social dynamic of the group. Once young people have built their social competencies and social identities to be more robust, the programme could shift to a community location, co-facilitated by the music mentor along with the music therapist. Playing in groups is a challenging task and many young people may require the ongoing support of a music mentor long after the need for therapeutic input has been met. Increasing the group’s independence by reducing and eventually eliminating the music therapist’s role and allowing the group to function as an on-going facilitated community-based music group could serve as a long term goal.