The survey method was selected to hopefully capture a comprehensive international view of the professional perspectives and approaches of music therapists working with families. The survey questions were developed by the authors through a series of steps with the intention that responses could be completed anonymously by participants via an online platform. The first step involved a series of research meetings. The authors identified key issues (Smith et al., 2016) and discussed differences in terminology according to their own international perspectives and cultural contexts. Through these discussions, diverse definitions and experiences of educational and theoretical frameworks, clinical populations, and music therapy methods were explored. The multiple-choice questions were designed to be easy and quick to answer. Since there was no budget for translation to multiple lanugages, the questions needed to be clear and concrete, and written in accessible English language expression for a professional and multi-lingual audience who are experienced with accessing literature, training seminars and conference presentations in English (Smith et al., 2016).

The second step involved a pilot of the questions. Since the survey was targeted to professional music therapists who define themselves as working in a family-centered way, the authors approached several colleagues from an online support group Music Therapy with Families Network and asked them to complete the questions and provide feedback. The group consists of professional music therapists working with families, many of whom are also experienced researchers in the field, and whose first language is not necessarily English. Altogether, nine evaluations of the pilot questions were received between March and April 2018. The authors then worked to refine the questions into their final format taking into account the feedback provided. The final version of the survey consisted of 22 questions (see Table 1).

The third and final step involved the roll-out of the online survey via Webropol. The survey was open from 13.9.2018 until 7.1.2019. An invitation to participate in the survey was published in several closed Facebook groups ² including the Music Therapy with Families Network (275 members), Music Therapy in Child Welfare (128 memembers), Music Therapists Unite! (5901 members), School Based Music Therapists (335 members), Music Therapists Working in Mental Health (953 members), and Music Therapy and Hospice & Palliative Care (1112 members). In addition, national Facebook pages for professional music therapy associations were invited to post an invitation, including China, India, Latin America, Spain, Australia, Israel and Finland. E-mail invitations to participate were circulated to members by the World Federation for Music Therapy, European Music Therapy Confederation, and British Association for Music Therapy. All announcements and invitations were posted up to three times. Individuals were also encouraged to forward the invitation to other colleagues. Despite the high numbers of people in each professional group, it is likely that the same people were members of multiple networks/groups. Therefore, it is not possible to estimate the final invited sample size.

At the beginning of the survey, the respondents were asked to authorize that their data can be used by the research team for the purposes of the study. The study follows the ethical codes of the University of Jyväskylä, Finland. While the online platform did not collect names and contact information, any survey responses in the open comments were checked and deidentified prior to analysis. The survey questions are provided in Table 1.

A total of 134 people responded to the survey. Of these, nine people indicated that they were not trained music therapists, and were therefore removed from the analysis. The final number of complete responses was 125.

Of the 22 questions, 19 were multiple choice and three allowed a free open-text answer. The respondents were asked to answer every question, with several multiple choice questions including an “other” option that also allowed for further explanation via an open-text field (see Table 1). The complete survey is provided in Appendix 1.

The first step when analysing the data was to examine the “other – please describe” free-text answers to the multiple-choice questions. The first author read through the free text and determined if the answer could be incorporated into the existing categories. If it could not, a new category was proposed and discussed by all authors. In this case also those entries mentioned only once were categorised as their own, aiming to present the picture of the data as authentic as possible. The meaning of some answers were unclear, provided feedback on the survey question, or more conversational in nature and were excluded.

Next, the three open-ended questions which invited a free-text response underwent a qualitative content analysis (QCA). Using the guiding question “What is intended to be said?” (Bengtson, 2016; Bruscia, 2016), the first author worked to systematically analyse and classify the text into an organised and concise summary of key categories (Bruscia, 2016; Erlingsson & Brysiewicz, 2017). The systematic coding was carried out in an inductive way in order to identify meaningful themes that addressed the research questions (Bengtsson, 2016). The first round of coding was broad and aimed to stay faithful to the original text and expressions of the participant. Next, the codes were categorized by grouping related codes together, and discussed amongst all authors. Finally, the frequency of comments related to each category was descriptively analysed.

The respondents were mostly female (90%) and aged between 30-39-years-old. There were no participants over 70 years of age (Table 2).

Most respondents stated their highest qualification in music therapy to be Masters (44%), followed by Bachelors (21%) and Doctoral (18%). Only 2% of respondents indicated that they had a pre-Bachelor (sometimes called ‘clinical training’) qualification. Further, 19 respondents had aquired additional music therapy training, including GIM (Guided Imagery and Music Bonny Method; n = 4), NICU (Neonatal Intensive Care Unit music therapy; n = 3), NMT (Neurologic Music Therapy; n = 3) and APCI (Assessment of Parent-Child Interaction; n = 1).

Geographically, most respondents reported that their first qualification was undertaken in Europe (n = 54) and North America (n = 43), followed by Oceania (n = 16), Asia (n = 9) and Latin America (n = 3). There were no respondents from Africa. In response to the question “In which country are you currently practicing music therapy?” there was no significant difference compared to the respondents’ country of qualification (Figure 1).

The highest number of respondents reported to have begun working as a music therapist within last 7 years (44%, n = 55), while 28% of the participants had been working for over 18 years (n = 49). Further, 72% (n = 91) of the respondents reported that they began working with families within 2006–2018. However, according to this sample, the more experienced music therapists reported working with families as early as the 1980s (Figure 2).

Music therapy practice is often highly varied, and this trend was reflected in the data. Many of the respondents reported that they work with several clinical populations. Therefore, there was a total of 381 selected answers to question 11. When “other” responses were added retrospectively to the initial options, there 415 populations selected by 124 participants (Figure 3).

According to these results, disability was the largest clinical population in which music therapists work with families. Of these, 16.9% (n = 70) of the respondents work with preschool aged children with disabilities and 13.7% (n = 57) with school aged children with disabilities. Mental health was the next most common population for music therapists working with families, with 9.4% (n = 39) working with children, 7.5% (n = 31) working with adolescents, and 2.7% (n = 11) working with adults. Families at risk/child protection were also highly represented, with 8.7% (n = 36) of music therapists working with this population.

From the “other – please describe” comments, three new categories were constructed, including 4.3% (n = 18) of respondents who indicated that they worked with families in hospice or palliative care settings. However, other responses were more difficult to categorise where they did not refer to a specific clinical population, such as “special needs,” “public school” or “mainstream children’s center and school.” The authors considered that these answers could be referring to children with behavioural problems, ADHD or learning disabilities or children with no specific diagnosis. Therefore, a “Children – general” category was established. In a similar way, “Older adults – general” was added as a category even though this was represented by only 0.2% of the respondents. Additionally, respondents indicated that they work with populations including emergency settings post-disaster and conflict, military families as well as asylum-seeking families.

While there is great variety, when clustering the results into broader categories, the dominance of certain populations became more apparent. According to these results, 35% of music therapists working with families work in the field of disability, and 20% in mental health (Figure 4). If the categories of families at risk/child protection and social care settings were combined, 12% of the respondents could be classified as working in this area. Similarly, 13% of participants work in medical settings with clients of all ages. Dementia care/seniors and hospice/palliative care may not be easily combined, since end-of-life care involves clients of multiple age groups.

Taking this broader view one step further, an approximate analysis of the age distribution could also be made. Based on the population descriptions, it seems that 79% of the respondents work with children and adolescents, while 21% of music therapists surveyed work with adults (Figure 5). To avoid ambiguity, the categories of hospice/palliative care, community based preventative programs, and social care settings were left out from this age analysis because the exact age was able to be determined.

The survey findings revealed that music therapy with families commonly takes place in community settings (n = 54; Figure 6). This category included i.e., music centres, music schools, community centres, and libraries. Hospital/medical settings, including hospice units (n = 53) and music therapy taking place at the client’s home (n = 50) were also common. Specialist multidiciplinary services clinics (n = 24) included i.e., family rehabilitation centres and centres specialized in pregnancy, birth and early parenting. From the “other – please describe” comments, one new category was constructed: “Residential care facility for older adults.” This category includes rest homes, nursing homes, assisted living communities and seniors home.

Respondents indicated there was a large variety of theoretical frameworks applied to working with families (Figure 7). The responses indicate that each music therapist on average has three theoretical influences in their work. The humanistic framework was the most salient with 72% (n = 90) of people indicating they align with this theory, including more specific approaches such as wellness based theories, validation therapy, and existential and phenomenological viewpoints. Developmental frameworks (n = 55) included play-based interventions and Floortime. Psychodynamic (n = 48) and resource oriented (n = 41) approaches were both well represented. Integrative (n = 39) and systems/ecological oriented (n = 34) were nearly equally often mentioned as well as neurological (n = 19) and behavioral (n = 19) approaches.

From the “other – please describe” comments, five new categories were constructed. Three respondents described their approach as based on attachment theory. The authors debated whether this approach could be considered part of psychodynamic theory, but ultimately could not be sure given that the participants had included this answer within the “other” response. The “narrative” framework (n = 3) was also included as a new category. Only one respondent described “mentalization,” and similarly the authors debated whether this approach could be considered as belonging to the psychodynamic framework. However, it seemed important to emphasize this approach, especially when working with families, and therefore it remained as a separate category. “Music-centered” (n = 1) and “interactive” (n = 1) approaches were included as their own categories as well since both seemed to accent particular features of their framework.

For question 14, “What music therapy methods do you use when working with families in music therapy sessions,” multiple answers were possible resulting in 634 choices from the 125 participants. This equates to an average of 5 methods per person, suggesting that a large variety of music therapy methods are relevant to working with families. The most commonly reported methods were improvisation with instruments (n = 115), singing pre-composed songs (n = 101) and structured activities with musical instruments (n = 91). Also, improvisation with voice (n = 85), music listening (n = 79), music and movement (n = 78) and song writing (n = 63) were commonly reported.

From the “other – please describe” comments, two new categories were constructed. The first was integrative methods and integrating musical activities, where different methods were used in a holistic and dynamic way depending on the needs of the family. The second was including other non-musial methods in music therapy sessions, such as the use of pictures, play, meditation, story telling, and Eye Movement Desensitization and Reprocessing (EMDR).

These results could also be clustered into broader categories as follows: 1) singing, including pre-composed songs and improvisation with voice; 2) playing instruments, including structured activities with musical instruments and improvising with instruments; 3) music listening, including Guided Imagery and Music (GIM); 4) music and movement; 5) song writing; and 6) other. When responses were analysed from this broader viewpoint, singing (29%) and playing instruments (33%) together accounted for 62% of the data (Figure 9).

According to this survey, “consultation and discussion” was the most popular non-music-based technique, with 82.2% (n = 104) of respondents stating they use this approach with families in music therapy sessions (Figure 10). This approach included several ways of working, including therapeutic discussion, verbal processing, reminiscing and life review. Imaginative play with toys (n = 53), art-based methods (n = 48), and playing games with rules (n = 41) were also used frequently.

“Techniques from other therapeutic approaches” was a new category developed during the analysis of the free-text responses. This category (n = 9) included approaches such as Theraplay®, narrative exposure therapy, Adaptive Mentalization-Based Integrative Treatment (AMBIT) and Mentalization Based Treatment (MBT), trauma-informed care approaches, and cognitive therapy. The integrative methods category was also retrospectively added to acknowledge the flexible, shifting and dynamic way of working described by one respondent. In addition, two other categories were added based on the free-text responses, including: multisensory activities (n = 1), meaning multisensory actions (lifting, waving); and interactive play (n = 1) including the use of early childhood play/games between the child and the carer.

The most selected answer to question 16 “Which of the following models best describe your work”, was “family members are active participants in music therapy sessions with the child/adult client” (93.6%, n = 117). Forty of the respondents (32%) reported that family members were present but not active in music therapy sessions. Taking these two categories together, it is therefore much more common for family members to be present in the music therapy session with the child/adult client than not. Even so, 33 participants (4.8%) reported that they conducted “separate/additional counselling sessions” for family members.

Analysis of the free-text responses highlighted that people also use a combination of models. Therefore, a new category of "integrative methods" was created to reflect this approach. Further, in the free-text response, one person described a model where family members participated in separate music therapy sessions provided by another music therapist. A new category was created to capture this reponse.

When these results were clustered into broader categories, results showed that 77.0% (n = 157) of respondents stated that family members pariticipate in music therapy sessions, either actively or more passively. Counselling sessions provided by the same music therapists in individual or group meetings appeared in 19.1% (n = 39) of the answers. Family members observing the session from outside the therapy room, along with family members’who received separate music therapy sessions, were clustered into the “Other” category and covered 3.9% (n = 8).

With the earlier clustered data from question 11 indicating that 79% of respondents work with children or adolescents, it is perhaps not surpising to see that 86.4% participants indicated that the parent(s) (n = 108) were most often present in music therapy sessions, followed by sibling(s) (43.2%, n = 54). For those music therapists working with adult clients, the data also shows that the partner/spose is included 32.8% of the time (n = 41). There may also be other extended family members (n = 34) and grandparent(s) (n = 30) included in music therapy sessions.

The question concerning counselling aimed to map how frequently separate counselling sessions with family members occur. However, it should be noted that 54.4% (n = 68) of respondents stated that the question was not relevant to their work. Therefore, the actual analysis included only 59 answers (Figure 14). According to this data, separate sessions for family members typically take place less frequently than sessions with the client (n = 39). Only one respondent (n = 1) mentioned that the counselling sessions take place more frequently than sessions with the child/adult client. One new category was constructed based on the free-text analysis: The frequency varies depending on the client’s needs (n = 6). Again, in this question music therapists seemed to advocate for flexibility in their practice and explained that the frequency depends on the demands, goals, context and needs of different cases.

When asked about the most common techniques used in these separate counselling sessions, discussion and consultation was highly reported (n = 68; Figure 11). Music therapy methods were also used widely within counselling sessions. Improvisation with instruments (n = 38), music listening (n = 28), song writing (n = 24) and improvisation with voice (n = 22) were all mentioned. From the non-music-based techniques, the use of video feedback (n = 19) was most common. However, similar to the question above, 40% (n = 50) of respondents chose the option “not relevant to my work.”

While analysing the free-text response, four new categories were constructed: techniques from other therapeutic approaches (n = 3) which included Mentalization Based Treatment (MBT) for families, breathing activities, and mindfulness. Modelling was formed as a category of its own (n = 1), including modelling Applied Behavior Analysis (ABA) techniques. The use of a self-report assessment form, Spence Children’s Anxiety Scale (SCAS) for parents was placed in an idependent category of self-report assessment forms (n = 1). Using movement to music was categorized in the integrating musical experiences with other methods category (n = 1).

Question 20 provided opportunity for a free-text response: "How would you describe your role as a music therapist working with families?" There were 105 responses to this question, which underwent a qualitative content analysis. The analysis generated 159 codes which were then further grouped into 12 categories. Table 3 shows the final categories along with the numbers of individual codes within each category and a summary description incorporating exemplars of the words used by the respondents.

Table 3

The role of the music therapist when working with families (n = number of codes)

A descriptive quantitative analysis revealed that the first category “To share their expertise: as a counsellor, teacher or guide” included 45.7% of the responses. The role of the music therapist was described as “to support” in 35.2% of answers, whereas the role “to provide, create, and offer” followed with 25.7% of responses.

Figure 16

Role of music therapist when working with families (n = 105)

In response to question 21, respondents identified that there are some specialist training courses in working with families in different countries. The free-text responses included the following training courses: assesment of parent-child interaction (APCI) by Lindahl-Jacobsen (i.e., Jacobsen et al., 2014) originally developed in Denmark; a short course in family centered music therapy and dialogic parent counselling by Tuomi and Jordan-Kilkki from Finland (Jordan-Kilkki & Tuomi, 2016); specialist courses focused on music therapy in hospice settings (no trainers or names were provided); a short course in music therapy and families from Spain; neonatal intensive care unit (NICU) music therapy training (www.nicumusictherapy.com); and the “Sprouting Melodies” training model available in the USA (www.sproutingmelodies.com). In addition, respondents broadly described workshops, visiting lecturers and short courses taking place in various locations around the world for music therapists working with families.

In response to question 22, there were 92 participants who provided their perspective, insights and ideas. The high level of engagement in this question perhaps indicates the passion these respondents have for promoting further skills and training in music therapy with families. From these 92 responses, 120 codes were identified through a Qualitative Content Analysis, and from these codes 11 categories were formed. Some respondents indicated that short courses (n = 5) and advanced training seminars (n = 5) for qualified music therapists are needed; however, others indicated that training should also take place within clinical placements (n = 4) and as part of initial music therapy qualification courses (n = 3).

The categories describing the focus of future training courses in music therapy with families could be clustered into three main themes: Theory, Practice and Context (Table 4).

Table 4

What should be included in music therapy training programs and continuing education (n = 92)

The results identify that the role of music therapists working with families is broad and versatile. Lindahl-Jacobsen & Thompson (2017b) mapped out a model for the therapist’s role (also described as their “stance” or “position”), to encourage music therapists to consciously consider their approach when working with families. Their model proposed that there are three continua interacting together that the therapist might reflexively consider, including: 1) the way the therapist guides and challenges the family, ranging from a more expert position through to an equal partner; 2) the way the therapist shares knowledge and assists the family, ranging from a more directive approach through to more supportive problem solving; and 3) the degree to which the therapist engages with the family, ranging from a more distant outsider to a close insider. Rather than a protocol for how to work with families, Lindahl-Jacobsen and Thompson (2017b) stress that there is no “best” stance but merely a “best fit” for each family and context.

In reflecting on the analysis of the free-text responses to question 20, the 12 categories could be mapped within the first two dimensions of Lindahl-Jacobsen & Thompson’s (2017b) model when they are presented as four quadrants. When considering the results through the lens of this model, music therapists adopting a more supportive–expert role are in the majority (n = 95), followed by the directive-expert role (n = 57). However, the work cannot be interpreted in a binary way as any model might imply. The respondents in this study frequently highlighted how their approach is more likely to be dynamic and responsive to the context.

With music therapists belonging to numerous professional groups and no single international registry for qualified music therapists available, it was difficult to estimate the expected sample size. The lack of statistical data for the profession may contribute to challenges with validity and have implications for study replication. While a variety of countries are represented in the sample, the fact that the survey was only available in English may have been a barrier to participation. Future studies should include funding to enable translation of surveys to several international lanuauges to promote participation. In addition, funding would have enabled access to resources to support recruitment and advertising which may have increased accessibility and the number of responses.

Formulating multiple choice questions for an international audience is also challenging. Despite careful consultation in the pilot stage, different traditions, terminology and cultural considerations might not have been adequately included. This challenge may be reflected in the need to add new categories during the analysis of the free-text answers to the multiple choice questions.

While surveys are useful in collecting a breadth of persectives, depth is limited. For example, the results do not explain when, how and why (or why not) particular music therapy methods are used with families or how they are put into action. A follow up interview study could further explore these deeper questions.

Within the data, there are valuable suggestions for future training and education of music therapists who wish to work with families. For example, within the non-music-based methods, verbal facilitation skills are commonly used, yet respondents see this area of practice as needing further training. These results are echoed in previous research from NICU settings (Gooding & Trainor, 2018). While there is some music therapy literature exploring the use of verbal facilitation skills (i.e., Amir, 1999; Gooding, 2017; Lindblad, 2016; Nolan, 2005) more research is needed.

Beyond the profession of music therapy, the importance of therapists adopting mentalization approaches to increase the family’s capacity for reflective functioning is highlighted in the broader research into family work across clinical populations (i.e., Dimitrova et al., 2016; Fonagy, 2012; Fossati & Somma, 2018; Kalland et al., 2016; Pajulo et al., 2012; Philipp, 2012; Solbakken et al., 2011). The ability to mentalize is seen as a crucial part of parenting and is especially important when there are challenges in the child's development. In this survey, only one respondent mentioned including mentalization theory as part of their approach. Within the music therapy literature, mentalization is more commonly described in work with adults with mental health issues (Hannibal, 2014; Hannibal & Schwantes, 2017; Strehlow, 2016). In the field of music therapy with families, there have only been preliminary discussions about incorporating mentalization theory as part of recent conference presentations (Lindahl-Jacobsen et al, 2018; Tuomi, 2018, 2019). It is important to acknowledge that working in this way requires advanced training and/or counsultation with other professionals from this field, such as family therapists. In addition, given that the broader field of family therapy includes mentalization as a key theoretical framework, there is scope for further research in this area in music therapy.

In terms of the role of the therapist in working with families, this survey only provides the therapists’ perspective. Studies exploring the outcomes of family-centred sessions have demonstrated that parents and family members often gain knowledge and skills from participating in the sessions (Thompson, 2018; Schwartzberg & Silverman, 2017; Warren & Nugent, 2010), or from receiving parallel counselling sessions (Blauth, 2017; Gottfried, 2016). In either approach, the music therapist’s facilitation style was important to the perceived success of the sessions (Edwards, 2014; Nicholson et al., 2008; Thompson, 2018). Future studies should consider researching the role of the the therapist from the family’s perspective.

In addition, the survey results cannot provide a deeper insight into who is considered the “client” in family-centered music therapy sessions. In other words, is the focus on the child/adult client, on the parent/carer or on the whole family? When reflecting on the analysis to the open-text questions, it seems that the participants conceptualised their work with families as involving a child/adult client who are accompanied by others who share the session with them. This topic needs further research to better understand practice, since there are flow on ethical implications for determining the goals/focus of therapy, and for raising awareness about the potential benefits of music therapy with families. Further, more research exploring how music therapists interact with family members who are not present within the client’s session, and who are not receiving parallel services, is needed.