This project was characterised by its iterative nature and a great degree of uncertainty between the iterations regarding whether the necessary resources would be made available for the project to continue. Due to slow decision-making processes in the organisation and much uncertainty regarding participants’ work contracts each year, we had three to four weeks to prepare for each new iteration. Premeditation of a research design thus was not the main focus, as the practical matters of preparation seemed more pressing (e.g. the actual scheduling of lessons, group work, and preparing the online course, among others).

In the first two iterations, we initially wanted to interview the participants before they joined the project, but this was impossible given the above reasons and the constrained resources. The most practical approach in each iteration was to execute the ‘teaching’ part and then interview the students to inquire about their experiences as an extended version of the regular ‘evaluation of teaching’ that is conducted every year, with a sharper focus on the adjustments that had been made within the project to the regular tuition mode. The interview guidelines are included in the appendixes. In addition to the student interviews, several reflection sessions and teacher interviews were conducted. Table 2 below illustrates what data was collected, when it was collected, and who was responsible for the data collection.

A reflection session halfway through Iteration 2 was very useful, both in terms of yielding the observations of students in the middle of their learning and in terms of allowing the project team to adjust how the teaching would be delivered. The lesson observations that were attempted in Iteration 3 unfortunately could not be conducted to a satisfactory degree due to scheduling challenges. The few lessons that were observed yielded many interesting observations, however, and would be an asset for further research.

The students for the interviews were chosen to represent as broad a range of students as possible in terms of their year of study, background, and gender, among other factors. In the first iteration, the students were chosen so that one student from each year of bachelor studies (four students) and one master’s programme student would be included. The group included a mixture of Norwegian and international students.

All the students were invited to the reflective session in Iteration 2. We offered pizza to increase the likelihood that more students would come. The turnout represented approximately half the participating students from each instrument group. This session included students from every year of study, with a good mixture of Norwegian and international students. As a facilitator, I ensured that all the students would share their thoughts and experiences, and I encouraged them to report on both positive as well as challenging or negative aspects of their participation and learning; the students seemed to be comfortable with reporting on both sides of learning. One variance was that the horn students who attended this reflective session differed from those who were interviewed for Iteration 1.

The students for focus group in Iteration 2 were chosen to represent different years of study and, where possible, different lengths of participation in the project (and thus exposure to the adjusted main-instrument tuition model). By this time, some of the horn students had participated in two iterations, with one also having received AT lessons during the initial project documented by Jørgensen (2015), which essentially made this her third iteration.

All three main-instrument teachers were interviewed in Iteration 2. The AT teacher was not interviewed. In Iteration 3, the interviews were only conducted with those teachers who had completed the full amount of group work and whose students had taken the full number of individual lessons at the time of the scheduled interviews. The three teachers interviewed in the previous iteration were not interviewed again.

All the interviews, focus groups, and meetings were audio and/or video recorded and transcribed verbatim; some of the interviews and other encounters were conducted in Norwegian, while others were conducted in English. The quotations have been edited lightly for clarity for this report, although the editing has not affected their meaning in any way. The original transcriptions are available upon request. The findings presented in this report are based on my own analysis using the qualitative research tool NVivo. A partial analysis of the dataset in Iteration 2 was also conducted by our research consultant in parallel using HyperResearch. These analyses were conducted independently and were later compared; both seemed to yield similar findings.

In my analysis, the teachers’ and students’ transcripts were analysed separately, and an initial reading of the material was undertaken within each group. The interview guidelines already pointed towards certain aspects of the project that we were interested in illuminating and thus contained a certain degree of bias (see the appendixes for the interview guidelines). While analysing the data, however, I did try to keep an open mind for new insights that might emerge. A round of open coding was then undertaken; any utterances with distinct meanings were marked, and a code summarising the content of that utterance was created and assigned to the utterance. As the number of codes grew, I created overarching categories to group the lower-level codes.

After the initial open reading, the second step was to ‘saturate’ the emerging categories, which entailed the undertaking of several new readings of the material. This time I selectively looked for utterances that related to the codes and categories that had emerged in the first step but that had not initially been included in them. During this step, most of the initial categories were able to be saturated; some had to be renamed or regrouped to better reflect the material, and a couple of new categories also emerged. The most saturated categories are presented in this report.

I have played many (and sometimes conflicting) roles in this project. I took the initiative to start this project, with the support of my colleagues, and I acted as a project leader and manager throughout. I have also participated in data collection during the analysis as a participant researcher. I have served as an informant and was interviewed by the research consultant in Iteration 2 because of my role as a main-instrument teacher whose teaching setup had been altered according to this model.

Simultaneously playing so many roles did present a few challenges, but several opportunities also arose that otherwise would not have been available. The difficulty of balancing my faith in the project’s ideas with keeping an open mind in analysing and interpreting the data was ever-present. Lobbying to acquire resources to assure the project’s continuation with impassioned arguments sometimes made it difficult to maintain an analytic distance. At the time of writing, however, no further lobbying is necessary, and taking a more detached stance now seems possible.

While there is the possibility of bias in how the model was developed and the research conducted, the increased number of participants and the engagement of a research consultant in the later stages of the project may provide indications that the project’s findings might be relevant to a larger context beyond my own teaching practice.

Having immediate access to the actual practice and having an ‘insider’ understanding of the practice presented possibilities that might not have been visible if using an outsider, detached perspective. Being in the middle of the teaching practice (and having to make sense of how the new approach could be integrated), seeing how the students responded, and noting what challenges they faced provided questions that we could then ask our research informants. Because the research was more explorative at the outset, determining which questions to ask seemed like an important goal, which I think the project has succeeded in. My closeness to teaching the main instrument might provide some useful insights into how and why the project is relevant more broadly. The project showed that nuance is very important in the material foregrounded in this report; the proximity to the teaching/learning events of the project seems like an asset.

In the findings that present the teachers’ perspectives, only the utterances made by my colleagues will be presented in order to maintain a certain distance, although full objectivity will surely not be possible because of my close collaboration with these colleagues. On the other hand, many of the findings can be triangulated, as they are illustrated from both the teacher and student perspectives and show a good degree of agreement.

Finally, using precise and appropriate language has been a challenge, both in presenting the project in general and in writing this report. Given the large amount of tacit knowledge in the field, and because everyday language is often insufficient for expressing various nuanced points, some of the explanations and expressions might appear cumbersome and contrived to the reader. Because different senses of some words are discussed later, the reader might rightly criticise the lack of consistency in referring to the different senses. The reader is left to judge the credibility of the findings and the aptness of the language.