Critical aspects to consider when establishing collaboration between school leaders and researchers: two cases from Sweden

ABSTRACT

The aim of this study is to explore what happened – and why – in a collaboration between ourselves as researchers and school leaders from two different schools in Sweden. The theory of practice architectures is used to frame the study. The goal of the collaboration was to enhance the scientific foundation of school practices, which motivated the use of action research as the methodological approach for designing the collaboration. The results allow us to propose a number of critical aspects to take into consideration when establishing a collaboration between researchers and school leaders. The first aspect concerns aiming for co-ownership and an equivalent collaboration. These involve trustful and respectful relations, empowering individuals in their professional roles. The second aspect is linked to attention to local infrastructures and the importance of organisational settings that facilitate a holistic approach and a shared understanding of the aims of both the research and the collaboration. The third and final aspect is related to school leaders’ legitimacy to organise research-based school improvements. School leaders are responsible for educational quality in schools, and this makes them best suited for leading school improvements. However, theoretical frameworks and methods introduced by researchers can support the ambition to enhance the scientific foundation of school practices. We advocate the development of practices inhabited by school practitioners and school researchers within the educational complex. In addition, we promote the creation of healthy environments in the form of arrangements that create spaces where equivalent collaborations, shared responsibility and co-ownership may grow.

KEYWORDS:

• Collaboration
• co-ownership
• equivalence
• school leader
• theory of practice architectures
• action research
• school improvement

Introduction

This article focuses on collaboration between researchers and school leaders, using action research as the methodological approach for setting up the collaboration. The study was conducted within a Swedish national project carried out under the acronym ULF (Utveckling, Lärande, Forskning [Development, Learning, Research]). ULF was founded by the government and is, therefore, a prominent project in the current Swedish educational context. The government’s main goal with respect to ULF is to enhance the role of research in the daily work of teachers and school leaders by making research more directly applicable in practice. This is mainly carried out by stimulating the development of collaborations between universities and schools. This is in line with what Carr and Kemmis (1986) noted a long time ago: that school leaders and teachers have to be involved if we really want to transform the work that is carried out in local schools. There is no consistent definition of this kind of collaboration (Rudge 2018), and this article does not examine its various meanings. Instead, the practice of collaboration is explored. In particular, this study seeks to critically examine the necessary conditions for the kind of collaboration that is requested within ULF. The self-experienced collaboration between ourselves, as researchers, and school leaders from two different school settings constitutes the foundation. The term school leader is used as a collective term, referring to both principals and school managers.

The use of the theory of practice architectures (Kemmis et al. 2014) as a theoretical lens provided an accessible and concise language for describing and understanding the collaboration. The theory steered the analysis towards what actually happened in practice and allowed us to identify empirical connections between different practices and different arrangements. In the form of narratives, we were able to explicate other coexisting practices and illuminate the arrangements – in the form of practice architectures – enabling and constraining our collaboration. The theory brings conditions such as norms, practical set ups, and relations of power and solidarity among participants into focus (Olin and Pörn 2021).

According to the theory of practice architectures, practices are, ontologically, located in specific sites, with their own specific practice landscapes and practice traditions. Landscape refers to the particular practice setting where different practices coexist and where there are often overlapping sites of practices (Mahon et al. 2017). Practice traditions, on the other hand, refers to a form of collective memory – or norms – regarding how to do and understand things. The ULF project provides a practical setting in the form of resources, goals and directives for how the work is to be carried out. ULF also brings along a tradition, in the form of the belief that teachers and school leaders can come to terms with problems and shortcomings that exist within local schools through the support of research. In Sweden, this belief also appears in the Educational Act of 2010, which specifically points out that education should be based on research and proven experience (SFS 2010, 800, chapter 1 § 5).

As outlined, this study is part of a comprehensive ULF project, which is an important coexisting practice. Participation in ULF is voluntarily, and the project was initiated by a team of researchers with funding from ULF. The ULF research team is a multidisciplinary team consisting of researchers in didactics (in English more commonly pedagogy) and school leadership. The focus of the comprehensive ULF project is on teachers’ practices, more specifically teachers’ collegial planning and preparation of lessons. Teachers are important actors in schools – in the so-called educational complex – which will be outlined in the theoretical framework section (Mahon et al. 2017). Even if the teachers were not the focus of this study, their participation is mentioned in the results, as their practices affected our collaboration.

Research shows that school leaders play an important role in students’ learning (Jarl, Blossing, and Andersson 2017; Robinson, Lloyd, and Rowe 2008), which motivated this collaboration focusing on school leaders’ responsibility for arranging the settings for teachers’ professional and collegial work (Timperley 2011). The intention of our part of the collaboration was to enhance the role of research in school practices, which is in line with the ULF project. More specifically, this would be done by entering a joint action research project supporting the school leaders in leading and organising teachers’ collegial planning and preparation. During the action research process, questions were raised about how the school leaders understood their role in the collaboration. This turned our attention to the prevailing expectations within ULF concerning collaboration between researchers and practitioners as a prerequisite for making research applicable to teachers’ and school leaders’ daily work.

This study provides a close exploration of a collaboration involving practitioners and researchers. The aims were to explore a specific collaboration between researchers and school leaders and to expand knowledge of what enables and constrains such collaboration. Our research questions were as follows:

(1) How did the collaboration between the school leaders and the researchers emerge in practice?

(2) What shaped the collaboration between the researchers and the school leaders?

The focus on collaboration as a practice drew the attention away from the participants and towards what actually happened in the collaboration and why things happened. The contribution of this study lies in providing critical aspects to consider when establishing a collaboration between researchers and school leaders.

Developing collaboration through dialogue

As mentioned, the ambition to enhance the scientific foundation of Swedish schools calls for collaboration between researchers and practitioners. As the professional practices of these groups are separated, their interests are often difficult to unify (Levinsson 2017). Still, there is a long tradition of focusing on issues around collaboration between practitioners and researchers, and there is interest in how to improve this kind of collaboration (Ainscow, Booth, and Dyson 2003; Bruce, Flynn, and Stagg-Peterson 2011; Carlgren 2009; Edwards-Groves, Grootenboer, and Rönnerman 2016; Forssten Seiser 2019; Olin, Karlberg-Granlund and Furu 2016; Pennanen et al. 2017). However, as Olin and Pörn (2021) noted, a limitation of many of these studies is that the focus is mainly on practitioners. By exploring collaboration as a specific practice, we focus on neither practitioners nor researchers; instead, the focus is on what happened in the collaboration as a specific practice in a specific site.

Buick et al. (2016) identified six strategies of importance for developing a close and equal collaboration between practitioners and researchers. In this study, these strategies were used for exploring what happened in the meetings between ourselves and the school leaders. How this was done is described in the data and method section. The first strategy focuses on the need for shared agreements regarding the purpose of the research. Second, collaborations require enabling governance arrangements, such as joint accountability for the work. The third and fourth strategies emphasise a joint approach to data analysis and research dissemination. The fifth strategy highlights the need for the mindful composition and management of the participants. The final strategy is connected to personal relations, including the need for trust and respect.

Equivalency in collaboration occurs when researchers and practitioners engage with one another over the long term to produce sustainable knowledge (Buick et al. 2016; Forssten Seiser 2019). A collaboration where participants act and feel like equals requires a dialogue that bridges professionals’ specific cultures. Studies have shown that increased dialogue can improve universities’ learning capacity (Strandli Portfelt 2006) and enable more equivalent collaboration in action research projects (Edwards-Groves, Grootenboer, and Rönnerman 2016; Francisco, Forssten Seiser, and Grice 2021). However, genuine dialogue is challenging and usually occurs only in short sequences, as researchers and practitioners often remain in their different professional roles. This division may be bridged by creating a third organisational logic that, according to Kieser and Leiner (2011), goes beyond traditional logics. Such a third organisational logic can, from an ecological perspective, be described as a niche, that is, a practice surrounded by enabling practice architectures (Kemmis et al. 2014). From an ecological perspective, practices are seen as living organisms that are born, grow and die in an ecological relationship with other practices (Kemmis et al. 2014). The survival of a practice is dependent on the surroundings. In unhealthy environments, practices will slowly vanish and die, while in healthy environments, they will flourish.

There seems to be a quality in the relatings and dialogues between researchers and practitioners in collaborations that needs to be further understood and explored. Here, we turn to Habermas, who perceived intersubjective agreements to be the basis for mutual understanding (Habermas 1996). The reading of Habermas encouraged us to use what he described as communicative spaces as a role model for designing our regular meetings with the school leaders. By opening up communicative spaces, partners can develop a genuine and meaningful dialogue, interpreting their contexts and practices within deeper common understandings (Habermas 1996). Another ambition within communicative spaces is to establish unforced consensus about what needs to be done to improve practice. Such consensus is achieved through dialogue over time (Kemmis, McTaggart, and Nixon 2014). By focusing on the relationships between the participants, respect, trust and common understanding can develop and positively influence communicative action in the collaboration (Edwards-Groves, Grootenboer, and Rönnerman 2016).

Theoretical framework

Fundamental to the theory of practice architectures (Kemmis et al. 2014) is the attention given to the arrangements that enable and constrain specific practices in a specific site. According to this theory, a practice is understood as a socially established cooperative human activity constituted by the sayings, doings and relatings that hang together in a joint project. A project encompasses the intentions or aims that motivate the actions undertaken in the conduct of the practice. The collaboration between researchers and school leaders with the ambition of enabling research that permeates school practices was the project that motivated the practice in this study.

The notion of practices hanging together in a project is critical for ‘identifying what makes particular kinds of practices distinctive’ (Kemmis et al. 2014, 31). Additionally, the theory holds that a practice is prefigured by the practice architectures that are present or brought into the site of the practice. Practice architectures are the particular arrangements that together shape, and are shaped by, the practice. Practice architectures appear in three intersubjective dimensions and are made up of different forms of arrangements (Kemmis et al. 2014; Mahon et al. 2017). Cultural–discursive arrangements operate in the semantic dimension and appear in the medium of language. These arrangements enable and constrain the sayings in and about a site. For example, they affected what was said (or not said) in the conversations that were undertaken during the meetings between the researchers and the school leaders. Material–economic arrangements operate in physical space–time and appear in different activities. They enable and constrain the doings in a site. For instance, they affected when and where the meetings took place and the extent to which these conditions supported dialogue and collaboration. Social–political arrangements operate in the social dimension and appear in the form of power and solidarity. They enable and constrain the relatings in the practice. For instance, they affected who led the meetings, who was listened to, and what or who was ignored. They reveal how people relate to one another as well as to artefacts, both inside and outside the practice.

The theory of practice architectures also gives importance to the notion of leading in relation to the other practices that constitute the educational complex. The educational complex has been identified as a group of five practices with a long tradition of being interconnected: students’ learning, teachers’ teaching, professionals’ learning, school leaders’ leading and researchers’ researching (Mahon et al. 2017). The different practices in the educational complex are interrelated in that as one practice unfolds, there emerge practice architectures that enable and constrain other practices. In this article, we focus on how researching became practice architecture for school leading and vice versa, exploring the arrangements that shaped, and were shaped by, the collaboration between the researchers and the school leaders.

Data and method

As noted, methodically, this study focuses on the collaboration between researchers and school leaders designed as action research. In School A, the collaboration involved nine school leaders in a Swedish upper secondary school and two researchers. The school had 2,000 students and 260 staff members. The school leadership team included seven principals and two school managers. Five of the principals and the two managers participated in the project. School B had 530 students and 85 staff members led by a team of three school leaders, including one principal and two assistants. The collaboration in School B was conducted between the principal and one of the researchers involved in School A. In School A, the school leaders and the researchers met seven times over 14 months. Due to the ongoing pandemic, the last four meetings were online (Zoom). Focus group discussions Yin 2014) were used as the main method of data collection. In School B, the participating principal and the researcher met six times over 12 months, and four of the meetings were Zoom meetings. The recordings from the meetings (9.5 hours in School A and 6 hours in School B) were transcribed.

The transcribed recordings were analysed in three analytical activities: 1) data condensation, 2) data display and 3) conclusion drawing (Miles, Huberman, and Saldaña 2014). Figure 1 is a schematic description of the analysis process, illustrating how the theoretic frameworks were used in the analytical process. This is described in detail in the following sections.

Figure 1. The analysis process.

The first activity (data condensation; see Figure 1) was a selective and focusing process. In this activity, the sayings, doings and relatings (Kemmis, McTaggart, and Nixon 2014) were coded in line with Buick et al.’s (2016) six identified strategies to explore what happened at our regular meetings. This first stage of the analysis was conducted using the NVivo 11 data program. To make the coding sharper, we individually coded the same data and then compared and discussed our coding. The two cases were coded and analysed independently due to their dissimilar organisational settings.

The coding led to the second type of analysis activity (data display; see Figure 1), in which the data from the two cases were compressed into matrices for an overview of the similarities and differences in terms of what happened in the two cases, including over time.

In the third analysis activity (conclusion drawing; see Figure 1), thick descriptions Yin 2014) were constructed in the form of two cases, or narratives, that describe the sayings, doings and relatings (Kemmis, McTaggart, and Nixon 2014) in line with Buick et al.’s (2016) identified strategies for developing a close and equal collaboration between practitioners and researchers. In addition, the three intersubjective dimensions where different enabling and constraining practice architectures appear (Kemmis et al. 2014) were used to frame the narratives. In the semantic dimension, practice architectures emerged through the language and concepts used at the meetings; in the physical dimension, they were visualised in the form of different doings; finally, in the social dimension, they emerged in the form of descriptions revealing how the participants related to one another and to artefacts inside and outside the collaboration. Quotations from the participating school leaders have been added to strengthen the validity of the narratives.

Results

Turning to the first research question, the results are reported in the form of two cases in which we describe and compare what happened in the collaboration between the researchers and the school leaders in School A and School B.

Like most Swedish schools, both schools had organisations for school improvement, built on collegial groups that are led by teachers having an assignment as first teacher and, in some cases, lectors (a teacher with a PhD degree). The first teacher assignment involved being a middle leader who, together with the principal, leads school improvement projects at the local school.

The case of school A

Some of the lectors and first teachers had heard about the comprehensive ULF project concerning teachers’ collegial planning and preparation. As there already existed professional relationships between them and some of the researchers in didactics in the ULF research team, this became the foundation for introducing the project into the school. Information about the project was spread among all the teachers, and it also reached one of the school managers. It was up to the teachers if they wanted to be part of the project; the only obligation was that they had to participate as part of a collegial team of teachers teaching the same subject. The process was fast, and teacher teams in mathematics, history and technology quickly entered the project.

The school manager considered the project interesting, as is could be used as part of the process of building a solid school leadership team. This was of interest due to the fact that six of the seven principals in the school leadership team were relatively newly employed. Our part in this ULF project was to support the school leadership team by providing action research as a foundation for leading and organising the teachers’ work. This was communicated at the first meeting with the school leadership team. Five of the principals and both school managers ended up participating in the project. Three of the principals had teachers from their organisations participating in the project.

In the initiation of our collaboration, it became clear to us that the principals were not familiar with the ULF project, nor did they know that some of the teachers had entered the project and that the work had already begun. Therefore, we, as researchers, gave a short presentation of the ULF project and described how it concerned teachers’ collegial planning and preparation in subject teams. During the presentation, many questions were put forward by the principals. After the presentation, the school leaders were asked how they perceived the ULF project. After five minutes of individual reflection, everyone got to say something about their understanding, and the main perceptions were positive.

There was a common understanding that the teachers’ part in the project could be used as a strategy to reduce teachers’ anxiety about the new school organisation. For us, this was a conclusion that needed some clarification. The school leaders explained how the school was in the middle of a merger process, where two upper secondary schools were becoming one. They also described the intention of breaking up the schools’ traditional norms, routines and structures. The aim with the new organisation was to develop a collaborative learning culture with a focus on students’ learning and well-being. In particular, teachers were being regrouped into cross-curricular teams around students instead of, as earlier, around subjects. In addition, teachers were expected to work collegially rather than individually. The school leaders explained that there was a tension connected to this process and that some teachers were resistant to working collegially in cross-curricular teams.

The collegial ambition of the new organisation was in line with the ULF project, but the focus on subjects was, in some way, contradictory, although this could be regarded as something positive by teachers who did not approve of cross-curricular teams.

The meetings were conducted regularly during every second month, and early in the process, it became obvious that there was a great frustration present in the project. In particular, some of the teachers criticised the responsible principals for not providing the necessary conditions for their participation, mentioning above all their lack of time. The principals’ response to this criticism was that it was difficult to organise the conditions for a project of which they were unaware.

At one of the first meetings we, as researchers, suggested that our collaboration should be based on action research, with regular meetings as a foundation. Another suggestion was that the collaboration should include leading actions and efforts aimed at enhancing local schools’ infrastructure and overall capacity to practise school improvement (Blossing et al. 2015; Hopkins 2017). In particular, it was suggested that these leading actions should be conducted by the school leaders between the meetings. Both suggestions received a positive response. We provided a short theoretical presentation about schools’ capacity to improve (Blossing et al. 2015). As a result of the presentation, the first leading action that was negotiated was an examination of how the school had acted in previous improvement projects. Within the theoretical framework, this action is called an examination of the school’s improvement history (Blossing et al. 2015).

During the process, the school leaders slowly became more active and engaged at the meetings, even if some of the school leaders were more involved than others. This resulted in the meetings gradually changing in character and began to involve dialogues, inquiries and joint analyses often connected to theoretical frameworks presented by the researchers. Intersubjectivity (Habermas 1996) could be discerned in the form of the theoretical concepts used by both the researchers and the school leaders.

In one meeting, there was a vivid dialogue about the researchers’ meta-analysis of the school leaders’ individual analyses of the school’s improvement history. Both the researchers and the school leaders raised questions as well as giving answers. The school leaders corrected us, as researchers, when we got something wrong concerning the school organisation, and they emphasised that we had used the wrong concepts in describing the school’s grouping system. This pattern was in line with what Buick et al. (2016) labelled as a sense of being equals. We, as researchers, and the school leaders were all experts, but in different areas.

Joint meta-analyses also revealed that there were two dissimilar school cultures existing in the school as a result of the two schools merging into one. Another pattern consisted of the differences in how the principals led their organisations. In response to this, one of the managers emphasised that ‘it is very important that we [as a leading team] share the same vision’ (School Manager 1, Meeting 3).

Halfway through, the ULF project was extended, and new teacher teams were initiated without involving the responsible principals. Over time, it became more and more obvious that there were several dilemmas connected to the connections (or lack of connections) between the different teams and collaboration in the comprehensive ULF project: ‘I think it’s embarrassing to ask, but now I’m asking it anyway. The project with the teachers and this that we are in, is it the same?’ (Principal 1, Meeting 3).

Then, the COVID-19 pandemic hit the country, which had a huge impact on what was happening in the school and in the school leaders’ daily work. With very short notice, the school leaders had to reorganise their school, and our meetings changed from physical to online. Time was devoted to talking about the pandemic and its consequences for the school leaders, teachers and students. Joint analyses (to be compared with Buick et al. 2016) were still able to be carried out, and conclusions were also drawn, even if genuine dialogue was constrained, mostly due to technical issues. One of the principals made the following reflection connected to engagement: ‘As I’m not the head of any of the teachers who are participating, I do not think I’m as involved between our meetings’ (Principal 2, Meeting 4). Regardless of whether or not the principals had teachers engaged in the project, the principals’ engagement with teachers’ actual work emerged as small, and some of the teachers’ frustration was still present after months of collaboration. According to one principal, some teachers had left the project, despite time being scheduled for their participation.

As the teachers’ preconditions for participating in the ULF project were an issue often discussed, one meeting was dedicated to a joint reading of an article (Nordgren et al. 2020) that reported on a survey on how teachers and school leaders experience the conditions for teachers’ planning and preparation. The survey showed that teachers found the necessary preconditions absent, while principals perceived that teachers had time within their ordinary working hours. The reading of the article started an intense discussion in which the school leaders tried to find explanations for these contrasting opinions. One explanation was found in the teachers’ time agreements. Another explanation was connected to traditions and culture: ‘Upper secondary schools are very individualistic organisations; teachers have their lessons, their subjects, they work alone in the classroom’ (Principal 2, Meeting 5). This discussion turned into a reflection on how difficult it was to lead and accomplish pervasive changes and improvements in schools, with the school leaders referring both to their own experiences and to school improvement research (Blossing et al. 2015).

As a strategy to reach co-ownership and collaboration, we, as researchers, were keen to acknowledge practical and theoretical knowledge as equal. We often reinforced what the school leaders said, and by asking questions about the school’s improvement capacity, we wanted to inspire the school leaders to systematically analyse their organisations and leadership. This was appreciated by the principals: ‘It can feel quite good to discuss in this way because we do not have time in our ordinary daily work. It is very rare to have one-and-a-half hours when we put aside everything else and just talk about a special project’ (Principal 3, Meeting 5).

By the end of the process, there had been positive changes in the collaboration between ourselves and the school leaders, which also had positive effects on the teachers’ conditions for participating in the ULF project. In the closing meetings, the school leaders were more informed about their teachers’ work in the project and could report that the teachers found the ULF project very constructive. In our collaboration, we reached an unforced consensus (Habermas 1996) that organised time for teachers’ planning and preparation is a necessary precondition: ‘There is a need to develop a methodology for how subjects should work in the new school. The teachers in the project are in the process of getting a little closer to developing that methodology’ (Manager 1, Meeting 5).

The case of school B

The first meeting with the principal in School B was quite different from the meeting with the school leaders in School A. In this school, the principal was the first to hear about the project and the one that announced the project to the school. At our first meeting, the principal organised the teachers who wanted to participate into subject teams. Some of the participating teachers had assignments as first teachers, so they were used to leading school improvement projects.

Unlike the principals in School A, this principal was well informed and played an active part right from the start. The principal organised the teachers into subject groups, negotiated and distributed time for the teachers’ participation in the project, and reached an agreement on how to compensate the teachers in case of extra time used.

Even if the frustration in School A was not present in School B, the principal was a little confused about the teachers’ participation. S/he found the ULF project quite different from other school improvement projects, as these always included all the teachers and were mandatory. Like the school leaders in School A, s/he was unsure about the work that should be done by the teachers: ‘I’ve organised the groups and made sure that there would be time scheduled, but since then, I have not been involved any further. I know nothing about the actual work’ (Meeting 2).

In the same way as the school leaders in School A, the principal found it important that the researcher should know about the school’s organisation. At the first meeting, s/he clarified how the teachers were organised in teams around students, but unlike in School A, there was a long tradition of this. S/he also described the two-teacher-working model, which meant that the teachers always taught in pairs. According to the principal, there were many benefits of two teachers being together in the same classroom; however, the principal also pointed out that the model had some disadvantages. One was the difficulty finding time for teachers’ planning and preparation, which was why this project had caught the principal’s interest. When the principal described the daily work at the school, it appeared as systematic, with clear structures, visions and goals.

The principal was familiar with schools’ improvement capacity (Blossing et al. 2015; Hopkins 2017) as a theoretical framework, as this is used in the Swedish training program for school leaders, which s/he participated in during the collaboration. It turned out that an examination of the school’s improvement history was a task within the program, so the principal was able to conduct this both as part of the training program and as a leading action in our collaboration. The result was a well-performed examination, showing that the school had a routine whereby new projects were often tried out on a smaller scale before being evaluated and eventually included in the whole school. The examination also emphasised that every new project was frequently assessed in relation to students’ learning conditions.

The principal expressed that s/he appreciated collaborating with a researcher, as this had resulted in some self-reflections: ‘When you ask questions to clarify, I must reflect on why we do this … and in what way we do it … this broadens my understanding and makes me think … . It helps me to reflect, so we can move forward’ (Meeting 3). A topic that was discussed during the meetings was the school leaders’ function in school improvement, with the principal emphasising that if the school leaders were not engaged, ‘the work will be isolated, and there will be no shared vision. I feel that there is also a risk that it will not be equal for the students’ (Meeting 5).

In the same way as in School A, the COVID-19 pandemic meant big changes in the daily life of the school, and it also affected the principal’s engagement with this project: ‘I would say it is a bit different now; we must prioritise keeping the school running, so this project has fallen a bit into the background’ (Meeting 3). Another similarity was the existing gap between the participating groups in the comprehensive ULF project: ‘I have not really known how … about which part of this I should be involved in and what part the researchers do with the teachers … . I would have liked to have had some form of feedback, as we usually do in different school improvement projects’ (Meeting 3).

The last meeting was mostly devoted to a reflection on the collaboration process. One conclusion was that it would have been positive if the different parts of the ULF project had met as a strategy to close the gap. Another conclusion was that the principal saw the researchers as the project leaders and as owners of the project: ‘I took a smaller role in this project because I had not fully understood the aim … . I did not exactly know what and how things should be done … I feel that I had a much smaller leadership role in this improvement work than I have in all other improvement work’ (Meeting 6).

In the next section, we use practice architectures (Kemmis et al. 2014) to analyse what shaped the collaboration between the researchers and the school leaders and to explore why it was so difficult to create a co-ownership of the project.

Practice architecture that enabled and constrained the collaboration

Our findings across the two cases reveal that particular social–political arrangements are key to enabling and constraining the kind of collaboration that is advocated in this study. One such arrangement is connected to the need for shared agreements regarding the purpose of the research, while another is the importance of joint accountability (Buick et al. 2016). Neither were fully achieved in the current collaboration between ourselves and the school leaders. But there were improvements in the social dimension. These and other indicators will be addressed in the following section.

The initiation of the ULF project was an arrangement with impact on our collaboration. The different initiation approaches influenced what happened at the meetings. In School A, time was spent clarifying and trying to understand conditions in the ULF project and discussing teachers’ reactions. This was not at all the case in School B, where the meeting could instead immediately focus on how schools’ capacity to improve (Blossing et al. 2015) could be used as a theoretical frame in leading the ULF project.

The initiation was also crucial from a school leading perspective, emerging in the dimension of power (Kemmis et al. 2014). While the principal in School B had control over the teachers’ participation, the principals in School A were not involved at all. Instead, some of the teachers took the lead: ‘We have talked about how we were second on the ball here. So, in that way, the initiation may not have been the very best’ (Principal 3, School A, Meeting 6).

One explanation for the differences in initiation at the schools might be that there were existing relations between teachers and didactics researchers in School A, but not in school B. Another explanation is to be found in the analyses of the local schools’ improvement histories. In School B, there was a long tradition of the principals leading school improvement projects and being involved and engaged in teacher teaching. The tradition in School A was quite different, with a strong individual tradition of autonomous teachers, who were used to taking the lead in different projects without engaging the principals.

The results show that material–economic arrangements, in the form of organisational structures, affected the collaboration in both positive and negative ways. School A was a significantly larger organisation, and the collaboration involved a school leadership team that was busy uniting two schools. The school managers were heads of the principals and also more engaged in the ULF project than the principals. This social–political arrangement shaped how we related to one another in the meetings. In School B, interactions between researchers and school leaders were easier, with one principal leading a well-known organisation with systematic routines and clear norms.

An arrangement that emerged as a challenge was the connections, or lack of connections, between the different parts participating in the ULF project. This emerged in the form of uncertainty among the principals regarding how our collaboration (between school leaders and researchers) was linked to the teachers’ work and collaboration with the researchers in didactics. Another dilemma that applied to both schools was that the ULF project entailed an organisational challenge, as it did not fit into any of the schools’ formal organisational settings or infrastructures (Blossing et al. 2015). When the school leaders visualised their organisation in the form of a drawing, it was impossible to include the organisation of the participating teachers’ collegial planning and preparation in the formal settings. Adjustments in the internal infrastructures were more complex in School A since the participating teachers belonged to different principals. Another constraining arrangement was that the project only involved some of the teachers, with the consequence that existing subject conferences could not be used in the project. Therefore, solutions were required that went beyond the formal infrastructure: ‘Time needed to be set aside outside working hours so that they would have time to carry out planning for what is included in the project; otherwise, if everyone had participated, it would have been possible to use subject conferences to a large extent’ (Principal, School B, Meeting 6).

Voluntariness is an arrangement that is linked to the ethical perspective in research, which requires that participants participate of their own free will. This constitutes an important factor that needs to be emphasised in collaborations between researchers and practitioners. It is an approach that differs from ordinary school improvement projects. From a school leading perspective, this makes research projects complicated to lead and implement in formal organisational settings. As one principal said, ‘things happen without my knowledge’ (Principal 3, School A, Meeting 3).

Time, in the form of material–economic arrangements, enabled the development of dialogues within the collaboration. Dialogue was used as a strategy for achieving intersubjectivity (Habermas 1996). By enabling the development of intersubjectivity, as a cultural-discursive arrangement, continuous dialogues allowed meaningful and professional conclusions to be drawn. The fact that some school leaders admitted that they had problems getting the big picture regarding the ULF project can be seen as a sign of trust, respect and a close interpersonal relationship, encouraging the participants to speak freely (Buick et al. 2016).

A social–political arrangement that emerged as facilitating an equivalent collaboration was the composition of the group (Buick et al. 2016). In our collaboration, the school leaders clearly took the role of experts regarding their school organisations, while we were regarded as experts in the theoretical concepts and methods that were used in the process.

Discussion

The aim of this study was to explore collaboration between researchers and school leaders. The results show how this kind of collaboration does not just occur. This has already been established in previous studies (Buick et al. 2016; Forssten Seiser 2019; Levinsson 2017), but often with a focus on the practitioners (Olin and Pörn 2021). In this study, the focus was on collaboration as a practice, which draws the attention away from the participants. The focus was on what happened, as well as on the enabling and constraining arrangements enabling a deeper understanding of why things happened. This is a perspective that avoids the identification of scapegoats.

A limitation of the study was the non-existence of a holistic perspective, which emerged as a lack of intersubjectivity and common understanding within the ULF project, not least among the researchers in the multidisciplinary team. Closer collaboration in the researcher team could have prevented some of the uncertainties that were expressed by the school leaders during the whole process. In addition, COVID-19 turned out to be a challenge that no one could have foreseen.

By exploring this collaboration as a merging of two established practices within the educational complex (Mahon et al. 2017), the idea emerged of a third practice (inspired by Kieser and Leiner 2011) populated by researchers and school leaders who are engaged in a close and equivalent collaboration. Adopting an ecological perspective provides us with the concept of niche (Kemmis et al. 2014) to describe the creation of a healthy environment, which can be seen as a critical aspect to consider when establishing this kind of collaboration.

Aim for co-ownership and an equivalent collaboration

The results show how methods that support dialogue, intersubjectivity and unforced consensus enable the progress of an equivalent collaboration (see also Edwards-Groves, Grootenboer, and Rönnerman 2016; Francisco, Forssten Seiser, and Grice 2021). In addition, the results reveal how a lack of shared agreements regarding the purpose of a collaboration constrains, or even prevents, co-ownership.

Co-ownership and an equivalent collaboration involve trustful and respectful relations, empowering individuals in their professional roles. The development of a shared language and a common understanding builds interpersonal trust, which instils a sense of belonging. It also addresses the hidden power dynamics that are traditionally found in the relations between practitioners and researchers. This component can be silent, but it is important, as theoretical knowledge has the tendency to be regarded as more true and more significant than practical knowledge. But this phenomenon can be counteracted by everyone being regarded as an expert and an equal in the collaboration.

Pay attention to organisational infrastructures

The results show how a lack of co-ownership risks research projects not being embedded in local school organisations. The work that was conducted by the teachers in the ULF project was organised outside formal settings in both School A and School B. The consequence of this was the immanent risk that the efforts from the teachers’ work would not be shared or made available to others in the schools. Therefore, we strongly recommend that infrastructures (Blossing et al. 2015; Hopkins 2017) that enable collaboration between researcher and school practitioners are in place early in the process. This applies to both local schools and universities. Essential academic infrastructures include mechanisms to bridge conceivable gaps between researchers from different research traditions, while school infrastructures are required for research projects to become part of the everyday work.

We promote a holistic infrastructure that facilitates gatherings between all participants in a current research project. We encourage approaches such as collectively raising research questions and carrying out joint analyses enabling a shared understanding of the research and the collaboration.

Enhance school leaders’ legitimacy to lead research-based school improvements

This finding is in line with earlier research on how school leaders is a powerful factor in school improvement (Jarl, Blossing, and Andersson 2017; Robinson, Lloyd, and Rowe 2008) showing how school leaders are the best suited for leading school improvements at the local school level. They are also the experts on their school’s infrastructure. The study reveals how different approaches for introducing research projects have a large impact on school leaders’ involvement and how the voluntary nature of teachers’ participation constrains a whole-school approach. Therefore, we argue that school leaders’ participation in the introduction is essential for enabling sustainable and pervasive research-based school improvement that includes the settings for all teachers’ professional and collegial work (Timperley 2011).

Finally, this study shows how a collaboration, designed as action research, utilises a critical approach that supports both school leaders and researchers in understanding how practices and individuals are shaped (and shape) based on habits, adaptions, ideologies and traditions (Kemmis et al. 2014; Mahon et al. 2017). To enhance the scientific foundation of school practices, we advocate the need for practices that are facilitated by organisational infrastructures and inhabited by school practitioners and school researchers engaged in equivalent collaborations built on shared responsibility and co-ownership.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Karlstads universitet and Vetenskapsrådet [Dnr 2021-04766].