Enhancing system thinking - a superintendent and three principals reflecting with a critical friend

ABSTRACT

There is a growing body of action research on school improvement and development. Many of these studies have found the use of critical friendship valuable in relation to teachers’ or principals’ practice. Yet, little attention has been paid to critical friendship at the school district level. This study explores how a superintendent and three principals changed their understandings through interaction with a critical friend. It draws on reflective conversations, written reflections and observations of meetings between the participants, and is framed within practical action research aimed at changing the participants’ understandings. We found that through a trust-building process the critical friend acted as a catalyst for systematic reflections, helping leaders to set aside time and space to reflect. This has contributed to new ways of talking and thinking and has enhanced the leaders’ system thinking.

KEYWORDS:

• Critical friend
• dialogue meetings
• reflections
• school districts
• superintendent
• system thinking

Introduction

A growing body of research utilises action research as a tool to improve and develop schools. Critical friendship, a form of intervention found within action research (Kember et al. 1997), is also gaining popularity. However, most of the studies on both action research and critical friendship are on school-level, that is, they are concerned with the teachers’ practice (e.g. Costa and Kallick 1993; Kember et al. 1997; Wennergren 2016), and fewer studies are engaged with school leadership issues, involving principals (e.g. Swaffield 2008, 2015; Gurr and Huerta 2013). We have not found any research on critical friendship involving the school district, the municipality or the local education authority.

Nevertheless, there seems to be a broad consensus that school districts can play a significant role in supporting local schools in their change and development endeavours (Leithwood, Anderson, and Louis 2011). Several empirical studies encourage school district superintendents to collaborate with principals on developing a shared understanding of goals, through dialogue in designed professional forums (Lee, Louis, and Anderson 2012; Louis 2015; Paulsen and Høyer 2016; Henriksen 2018).

In this study, we follow a Norwegian municipality – which constitutes a school district and a local education authority – that supports schools through dialogue meetings with the purpose of improving pupil results. The dialogue meetings can be understood as a designed community of practice (Brown and Duguid 1991, 2001; Wenger 1998) set up at the purpose of supporting practice-learning in the form of clarification of intention, mutual adaptation, integration of goals, sharing of knowledge and setting direction for practice among actors working in disperse locations (Printy 2007; Hjertø, Paulsen, and Thiveräinen 2014).

We investigate how a critical friend (Swaffield 2008) aids the superintendent and three principals with reflections on dialogue meetings and asks critical questions to their practices. In general, a superintendent in Norway is a) the principal’s leader and is responsible for primary education within the municipality, b) a member of the municipal administration’s leadership team, and c) subordinated to a political board (Paulsen et al. 2014a). The key question for the study is: How can a critical friend enhance educational leaders’ understandings on how to change and improve practice? Our findings suggest that through a trust-building process the critical friend has acted as a catalyst for systematic reflections, helping leaders to set time and space apart for reflection. This has contributed to new ways of talking and thinking, and it has enhanced the leaders’ system thinking.

Theory

Practical action research and critical friendship

In action research, the researcher intervenes with the intention of improving the practice under study. The intervention constitutes a spiral of action cycles with planning, acting, observing and reflecting. In reflections – the hallmark of action research – the participants systematically reflect upon their own practices to gain an increased understanding of how to change and improve practice (Carr and Kemmis 1986).

Within the various forms of action research, we will briefly pinpoint two, a) practical action research, which uses external facilitators/researchers who cooperate with practitioners in their change process, and b) critical action research, in which the practitioners themselves take charge of their own change process (Carr and Kemmis 1986, 203–204; Kemmis 2009, 470). We draw on practical action research with the first author as an external facilitator, inciting reflections from the participants, and with participants not involved in any part of the research process. The participants are action learners and not action researchers (Tiller 1999) as action learning happens among the participants in practical action research.

Action research deals with change. In theorising change, two main conceptualisations dominate. The first perspective maintains that change is linear, with clear cause-and-effect connections, or a complex struggle based on competing interests; the second viewpoint indicates that change is about how we understand the interconnections among people or about capturing visible improvements (Gunter et al. 2007). Yet, another approach to change is connected to the process models of organisational innovative learning (Kolb 1984; Engeström 1987; Nonaka and Takeuchi 1995). Kolb’s spiral of actions research and Engeström’s concept of expansive learning offers additional perspectives to that of Gunter et al. (2007). They are both based on exploring what takes place during a collective activity (e.g. a school development project) and the expansive learning is highlighting evolving contradictions. These approaches reflect the importance of focusing not only on what leaders can do, but also on what they apparently cannot and why they do as they do (Thomson, Hall, and Jones 2013), as for example documented in a Norwegian study investigating tensions in school development (Aas 2017a).

Kemmis argues that action research aims at not only changing practice but also changing the practitioners’ understanding and the conditions they practice in,

Transforming our practices means transforming what we do; transforming our understandings means transforming what we think and say; and transforming the conditions of practice means transforming the ways we relate to others and to things and circumstances around us. I will speak about these three things as “sayings”, “doings” and “relatings”. (2009, 463)

Thus, action research aims at changing the practitioners’ sayings, doings and relatings, that is, their understanding, their actions and how they connect and interact with their surroundings. Moreover, the sayings, doings and relatings are tied together and always in a change process, interconnected and dependent on each other. Inspired by a Norwegian action research project, where the researchers identified how principals transformed their leadership practice through their understandings, sayings and doings (Aas, Vennebo, and Halvorsen 2019), we will use Kemmis’ terms of sayings, doings, and relatings in our analysis in this study. Kemmis connects the terms to critical action research, but we maintain that it is also applicable to practical action research, especially when open reflections are as central as it is in this study.

The ability to identify how and why participants think the way they do seems important in action research. However, leaders seldom use reflections systematically as part of practice, but something additional to practice if there is time (Schön 1983; Seibert and Daudelin 1999; Robertson 2013). With reflection, it is here not meant Schön’s (1983) reflection-in-action, but rather reflection on action, that is, reflection on past experiences. Furthermore, just as organisations in general have difficulties dealing with double-loop learning, many leaders tend towards single-loop and not double-loop learning. The latter refers to questioning underlying norms and goals in an organisation and are essential for learning and development (Argyris and Schön 1978).

In this study, we understand reflection as a critical investigation that provide learning and development for the participants. Goleman (2006) argues that metacognitive skills can increase leaders’ capacity as problem-solvers, decision-makers and critical thinkers, enabling them to regulate their emotions, handle complexity and cope with conflicts. Additionally, Aas (2017b) and Robertson (2013) claim that metacognitive leaders perceive their leadership as opportunities for learning, developing their capacities in the process.

Since reflection is a key concept in action research, critical friendship can be used (see e.g. Kember et al. 1997). In fact, in describing practical action research, Carr and Kemmis state that ‘the facilitator’s role is Socratic: to provide a sounding-board against which practitioners may try out ideas and learn more about the reasons for their own action’ (1986, 203). This is similar to a critical friend who ‘assists through questioning, reflecting back and providing another viewpoint, prompting honest reflection and reappraisal, a seeing anew that may be challenging and uncomfortable, yet enhancing’ (Swaffield 2008, 323). The critical friend also acts as a shoulder to lean on. Support is more important than critique, but trust is also essential to achieve open and honest communication and psychological safety for thinking aloud (Paulsen and Henriksen 2017, 80).

The role of the critical friend is close to the role of a leadership coach. In studies of group coaching in professional development of principals, researchers have demonstrated how the group coach can contribute to principals’ learning by critical questions and reflections that can lead to new leadership practices (Fluckiger, Lovett, and Dempster 2014; Aas 2017b). Successful reflections are linked to leaders’ practice in schools, involving both personal and systemic perspectives and being part of a social learning process. Finally, a climate of trust and support among the participants open up for sharing experiences on an individual level (Brandmo et al. 2019). In a study of an underperforming school district, the researchers found that a close relation between appreciative inquiry and trust led to new initiatives in student achievement, respect, and community pride and involvement. In sum, these studies underline the importance for the critical friend to combine the inquiry role with a climate of trust (Tschannen-Moran and Tschannen-Moran 2011).

Dialogue meetings within school districts

Large-scale research from the US and Canadian context has shown a positive link between trust-based collaboration involving school districts and principals – on the formation and sustainability of professional communities in schools – which again relates positively to more focused instruction and better student achievements (Leithwood, Anderson, and Louis 2011). In this perspective, dialogue meetings are assumed to work as a vehicle:

It looks like the dialogue model with its focus on creating spaces for communication, cooperation between all participating groups, and a sense of ownership of all municipal goals among the teachers, is a far more effective model for influencing teachers’ practice. (Berg 2015, 382, our translation)

In Norway, the dialogue model, also called management dialogue, consists of dialogue meetings between the levels in a school district (e.g. local politicians, the municipal executive director, the superintendent, principals, team leaders, teachers, parents, pupils) (Jøsendal and Langfjæran. 2009, 151–152). It varies from one district to the other whether they have dialogue meetings and how this plays out in practice. However, the emphasis is on broad involvement and dialogue between equally important participants on how to develop and improve schools. Since Norwegian municipalities have received more responsibility for primary education in the last two decades (Hindberg 2009, 165), dialogue meetings have gained popularity. An R&D report (Jøsendal and Langfjæran. 2009) recommends municipalities to become more involved in improving and developing schools through ‘a management dialogue about the school’s content, results and methods. This dialogue has the potential to become the municipality’s and the schools’ strongest impetus for quality improvement’ (Jøsendal and Langfjæran. 2009, 8, our translation).

Research in Norway supports this view and maintains that both superintendents and principals appreciate cooperation through dialogue meetings; the superintendent acquires a better understanding of how the school works, and the principals receive support, feedback and advice (Engeland, Langfeldt, and Roald 2008, 191). Dialogue meetings have been seen as a forum for sensemaking where joint reflections build shared understandings and more democratic processes (Henriksen 2018). However, some studies (Roald and Øydvin. 2009, 139; Møller and Ottesen 2012, 182) have found that not all dialogue meetings help developing schools and many schools find themselves without enough support from their municipality. Some of the reasons might be that the municipalities have little administrative resources (Møller and Ottesen 2012, 182) and that superintendents focus more on economic and administrative tasks (Paulsen and Høyer 2016, 123). However, it is difficult to cooperate across levels in an organisation (see e.g. Pors 2012, 30), especially in a complex organisational hierarchy such as a municipality (Paulsen et al. 2014b), and particularly within education, where schools have traditionally been rather self-governing entities with considerably autonomous teachers.

Methodology

A small-medium municipality in Norway, with a little less than ten primary and lower secondary schools, took part in two national projects to strengthen the local education system, the Tutor Corps (Veilederkorpset) and the School and Municipality Development Project (SKUP) (Ministry of Education and Research 2013, 162). In these projects, the municipality received external help to build up a management dialogue, starting with dialogue meetings between the municipal administration and school leadership teams.

Through a project based on practical action research (Carr and Kemmis 1986), the first author participated in the first 18 months of implementing these dialogue meetings, acting as a critical friend (Swaffield 2008). The project was initiated by the superintendent’s expressed interest in involving action research in school development projects, hoping that someone could pose critical questions and bring in new perspectives. Just before the project started, the first author worked in the municipality under study. Thus, he was familiar with the participants and the setting, and was already trusted by the participants. When the project started, he was no longer an employee in the municipality and clearly stated his independence from the superintendent. If we mix Herr and Anderson’s (2005) action research positions, the role of the first author could be an insider going outside to collaborate with insiders.

The intervention started in the autumn of 2013 and followed the spiral of the action research cycle: planning, acting, observing and reflecting. First, the researcher was an external facilitator involved in the planning of the next dialogue meetings with the superintendent. Second, a dialogue meeting was held at each school with the researcher as an observer audio recording it all. Third, four key participants reflected on the meetings with the critical friend. Each half-year the cycle was repeated making it three cycles within 18 months, from the autumn of 2013 to the autumn of 2014 (see Appendix).

The reflections consisted of three parts. First, each participant wrote a reflection note where they reflected on the planning and function of the dialogue meeting, including strengths and challenges. Second, each participant had a conversation with the critical friend, with their own reflection note as the starting point. Third, each participant wrote a second reflection note addressing four questions concerning both the dialogue meetings and the reflections: a) what was good, b) why, c) what can be better, and d) how. The reflections were supposed to support and strengthen the dialogue meetings as well as giving crucial data for research. In addition to reflection notes and reflective conversations, collected data consists of observations, audio recordings and minutes from the dialogue meetings. A last reflective conversation was held in the spring of 2015, half a year after the third cycle of dialogue meetings. In addition to collecting more data, these conversations were also used to evaluate the action research project.

Since the project was not all set up from the beginning, the principals only wrote one reflection note each after the first dialogue meeting and only the superintendent had a reflective conversation with the critical friend (see Appendix). The superintendent was also too busy to write all the reflection notes. In hindsight, we could have improved our preparation and offered busy participants the possibility to audio record reflections, but the amount of data was still more than enough.

In the analysis (done with NVivo), the transcripts were categorised by induction. Since we sought to explore the participants’ thoughts on the reflections with the critical friend, the following categories were chosen: reflections, critical friendship, about dialogue meetings, and the researcher’s role. With these categories we wanted to find out whether the participants developed any new perspectives, changed their understanding, and whether the critical friend made any difference. We have tried to secure validity by disclosing both positive and negative results, by making all component parts transparent, so the study is open for inspection, but also, through triangulation – across different methods of data collection: observations, conversations, and written notes – and member checking – by letting the participants read and comment on the study (Creswell 2012).

The first author wrote a research log to keep track of ethical decisions and choices made throughout the process (Herr and Anderson 2005, 77). Any intervention in people’s lives should be taken seriously, respecting the privacy and anonymity of the participants (Creswell 2012, 88–90). All participants have volunteered and given written consents for the research. The Norwegian Centre for Research Data has approved the treatment of data in the project. It is hard to give anonymity to leaders in central roles, but since the superintendent was open for naming the municipality with its real name and giving up her anonymity, we could mention explicitly in publications that the first author worked in this municipality. To ensure anonymity to the principals, we have not disclosed names of the municipality or the schools and we have given fictive names and gender to the participants. Being invited by the superintendent, we had to be especially aware of ethical power issues (Herr and Anderson 2005, 36), maintaining an open collaboration with the principals, making sure we did not run errands for the superintendent. We argue that action research, with its stress on intervention, is ethically defensible since it tries to give something back to the participants that can guide them in their further practice.

Findings

Kemmis’ (2009) sayings, doings and relatings have been used to sort the findings and to analyse their significance in relation to action research. Sayings are easily traced in the findings since the collected data mainly consists of dialogues and since sensemaking is a pivotal activity in dialogue meetings (Henriksen 2018). Doings and relatings are harder to find, except for the fact that the dialogue meetings in themselves represent changes in doings and relatings as the participants’ practice have changed and they have brought new people together. Since we only have observations from meetings and not from the participants’ daily practice, we have taken the participants’ self-reported utterances of any change at face value. To acquire a deeper understanding of the participants’ experiences and reflections, the findings are presented for each of the four participants, one at a time in a narrative form (Toledano and Anderson 2017).

Principal Arvid – complexities of change

Throughout the process, Arvid was positive to the dialogue meetings and wrote, ‘I believe the dialogue meeting can be a good way to evaluate the school’ (note 1). However, the school had some challenges with following up the meetings. After the third dialogue meeting, when asked about how they followed up, Arvid said, ‘in which way we will continue working with the dialogue meeting, that’s where we are’ (conv 3). There was an intention to follow up the dialogue meetings, but after three dialogue meetings, the school still struggled with how to do it. There was a lack of change in practice and hence their doings remained the same.

Arvid was quite eager to include research in the process and wrote that reflections with a critical friend were important: ‘I have nobody to reflect with, no colleagues, only in the pauses at the principal meetings. There’s not enough time’ (note 2). Moreover, he wrote that it was good to ‘analyse the dialogue meetings with a critical friend [… because it gives] other perspectives [… and] a researcher’s eye on the challenges’ (note3). This suggests that the reflections changed Arvid’s understanding and thereby changed his sayings.

Arvid explicitly commended the reflective conversations. In a note, he wrote:

I can see that we now have become more familiar with each other and that it’s easier to enter the ‘critical friend phase’. Feel like we have now entered a new phase: Now we don’t feel each other out anymore, it’s a higher form of reflection, maybe not meta-reflection yet. (note 3)

Here, Arvid suggests that trust increased as the relationship to the critical friend grew, leading to more critical questions. Moreover, Arvid wanted to have ‘reflective meetings in the class, in the pupil council, in the parental council’ (conv 3). He changed his sayings as he changed his understanding of the reflections, finding them so valuable that he wanted to extend the practice with reflective conversations to other areas within the school.

A leadership lesson from Arvid’s experiences might be that change processes are complex. He says the dialogue meetings and the reflections are good, but there is not much change of practice in the school. Support from a critical friend makes it even more complicated, though probably also more sustainable, adding reflections that help him to see the larger picture instead of focusing too much on practicalities.

Principal Bodil – system leadership

Bodil was, in general, optimistic and said she felt ownership to the dialogue meetings. She thought the conversations were good and practice-oriented, ‘Very good and warm dialogue. Concrete and close. We managed to prioritise the right things’ (note 3). Moreover, she showed metacognitive skills when she wrote, ‘It’s good to adjust my thinking in relation to the critical friend’ (note 3). One such instance of adjustment was the realisation of her role as a system leader, connecting the school’s goals to the municipality’s goals. Bodil said, ‘it’s more like the superintendent’s main intention with the dialogue meetings, that is, to tie it all together, from the top to the bottom’ (conv 3). Bodil changed her understanding of the local education system, developing another way of talking about the school’s goals to her own staff, making the dialogue meetings a natural part of their school’s daily practice.

I feel ownership of the dialogue meeting, I feel it’s very important. And I feel that it works really well, like when I think back, I often mention it in the plenary meetings, that we reached this and that at the dialogue meeting, and Solveig expressed this in the dialogue meeting, I say as a natural part of everyday life. And then … in order for it to work in our school, it’s clear that the most important factor is that it’s rooted in the principal, and if the principal, in one way or the other, does not talk about it with warmth, then it will certainly have zero effect (conv 3).

Bodil referred to the dialogue meetings in plenary talks with the school’s staff, using it as a steering mechanism, expecting the meetings to guide team leaders and teachers, changing their practice and hence their doings. This is also a change of sayings as Bodil now showed an understanding both of the connections from the municipality to the school’s practice and of the importance to mention this explicitly to her staff. Moreover, we can see the interconnectedness to the relatings, as this also transforms her relations to the superintendent, making the superintendent’s statements more influential.

Bodil wrote, ‘It’s also inspiring that this process can be observed externally through the research project. We expect that this can help us do the right things to create a positive development for our school’ (note 1). She appreciated the reflections with the critical friend and said, ‘When you have these follow-up talks like now, it’s more for research purposes, but nevertheless, it works very well that it’s followed up before and after’ (conv 3). Bodil had no critique or suggestions for improvement to the critical friend. When asked in the reflection note: ‘What can be improved?’ she wrote, ‘Nothing really. This works very well’ (note 3).

After the first dialogue meeting, Bodil wrote about the lack of coordination, stating, ‘There’s been created some confusion about the logistics of this process’ (note 1). However, already after the next dialogue meeting, her system thinking was strengthened as she now wrote, ‘The process is good’ (note 2). This change was also confirmed in the reflective conversations. Moreover, Bodil wondered whether there was a dialogue with the politicians, ‘we just have to trust that there is a dialogue at the political level. [… Solveig] could inform the principals more about this’ (conv 3). Bodil expected that the system was intact with involvement of the politicians in the management dialogue, aligning the different levels with dialogues going upwards and downwards.

A leadership lesson from Bodil’s experiences is the realisation of the strength in being a system leader. When she deals with expectations from above and wants to implement them in the local school practice, she uses the superintendent’s sayings as support in handling the teachers. Whereas, in dealing with the superintendent in the dialogue meetings, she has the support of her leadership team.

Principal Dag – leadership development

Dag was an eager proponent of the dialogue meetings. He declared commitment to the dialogue meeting as a way to improve the school: ‘I find these meetings rewarding. I really enjoy that the municipal management comes to us and that we have the opportunity to take charge of the presentation of the school’s results and challenges’ (note 3). After the second dialogue meeting, a project was initiated, as they received resources from the municipality for teaching some pupils considered troublemakers in a secluded space with special teachers. This project changed the teachers’ doings. Dag was very happy with the superintendent’s ability to take action and referring to this project said, ‘one of Solveig’s strongest sides is, as I see it, that she is present when she attends those meetings, and she really takes them seriously’ (conv 3).

Among the participants, Dag wrote the most in the reflection notes. He repeatedly underscored that he liked the outside perspective on what to improve in the school, ‘As I have said many times, I like that somebody outside the school looks at what we do. Reflecting together after dialogue meetings or other activities is smart’ (note 3). Dag found the conversations with the critical friend constructive and good: ‘The conversations are useful; they are a foundation for reflection and learning’ (note 2). Dag contended that the reflections could lead to learning and improvement of the school’s practice, suggesting his understanding increased and his sayings changed.

Dag talked about the difference between teachers and leaders: being spot-on and close to practice and taking a step back and seeing the whole picture. Some team leaders were ‘more teachers than leaders and the transition towards taking a step back and grasping the whole picture is a challenge’ (conv 3). He continually encouraged the team leaders to take a leader’s perspective, suggesting that system thinking was important.

After the first dialogue meeting, Dag was frustrated with the lack of coordination from the municipality: ‘I think there are too many plans, too many different concepts, and too many processes going on at the same time’ (note 1). However, already after the second dialogue meeting, his impression was quite the opposite. On the same question he now wrote, ‘This is neat and straightforward. No problems’ (note 2). Dag’s system thinking had changed and this was also confirmed in the conversation with the critical friend. When he was questioned about bringing the dialogue to the political level, he said, ‘it could have been good to ask the question, what happens on the way to the politicians. Clearly, that is certainly interesting’ (conv 3). Dag changed his sayings by raising his awareness of the system as a whole.

A leadership lesson from Dag’s experiences might be that reflections is a form of leadership development, as experience and reflections together induce learning. Thus, Dag is clearly touching upon metacognitive leadership, learning and developing through reflections on practice.

Superintendent Solveig – system thinking

Solveig thought the dialogue meetings were practice-oriented and continually improved during the process. ‘It’s a learning process. It’s a tool to be a learning organisation, practising being both questioned – somebody has to ask those questions – keeping appointments so that both have the responsibility to keep the process going and moving onwards’ (conv 1). She was pleased that the principals appreciated the meetings, ‘They said the meetings were important, and they said the meetings had an effect, both that people felt they were seen, that there was a dialogue, many felt they could say things, and that it was followed up’ (conv 3).

As the participants communicated their thoughts, their relationship grew, and trust and psychological safety for thinking aloud was strengthened. Still, it is an open question whether the dialogue meetings inspired changes in practice. Solveig pondered over this,

Did the dialogue meetings lead to changes in practice? It’s hard to say because it has been in a context with many other things, but I believe it has led to more pressure on … a pressure on practice, that is, it has led to a pedagogical platform and the questions to the schools have hopefully become clearer. I believe we now have more courage to ask for results. (conv 4)

Solveig argued that the dialogue meetings were on the way to being more critical and outcome-oriented, and a step towards a learning organisation. On the other hand, she also expressed uncertainty of what they achieved, ‘it’s not very much a reflective meeting either, it’s kind of … we come and we ask and we take it with us’ (conv 1). This impression lasted throughout the process. In the last reflective conversation, she said, ‘Would it be just as positive if you’d come in and taken a cup of coffee and asked, “How’s it going?”’ (conv 4). Still, Solveig underlined the importance to involve the schools, ‘If we could start it all over again and do it differently, I believe we should have given the schools the responsibility for the dialogue meetings from the beginning’ (conv 4). If the schools were in charge of planning, setting the agenda and leading the meetings, they would put more emphasis on what they found crucial to discuss and probably be more eager to follow up the meetings. In short, they would take more ownership of the dialogue meetings.

Solveig had few direct comments regarding the critical friend. However, she stressed that ‘it’s important to sum up, to see things in relation to each other. You attain other perspectives when there is a bit distance’ (note 2). This is exactly what the reflection notes and the reflective conversations gave room for. Solveig made several comments supporting both reflection and dialogue meetings. Among others, she believed it was most important to understand and tie it all together, ‘I believe it’s the most important word we have, to see the connections […] it’s not that we do this and then that, but everything is connected’ (conv 1). In this perspective, critical friendship and dialogue meetings are a way for the participants to make sense of the system and to change their sayings.

The dialogue meetings were only held between the superintendent and the school leadership teams. Yet the plan was to extend the dialogue to all units in the municipality. Solveig had a systemic perspective on the dialogue meetings and she wanted to include the politicians in due time. She said,

The politicians and the local education authority are certainly a part of the management dialogue; they too want influence. So, in the whole system, we need to take care of it so we will take an extra look at that, in three years the politicians ought to be included. (conv 1)

A leadership lesson from Solveig’s reflections might be the importance of system thinking, connecting the schools’ work on school development, securing school ownership of the processes, while at the same time strengthening the focus on results, critical thinking and learning.

Discussion

Our research focus has been to investigate how a critical friend can enhance the participants’ understandings on how to change and improve practice. The context has been reflections on dialogue meetings in a Norwegian municipality, and we draw on practical action research where the first author has been an external facilitator in the role of a critical friend. Two topics for discussion arise from the findings. First, we discuss the role of a critical friend as a catalyst for systematic reflections and briefly ask the question if the superintendent can have this role. Second, we discuss if and how dialogue meetings and critical friendship can induce reflection and learning.

The critical friend as a catalyst for systematic reflections

In practical action research, the critical friend supports through questions and reflections and is supposed to be an eye-opener (Swaffield 2008). This happened in this study when the critical friend triggered the participants to reflect on their experiences through dialogue and open questions. The comments from the principals show that they liked being seen by an outsider, particularly since they spent little time on reflection and had few with whom to reflect. They wanted a mirror from the outside that could inform their leadership practice in a new way, and several of them expressed interest in the research part of the study. Consequently, the design of a critical friend supporting the dialogue meetings gave space and time for reflections.

Leaders do not see reflections as something that is part of their daily leadership practice, but something that comes in addition. Critical friendship represents a possibility for reflections that can improve leaders’ metacognitive skills, which gives them opportunities for learning to develop their capacities in the process (Aas, Vennebo, and Halvorsen 2019), confirmed by the participants in this study.

Studies from the Nordic countries show that superintendents act as both critical friends, coaches and mentors to their principals (Moos, Kofod, and Brinkkjær. 2016, 47) and that this is highly appreciated by the principals (Johansson et al. 2016, 153). In light of this study, it is important to ask if there are any differences in having the superintendent as a critical friend or having an external critical friend. A distinction between the two is first of all defined by the different positions. When the superintendent has power to give instructions to principals as subordinates, the asymmetric relations between them can create tensions. With an external critical friend, however, there should be fewer power tensions. An argument for superintendents as critical friends is that they have the necessary context knowledge and are in position to lead the dialogue meetings and decide the topics for discussion. An argument against combining these roles is that there might be a stronger focus on management, whereas with an external critical friend the focus might be more on awareness and learning.

Reflection and learning

In this study, we have seen dialogue meetings as a space where professionals meet to discuss their own profession, reflecting over educational matters, with joint analysis and collaborative planning. This has much in common with reflective professional inquiry (Stoll et al. 2006) and the kind of professionalism that Hargreaves and Fullan write about, ‘Where collaborative improvements and decisions […] are pushed forward by grown-up, challenging conversations about effective and ineffective practice’ (2012, 128). The superintendent’s purpose of the dialogue meeting was to improve the municipality’s support to schools, by following up development projects, focusing on results, and asking critical questions, much in line with the abovementioned theories.

When the superintendent characterised the reflections in the meetings as more or less an exchange of information, there seems to be a gap between her expectations and the theories and experiences so far. However, the superintendent was pleased that the principals appreciated the meetings. She saw the importance to establish a good and inclusive climate in the dialogue meetings. Trust was fostered among the participants as their relationship grew and they gained an increased understanding of each other. In other words, the interconnections among the participants increased (Gunter et al. 2007). This suggests a change in their relatings, but also in their sayings. From research on group learning and group coaching, we know that groundwork in advance is important for building trust towards each other (Aas and Flückiger. 2016). Still, if sharing information is the only thing happening in the meeting, the idea of the dialogue meetings as a space for learning is not yet achieved.

The reflections were mostly about the practical sides of the dialogue (e.g. number of participants, agenda-setting, minutes writing, evaluation); and about the schools’ content (e.g. best practices, standardisation, pupil results, shared goals). Sharing views on these matters is a first step in transforming the participants’ understandings. The participants had a keen interest in changing practices for the better and they appreciated to reflect on past experiences. This is an interesting finding set up against the action-orientation of leaders in general (Schön 1983) and how they tend to view reflection as distinct from their practice (Robertson 2013). Even if leaders seldom reflect systematically, the participants found it valuable when they managed to set time apart for systematic reflections.

Still, when the participants are concerned with doings, and less oriented towards changing underlying norms and goals, they are in a single-loop learning circle. If new ways of thinking about and acting on issues are going to lead to enhanced learning and new practices, it requires double-loop learning. However, the reflections were not just comfortable and unchallenging talks and they might have induced some double-loop learning and changes in the participants’ sayings. The leadership lessons infer increased system thinking by several of the participants. Bodil has definitely changed her sayings on how the school’s goals and the municipality’s goals were connected, and how her leadership needs to balance between the school and the municipality. Solveig points to the connections between all the school projects in the municipality, increasing her understanding that they all must work together.

Conclusion

The critical friend has with systematic reflections contributed to new ways of talking and thinking for the key participants. The critical friend has been a catalyst for systematic reflections, helping leaders to set time and space apart for reflection. Much of the reflections were on practical questions with single-loop learning. Still, the leadership lessons imply instances of double-loop learning and metacognition through system leadership and system thinking. The critical friend was mostly concerned with supporting the participants, building trust, and reserving critique for later, when the relationship allowed for it, in line with MacBeath (2006).

Dialogue meetings as sensemaking processes can improve cooperation between superintendents and principals, and thereby help bridge the gap between municipalities and the schools (Henriksen 2018). A critical friend can help to deepen the dialogue meetings by enhancing the participants´ understanding and the way they relate to each other. We conclude that the learning aspect of the study was present, but it could increase with more double-loop learning and metacognition. This could be the next step, taking the project further by building on the foundation of trust and good relations.

Disclosure statement

No potential conflict of interest was reported by the authors.

Appendix

Data collected

Table

Data	Autumn 2013	Spring 2014	Autumn 2014	Spring 2015
Dialogue meetings	All schools	All schools a	All schools
Reflection notes	3 principals (note 1)	All participants (note 2)	3 principals (note 3)
Reflective conversations	Superintendent (conv 1)	All participants (conv 2) a	All participants (conv 3)	All participants (conv 4)
Reflection notes		All participants (note 2)	All participants (note 3)

aDue to a technical error, these audio files were corrupted and could not be transcribed.