How formal initiatives to improve teaching can lead to more significant informal conversations and increased sharing practice

: University teachers grow professionally from conversations about learning and teaching with colleagues. Significant informal conversations can be facilitated through formal activities initiated from the institutional side. This case-study shows how a formal institutional initiative to enhance educational quality has facilitated more significant informal conversations. Such conversations power constructive feedback to the organisation, improve the formal quality development work at the institutional level, and increase the use of collegial experiences across the institution. We identify the formal initiative “Collegial sharing sessions” as particularly efficient for fuelling significant informal conversations within and across departments.


INTRODUCTION
Academic teachers develop through professional training, experience, interaction, and conversations about teaching and learning with colleagues. Formal and informal opportunities can both promote professional development, but it is the social, active, continuous activities related to practice that are most effective (Eraut, 2004;Van Waes et al., 2016;Webster-Wright, 2009). One such activity is the informal conversations between colleagues that often focus on issues relevant to all conversational partners (Thomson, 2013).
Informal conversations characterised by privacy, mutual trust and intellectual intrigue can be defined as significant conversations (Roxå & Mårtensson, 2009). This is where teachers allow themselves to be influenced by others and these significant informal conversations are important for academic development (Roxå & Mårtensson, 2009). Studies have shown that significant conversations usually occur in small networks of colleagues in close proximity, within organisational boundaries (Pataraia et al., 2014;Roxå & Mårtensson, 2009). The most common conversational topics are managing teaching and disciplinary content, but conversations also provide personal support and ways to improve teaching and learning (Pataraia et al., 2014;Pyörälä et al., 2015;Thomson & Trigwell, 2018).
Significant informal conversations do not necessarily develop spontaneously, but they can be facilitated (Horn et al., 2020;Van Waes et al., 2018). van Lankveld et al. (2016) investigated informal teacher communities to identify processes supporting professional development among medical teachers. Their findings revealed that establishing informal teacher communities facilitated sharing stories of failures or difficult situations; and possible reactions to those situations. Having a space for dialogue enabled sharing of tacit expertise. Further, their study showed that informal teacher communities created a sense of belonging and identification of common challenges and solutions. An important finding was that, within these communities, teachers experienced that their role as teachers was valued and recognised as important. Informal teacher communities supported the professional development and strengthened teacher identity (van Lankveld et al., 2016).
Significant informal conversations are important (Roxå & Mårtensson, 2009;Thomson & Trigwell, 2018), but professional and educational development through informal channels may have restricted influence beyond the local participants. Katajavuori et al. (2019) studied the collaborative practices of academic teachers in life science. They found that teachers interacted informally and formally with peers for personal development, sharing good teaching practices, teaching together, producing educational artefacts, and developing education. All categories of interaction had both informal and formal examples, but informal practices were described with more enthusiasm than more formal practices. Although valuable practice developed in informal settings, it was unlikely to spread beyond the small, localised, discipline-specific networks and become part of the institutional practice -unless collaborative practices were fostered, bringing them into formal processes (Katajavuori et al., 2019). The existence of formal channels for sharing good practices at the institutional level, such as a local conference, can support teachers in overcoming the barrier of going public (Mårtensson et al., 2011). Therefore, institutions could benefit from organising activities that facilitate significant conversations and create links between local informal networks across the institution.
Identifying efficient actions to facilitate conversations is not clear-cut, and we need to acknowledge the influence of context and culture, as well as other institutional and disciplinary factors (Becher et al., 2001;Roxå & Mårtensson, 2015;Trowler, 2008). The strength and interaction among influencing factors vary depending on which aspects of culture are being examined. Results from a large (55 000 respondents) American national survey among academic staff showed that institutional culture strongly influenced collegiality within departments (Lee, 2007). Commitment to teaching was similarly affected by institutional and disciplinary culture (Lee, 2007). More case-studies informing this topic, particularly across disciplines, are therefore valuable.
This case-study investigated elements of the formal educational quality work at a small researchintensive academic institution, the University Centre in Svalbard (UNIS). We collected background data on a formal initiative to improve educational quality called the "UNIS Learning forum" (LF) over eight years (2013)(2014)(2015)(2016)(2017)(2018)(2019), examining programmes and participant feedback to identify actions and activities relevant for increased collegial sharing and initiation of significant informal conversations. In 2018, the additional activity "Collegial sharing sessions" was introduced as part of the LF. These collegial sharing sessions seemed to initiate significant informal conversations, and we investigated this further through a survey among participants in 2018 and 2019. The survey collected information about motivation and expectations of collegial sharing sessions before attendance, and gains and experiences after attendance. We further probed participant experience by interviewing three participants.
Our objectives were to identify critical actions leading to more significant informal conversations and sharing of practice at the local and institutional levels and discuss the transferability of identified aspects to a broader audience. In particular, we wanted to investigate whether introducing «Collegial sharing sessions», which are low-stake sharing activities in smaller groups, influenced both the number of significant informal conversations and the number of conversation partners. We hypothesized that "Collegial sharing sessions" would lead to more significant informal conversations among a wider network of colleagues.

Case-study
The University Centre in Svalbard is a small, research-intensive academic institution in Arctic Norway (UNIS). UNIS offers STEM courses with an Arctic perspective at bachelor-, master-, and PhD level. Students would typically spend one or two terms at UNIS as part of their degree programmes at other HE institutions. In 2019 UNIS had 68 full-or part-time professors/associate professors, 15 PhD students, and 47 technical-administrative staff involved in teaching, fieldwork and administrative support, and 743 individual students. UNIS is a partner in two Centre for Excellence in Education (bioCEED since 2014 and iEarth since 2020).
An annual "Learning Forum" (LF) is held at UNIS to increase attention on educational quality. Over time, the format of LF has changed from lectures and one-way communication to more collegial activities facilitating conversations and interaction across departments through workshops and collegial sharing sessions (box 1).

Insider research
Most of the authors have personal experience of the LF, either as participants or organisers. Consequently, the researchers are part of, and situated in, the context and culture studied, conducting insider research on our own organisation and practice (Brannick & Coghlan, 2007;Costley et al., 2010;Trowler, 2011). Rather than being an objective observer, an insider comes into their research with an informed perspective and social situatedness that gives unique access to the topic and subjects (Costley et al., 2010;Hanson, 2013). This enables a deep understanding, but comes with potential challenges to the validity and reliability of the research, such as prejudice, issues of internal organisation and closeness (Coghlan, 2007;Hanson, 2013). This study was inspired by the experience of several of the authors participating in and developing the LF. The questions emerged, and the results were interpreted through a reflexive process and negotiation where the authors also examined their own roles and perspectives (Agee, 2009). An outside view was provided by including one author from outside the context studied.

Content and evaluation of the formal initiative "Learning Forum"
To collect background data, we studied the LF programmes and written participant feedback from evaluation forms (2013-2019; box 1). We collected comments from open-ended questions related to the collegial sharing sessions, and summarised feedback from relevant closed-ended questions.

Teaching conversation survey
We developed a survey specifically for the collegial sharing sessions. The voluntary and anonymous online survey (Nettskjema) was distributed to the 64 participants in the 2018 and/or 2019 LF. Questions focused on how collegial sharing sessions have influenced conversations about teaching (appendix S1). The survey included 15 main items: two measuring level of attendance and involvement, eight items about motivation and outcome of attending (including sub-statements graded on the Likert scale), and five items with yes/no/NA options. Items with graded Likert scales were rescored as: Fully disagree = 1, Disagree = 2, Neutral = 3, Agree = 4, Fully agree = 5. The survey was distributed (28 MAY/2020) and open for 34 days.
Since 2013, UNIS has held an annual "Learning Forum" (LF) to increase attention on educational quality. It started as a voluntary four-hour lecture-based event that has developed into a mandatory event lasting 2 days, with talks, workshops and collegial sharing sessions (Fig. 1). The first LF had 12 participants and only scientific teaching staff were invited. In 2019, there were 60 participants, including scientific teaching staff, administrative and technical staff and a small group of students. The LF has become the largest direct investment UNIS makes in educational and professional development. After each LF, the event was evaluated and the committee archived participant feedback and minutes. Participants have been invited to give feedback in a formal evaluation survey (2015-2019). The organizing committee used feedback from participants and staff, as well as trends in educational quality development, while planning upcoming LFs.
In 2018, collegial sharing sessions were added to the LF activities. A collegial sharing session is a low-threshold, informal event. The person(s) leading the collegial sharing session (the facilitator) introduce the topic (15-30 min), followed by about 30 minutes of discussion and feedback from participants. The topic of the session is usually closely linked to the facilitators own teaching practice (e.g. field teaching, active learning, course planning, assessments methods). The facilitator can share topics such as successful teaching activities, problems or challenges, or an idea they have yet to try. The number of participants per session range from five to ten to enable good discussions. Collegial sharing sessions were organized as parallel sessions and participants had to prioritize between sessions.

Data analyses
Data from the Teaching conversation survey were summarised and plotted with the Likert function included in the HH v.3.1 package run in R v 4.0.3 (Heiberger, 2020;R Core Team, 2020). We analysed the ranked Likert scale (1-5) by standard descriptive statistics (median with 25 th and the 75th percentiles calculated in PAST 3.01) (Hammer et al., 2001).

Interviews
To probe survey findings further, semi-structured interviews were conducted with three mid-career academic teachers representing different departments, disciplines, and gender. Interviewees were selected among sharing session participants using convenience sampling. We asked informants if they had answered the teaching conversations survey prior to the interview. Informed consent was given by all informants. Interviews were conducted individually and included a mix of closed and open-ended questions, allowing follow-up questions (Adams, 2015). Examples of questions are: "What were your initial thoughts when you were invited to join a collegial sharing session for the first time? Who did you talk to about teaching before your first LF, and which form did these conversations have? When you talk about teaching, what do you talk about?" Two or three of the authors conducted the digital interviews, of which one led the interview, and the other(s) took notes. Interviewers wrote a summary based on the notes, including their reflections from the interview. Each informant read and approved the summary and was allowed to comment and revise to address potential mistakes and misunderstandings. Only minor revisions were made.
The three interviews were analysed following a deductive approach where we categorised the interview data based on themes identified in the background data and survey; i) how the LF had influenced conversations generally, ii) motivation and expectations to join collegial sharing sessions, iii) gains and experiences after attending collegial sharing sessions, and iv) perceived links between collegial sharing sessions and significant informal conversations. Two authors did the first categorisation, which was then discussed and refined in conversation with a third author. Special attention was given to information that could enrich the survey results. Selected quotes were used to illustrate and supplement the teaching conversation survey data. The informants have been given pseudonyms to protect their anonymity.

Data handling
The background data, the teaching conversation survey data, and data from the interviews were collected in line with regulations for data handling in Svalbard, Norway. All stored data have been anonymised and will be deleted or stored without personal information when this project is completed.

Summary of background data on the formal initiative "Learning Forum"
We noted increased engagement in the LFs over time (Fig. 1). The participant evaluation forms invited suggestions for new topics for the upcoming LF. In 2015, 4 of 20 (20%) responders suggested new topics for next year. Over subsequent years, the number of suggestions increased to 14 of 37 (39%) suggested topics in 2019 (LF background data; supplementary Table S1). The participants' opinions on appropriate frequency, duration, and commitment to the LF changed over time. In 2015, 11 of 20 respondents (55%) reported it was sufficient to hold the LF every second year, whereas in 2016, 15 of 19 responders (79.3%) reported the LF should be held every year. In 2018, 10 of 19 responders (53%) reported that the LF should be annual and mandatory (Table S1).
Written participant feedback showed that most staff found the LF useful (Table S1). Contributions by UNIS peers were valued and appeared more relevant and closer to practice than the external contributions. Participants reported wanting more time to discuss -asking for less lecturing and institutional information and more time to learn from colleagues through discussion. Responding to this feedback, collegial sharing sessions were added to the LF programme in 2018. An open call for contributions got little response, and the organising committee and department leadership had to actively encourage people to contribute. The organising committee reported that it was easier to get contributions the following year.
More than 70% of the participants providing feedback in 2018 and 2019 found the collegial sharing sessions useful. Collegial sharing sessions were described as a positive and constructive arena for learning and obtaining ideas for improvement -for the facilitator and participants. Being inspired by and learning from colleagues within and across departments was frequently reported as a positive experience. Also, collegial sharing sessions were said to lower the threshold for discussing teaching and learning with colleagues. Some mentioned that discussions sometimes deviated from the topic and were less useful when dominated by one participant.

Teaching conversation survey
About half of the LF participants answered the survey (n = 33/64), of which 15 had joined collegial sharing sessions only one year, and 18 had joined sessions both years (2018 and 2019). Eleven respondents had led a session (survey summary; supplementary Table S2).
Most respondents agreed or strongly agreed that their motivation for participating was to learn something new about a topic, get input on a topic relevant to their teaching, and know more about the experiences of the colleague leading the session (Fig. 2). Most disagreed with the statement: "I joined because we were told to attend" (Fig. 2). The motivation for joining a session was similar for respondents who participated in both years (n = 18; Fig. 2).

Motivation and expectations
Most respondents agreed or strongly agreed that their motivation for participating was to learn something new about a topic, get input on a topic relevant to their teaching, and know more about the experiences of the colleague leading the session (Fig. 2). Most disagreed with the statement: "I joined because we were told to attend" (Fig. 2). The motivation for joining a session was similar for respondents who participated in both years (n = 18; Fig. 2 Respondents that had led a session agreed or fully agreed that they like to share and discuss their own and others teaching (median score 4 "Agree", n=11; scale 1-5; Table S2). Respondents participating in sessions agreed or strongly agreed that they like to share and discuss their own and others' teaching and looked forward to participating. Few respondents regarded teaching as private or felt intimidated by attending (Fig. 3). Several did not know what to expect from a collegial sharing session the first time, and had no strong opinion on whether or not they would be able to contribute to the discussion. The survey indicated that those participating for the second time were somewhat more confident about their ability to contribute to the discussion ( Fig. 3; Table S2).

Gains and experiences
Most respondents reported discussing the planned topic during the collegial sharing session (n=28/33), whereas five of 33 reported that they had only partly discussed the planned topic (Table S2). All respondents reported benefits from the sessions, both concrete gains, such as new ideas to improve teaching, and personal development, such as being more confident in discussing their teaching with colleagues. All felt the sessions created a safe space for mutual support and guidance (Fig. 4).
The survey showed that almost all respondents agreed or strongly agreed that they learned to know their colleagues better and, as a result, knew whom they could ask questions and discuss their teaching with (Fig. 4). Interestingly, those reporting that their collegial sharing sessions just partly discussed the planned topic showed the same high score on this statement (median score 4, "Agree"; n=5, scale 1-5; Table S2). Most respondents also agreed or fully agreed that collegial sharing sessions had revealed that many teachers desire to have these kinds of conversations (Fig. 4). Of those reporting more informal conversations after joining a collegial sharing session, about half said that these conversations had led to changes in their teaching (n=9/17; Table S2). Further, these conversations were with colleagues they did not discuss such issues with before and included colleagues from other departments (Fig. 4).
None of the respondents reported speaking only with colleagues who participated in the same collegial sharing session. Conversations seemed to occur in various settings where colleagues happen to meet (in the corridor, at the coffee machine etc.), and no arena seemed more important (Table S2).

Summary of interviews
The interview analysis gave results that support the findings from the survey (full interview summary; supplementary appendix S2). The informants agreed that they now have more informal significant conversations with colleagues than before they attended their first LF. They described significant informal conversations about teaching as rare before the first LF.

Robin: 'I had no network before. I don't recall talking about teaching techniques before…' Charlie: 'I did not have anybody to talk to about teaching. ….I have always been interested in developing my teaching, but I felt that I had to go outside of my university to find colleagues to talk to.'
All informants agreed that the LF in general, and the collegial sharing sessions in particular, have been useful. However, one informant felt somewhat intimidated by the format of the collegial sharing sessions the first time.
Robin: 'My first thought was "Oh no". We will be forced to talk and share. I was apprehensive about what was expected and whether I could contribute. My training is in science -not education.' The perceived gains from joining collegial sharing sessions differed. One benefit was that the sessions helped develop a vocabulary to support conversations about learning and teaching. Another was to have more people to talk to about learning and teaching, which contributed to a feeling of belonging. The interviews supported the impression that collegial sharing sessions had increased the number of conversational partners. However, the informants still experienced these informal conversations mainly within the departments. The informants experienced a change in topics of conversation after the introduction of collegial sharing sessions.
Robin: 'I don't recall talking about teaching techniques before… I used to talk about the curriculum, not the methods. Now there are a few people that I ask for ideas. I am definitely more comfortable now.'

DISCUSSION
Our study identifies collegial sharing sessions as an important mediator for more significant informal conversations. Introducing collegial sharing sessions as part of the Learning Forum resulted in more informal conversations among colleagues, and these conversations meet the criteria of being significant (Roxå & Mårtensson, 2009). Thus, as we hypothesized, collegial sharing sessions led to more significant informal conversations. The survey data also suggested that these conversations to a higher degree than earlier included colleagues from other departments, indicating more significant informal conversations among a wider network of colleagues.
Further, our results support that it is possible to facilitate significant informal conversations through topdown formal structures initiated by the institution (Horn et al., 2020;Van Waes et al., 2018). In our case, the implementation of collegial sharing sessions resulted from a feedback loop between organisers and participants. Initially, formal requirements led the institution to invest in an annual event to enhance educational quality. Over time, the institution developed the forum, following up on suggestions and feedback from the participants (Fig. 5). Our findings indicate that the feedback and suggestions became more informed over time, reflecting increased knowledge and awareness about educational quality among participating staff members. For instance, it was easier to recruit contributors to collegial sharing sessions in the second year, and fewer had doubts about their ability to contribute to discussions the second time (Fig. 3).

Collegial sharing sessions: low investment -high output
Lectures can work well for sharing one person's ideas with many, but the possibility for discussion is limited. In contrast, collegial sharing sessions reach fewer, and participants must choose between sessions and topics. Even though collegial sharing sessions limit the possibility of sharing ideas across the institution in one session, our data show that they facilitate discussions during the formal event and lead to more meaningful conversations after the event (Fig. 4). This is in line with Thomson (2013), which found that workshops and other formal meetings that offer opportunities for teaching discussions led to academics discussing teaching before, during and after the formal setting.
Although the collegial sharing sessions were a formal initiative, the structure of the sessions resembles an informal setting. The topic is close to the facilitator's own teaching practice. Further, a collegial sharing session requires little preparation and is not aimed at a broad audience; it is sufficient that six to eight others sign up. The "low stakes" might explain why the participants regarded these sessions as a safe place for sharing experiences -success or failure. Several survey respondents reported that discussions in collegial sharing sessions inspired new ideas to improve teaching and lowered the threshold for sharing their own experiences. In light of Vygotsky's socio-cultural learning theory, these collegial discussions became mediating tools in the participants learning processes, and over time expanded their zone of proximal development (Vygotsky et al., 1978).

To know your colleagues is not enough
Our study took place at a small institution, where most colleagues know each other. Still, significant informal conversations about teaching were rare or absent a decade ago. A recent study on school teachers in mathematics showed a clear relationship between the quality of learning opportunities in formal meetings and the subsequent development of advice-seeking ties in informal social networks (Horn et al., 2020). Horn et al. (2020) also highlight the importance of identifying efficient ways to provide formal support: '..merely allocating time for teachers to collaborate does not in itself ensure high-depth meetings.' The lack of a common language and motivation to start a conversation about learning rather than practicalities can be another factor. A recent study explored the significance of a formal pedagogical course on informal conversations about teaching and learning (Soomere & Karm, 2021). Their results

Institutional
(formal) Public indicated that formal pedagogical training created a common ground that facilitated informal conversations about teaching among participants also after the course (Soomere & Karm, 2021).
Another aspect that can explain the absence of significant informal conversations is the history of this particular academic institution. Until 2009, UNIS only employed scientific staff on three-year contracts (The University Centre in Svalbard, 2010). Despite proximity among colleagues, the employment terms created a rather ephemeral collegium. Compared to the present, the number of staff members was far lower a decade ago (65 in 2009 versus 130 in 2019) (The University Centre in Svalbard, 2010, 2020). Only one or a few people represented each discipline. This might have affected the level of trust and mutual understanding among colleagues, making it harder to create a stable network of conversational partners.

A formal-informal feedback loop promotes a knowledge-sharing culture
The teaching staff in our case-study have developed a sense of ownership of the LF and collegial sharing sessions. This is seen in their motivation to participate and gain from the activity, as well as in their direct feedback. As the LF has developed, the teaching staff have not only influenced the program, but become the facilitators and leaders of the sessions themselves. This is an example of going public locally (Mårtensson et al., 2011). Our results also resemble the findings of Dorner and Belic (2021). They studied how a series of semi-formal conversations about teaching developed from being important for individual academic development, to over time focused on collective learning and institutional transformations. They noticed that by facilitating these conversations, teachers recognized that they were not alone, and "thereby creating space to grasp the 'big picture' and articulating obstacles to individual and collective academic development" (Dorner & Belic, 2021).
In addition to the feedback loop created between the formal and informal arenas, activities like the LF and collegial sharing sessions can become a bridge to a third arena where teaching staff go public outside their home institution (Fig. 5), as indeed this paper is an example of.
Collegial sharing sessions have efficiently fuelled significant informal conversations and sharing among colleges, both locally and at the institutional level. We argue that introducing appropriate formal arenas for sharing in settings where informal conversations about teaching and learning are rare makes it possible to create motivation and opportunity for teachers to continue conversations in informal settings.
In the same way that extending the backstage conversation to the frontstage can support development (Mårtensson et al., 2011, Katajavuori et al. 2019, frontstage conversations and formal activities can support the informal backstage conversation and expand teachers teaching and learning networks (Van Waes et al., 2018;Van Waes et al., 2015). A sustained culture change must involve both the backstage and frontstage (informal and formal) domains. In our case, UNIS has made the LF a clear institutional priority, and participant commitment is high. Our results indicate that having collegial arenas for sharing might be more important than the specific topic, as those participants who deviated from the predefined topic also got to know their colleagues better and knew who they could continue to discuss their teaching with.

Limitations
This study examines one educational development initiative and the process of how it developed in a way that led to more informal conversations about teaching and learning. We are aware that other external and internal factors not discussed here may have impacted this process and the teaching staff.
The sample size is small, and our survey's response rate was 50%. The participants are limited to one institution in a unique setting. With these limitations in mind, we would still argue that the feedbackdevelopment process resulting in more informed informal conversations is guided by principles that could be useful and effective in other settings.

Implications of our work for others
Campfires are regarded as a good venue for meaningful conversations, but no one gathers around a campfire if there is no campfire space. Through the development of the LF, we managed to create a campfire. The LF was clearly important to facilitate a common language and to discover the need for more conversation space. Sharing sessions made participants come closer together, making conversations with the one sitting next to them easier around the campfire.
We argue that the key to success in our case is the combination of regularity and flexibility of the LF in general and the collegial sharing sessions in particular. The introduction of an arena for collegial sharing, with a formal structure and predictable occurrence, has the potential to fuel significant informal conversations and sharing among colleagues. Yet, this arena must have a flexible format that is responsive to participant feedback, needs, and suggestions. Allowing the format to be responsive and change over time creates ownership of the event, and enables participants to engage according to the stage of their own development.
The collegial sharing sessions mixed participants with different degrees of experience applying a scholarly approach to their teaching. Such a mixture seems powerful and might create significant conversations where more experienced teaching staff function as mediators and facilitators, in line with Vygotsky's socio-cultural learning theory (Vygotsky et al., 1978).
We believe another important aspect of successful collegial sharing sessions is to keep the conversational theme close to practice for the participants. The specific topic is not the most critical factor, but rather that the discussions are rooted in local practice with peers. This adds a sense of ownership and meaning for the participants.

SUPPLEMENTARY ONLINE MATERIAL
This supplementary material consists of the following: Table S1 -Summary of background data regarding the development of the Learning Forum and collegial sharing sessions Appendix S1 -Teaching conversation survey questions Table S2 -Count summary and descriptive statistics of the Teaching conversation survey Appendix S2 -Summary of three interviews