Learning from experience

‘Bring in the next case’

In a reference to education at American business schools, Mintzberg (2004) jokes about a recent Harvard MBA graduate walking into an office for his first day at work. The graduate comes in, sits down at his desk, and then exclaims to his secretary ‘bring in the next case!’ The essence of this joke is that many business schools teach students to solve teaching cases instead of preparing graduates for real work in the real world. They learn to solve cases, not to interact and deal with the dynamics of everyday organizational life. Students are taught to solve didactically designed cases, instead of learning to deal with the ambiguity and ambivalence that comes with reality. Therefore, the more serious question prompted by the joke is how means of (executive) education and training can be developed that better reflect the messy and ambiguous social realities of life and work in public service. Are there ways to bring the richness and ambiguity of public administration work and practice into the public administration classroom?

The same questions about the connection to practice can be asked for public administration research. Positivist research in public administration relies heavily on methods that delineate people who work in practice as ‘respondents’, or even worse, as ‘data’ (Whittington et al., 2003). To streamline the causal argument, contextual elements that define reality are ‘controlled’ as a variable or placed outside of the research design entirely. The characteristics that make life in public service complicated are left out of the model, in order to hypothesize causal models. This leads to methodologically sound but highly simplified accounts of reality, which then form the basis for discussions about the profession of public administration and its processes (Abbott, 2004; Flyberg, 2006; Novotny et al., 2001). This can produce good quality positivist public administration research, without the findings being of much significance to the practitioners inside public administration.

In search of proximity to practice

In order to develop more accurate and rich accounts of practice, some researchers attempt to move in ‘closer’ to where the work of public administration is done. That requires a balancing of two important values: the proximity to and distance from research subjects. Practice is often studied ex-post and through indirect methods, such as surveys, interviews and focus groups (Flyberg, 2006; Ragin and Becker, 2005; Yin, 1994). This ensures a level of distance of the researcher, but also diminishes the data’s bounty (Abbott, 2004; Jarzabkowski et al., 2007; Mantere and Vaara, 2008). Other research methods look for more proximity to practice; for instance, by conducting interpretative (Wagenaar, 2011; Yanow, 1996; Yanow and Schwartz-Shea, 2006) or participative research (Greenwoord and Levin, 2007; Reason and Bradbury, 2008) near or inside the world of practice. In research traditions such as action research (Greenwood and Levin, 2007), the researcher positions himself near and around subjects, to develop a look and feel of practice and take that into account in research; for instance, by interpreting observations and turning them into working hypotheses and practical theories (Reason and Bradbury, 2008).

However, proximity is not necessarily the same as the observation of real practice. Although a researcher may physically be amongst his subjects, he typically remains at a distance from what is described. For example, in a typical research, the researcher visits a building, is offered coffee, conducts interviews with a number of interviewees and takes notes that will later form the basis for his case description. Throughout the process the researcher sees interesting things, but in a staged setting, with people talking about practice instead of actually doing it. Proximity is not the same as ‘seeing real practice’; it is closer to where the practice happens, not necessarily closer to what practice actually is.

These are all familiar topics in the discourse of public administration research and in discussions about qualitative research, participatory research methods and interpretative research (Hajer and Wagenaar, 2003; Wagenaar, 2011; Yanow, 1996). Therefore, to add to these existing discussions, it is interesting to take an opposite route for balancing the gap between distance and proximity. What if we take a different route to come nearer to the practice of public administration? Instead of moving in closer to the real life places of work it is also possible to invite practitioners to bring their real life practical experiences and dilemmas into the staged setting of a training programme. Instead of looking at a place of work, we can ask practitioners to take their work into a place for learning – an executive education programme.

In doing so, the classroom setting of executive education is used as a space where practitioners can bring in their real life practice. They are provided a space that is distant from their everyday work context – they are literally in a different room, in a different context, amongst different people – and that allows them to reflect on their work. This reflection on the work by practitioners can then be used as ‘data’ for interpretative research. In doing so, the classroom of executive education becomes a research context for studying the practice of public administration.

This way of organizing and utilizing executive education may help to bridge several recurring boundaries of public administration education and research: education and practice, research and practice, education and research. By bringing practical experience of participants into the classroom for the purpose of education, an interesting space for research emerges, one that is built around interaction about real life problems and dilemmas of participants and that uses the interaction as data for interpretative pubic administration research.

A study of executive education and research in practice

This paper explores one such attempt to combine educating practitioners with doing interpretative research, in a classroom that researchers use as a place for research instead of researchers moving into – or nearby – a workplace. In this format the classroom remains a ‘room for education’; it is not a ‘social simulation of work’ (Bower, 2008; Kornberger et al., 2015; Mintzberg, 2004), but the classroom also acts as a place for interpretative research. Participants are not interviewed in quasi-proximity of their real work (e.g. in a meeting room at their offices), but deliberately take distance from their workplace to reflect on their work. They bring the practical dilemmas from their workplace into the classroom and engage in intense interaction with fellow participants, after which the dilemmas and the interaction become data for interpretative research by the faculty of the education programme.

The educational merit of this approach is that participants can introduce their practical dilemmas into the classroom and reflect on that ‘real’ experience for their learning. This flips the joke of Mintzberg on its head; the faculty combines the roles of education staff and interpretative researchers, who use the classroom as an empirical space for interpretative research about the practice of public administration. This type of research can be a new branch on the growing body of interpretative research methods and can be powerful method for teaching public administration to high-level practitioners – for who the combination of theory and practice is a crucial element of training.

In this paper we explore an example of such an educational format, which was used in executive public administration programmes in the Netherlands: the so-called learning atelier. We will first explore the principles of teaching and research of public administration in the learning atelier. Secondly, we will present how we organized the learning ateliers, present the various didactical forms we used and how these translate into research and reflect on what happened during the programmes. Thirdly, we will discuss the meaning and limitations of our findings, and reflect on potential for introducing this method in other programmes. We will not extensively go into the content of the ateliers, since this paper is primarily about the setting of the ‘classroom’.

Bringing practice into the classroom of executive education

The case study method is practised in many executive education institutes in a variety of courses (Kornberger et al., 2015; Pollitt and Op de Beeck, 2010; Van der Krogt, 2015). It is a didactically strong and interactive way of teaching that has helped millions of students to better understand theory. However, we argue that in order to train seasoned practitioners about the intricacies of working in the context of public administration there may be another way, one that not only helps understand theory and work on scripted cases, but that also activates the rich experiences, skills and expertise of participants for the learning process (Jarzabkowski and Whittington, 2008). The classroom becomes a social and temporal space for participants to share knowledge with fellow students. Literally, it is a time and place to learn, set apart from the everyday routines of the work floor, but at the same time closely related to the dynamics of practice.

Educational cases typically present a ‘reality’ that is scripted for educational purposes. The case is usually based on a real situation, but remodelled to support teaching a concept or theory. Moreover, cases are written products that present reality as a text or a set of numbers. The text is structured, with information partly hidden and deliberately overloaded elsewhere. Realities are presented in numbers, and historical context is presented in data-charts. Process is embedded in the scripting of the case, for instance when the case at some point ‘fast-forwards’ to confront participants with new developments or with the supposed results of choices made earlier in the case. Students are repeatedly asked, ‘what would you do?’, and sometimes have to act out the choices they made and strategies they proposed in a social simulation setting or role-playing. Cases represent some features of actual work practice to help students learn (Bower, 2008). They are closer to the real life situations of practice, but remain artificial simplifications of it (Balogun et al., 2003; Liang and Wang, 2004).

There is much that is positive to say about case-teaching. Instead of learning theory from a lecture, book or PowerPoint presentation, case-teaching brings students closer to an interactive and practical model of education. As a result, concepts come to life. Insights are not neatly prepared for students, who are encouraged to actively develop their own conclusions through exercise and practice: they literally have to work for it. In addition, case-teaching methods confront students with the consequences of their choices; the question ‘what would you do?’ is often followed by a response by a fellow student or professor in reaction to any suggested move by students.

In spite of these various benefits, and especially with regard to teaching experienced practitioners, there remains essential and rich unexploited potential that begs the question: how far can the practical input of the students themselves be used to develop knowledge and create better training programmes? Is it possible to not only bring educational cases into teaching programmes, but also to use the practical experience of participants as cases for training?

Classroom learning as ‘data about practice’ for research

Case research has been the basis for many studies of public administration practice, especially in the more qualitative branches of research (Ragin and Becker, 2005; Yin, 1994). There is a growing body of interpretative research, and action research, that attempts to move research closer ‘into’ the world of practice and to collect data as much from first hand as possible (Flyberg, 2006; Flyberg et al., 2012; Novotny et al., 2001; Yanow, 1996). Researchers go out to observe practice, conduct interviews, take field notes and use that to write a ‘case description’. In action research the researchers often act as members of the projects they study; they combine practical work with doing research, and for that matter study the outcomes of their own work (Greenwoord and Levin, 2007). Applying such methods has been very helpful in developing a different line of research – and theories – besides the more traditional, ‘normal’ positivist and often highly quantitative research tradition.

There is evident value in these types of research, just as there is merit in positivist research. However, there are also limitations to this model (Abbott, 2004; Flyberg et al., 2012). Cases are about practice, but they only partially reflect the messy nature of reality (Flyberg, 2006). Interpretative researchers who move into practice, or become part of it by doing action research, attempt to come nearer to practice. They look for proximity, they want to be ‘close’ to where the practical work is, in order to have a better look. In the educational format we look at in this paper, researchers take a different approach to study practice and to achieve proximity, not by visiting a practical reality and describing what is seen there, but through interacting with practitioners who literally walk into the classroom and bring their practice in with them. Then, proximity does not come from researchers moving into practice, but from practitioners who walk into the research – which is in this case the setting of a classroom. The practice that can be observed in such a staged classroom setting is not quasi-work, but is the temporal, physical, intellectual and emotional space for practitioners’ interaction about their work. In such a setting practitioners discuss how they solve problems, struggle with the practical dilemmas of work and think about a possible repertoire for dealing with dilemmas. Such direct observation – of a context ‘away from work’ that allows practitioners to reflect on what they really do – can add new insights to the field of public administration.

This research approach can best be described as a part of the body of interpretative research (Hajer and Wagenaar, 2003; Wagenaar, 2011). In part, it is also participatory research, because researchers engage directly with the practice they study – the participants. Such interpretative research can provide access to rich material and holds the potential for new insight into practice (e.g. Mantere and Vaara, 2008). However, this type of close access also creates methodological difficulties. The researcher’s proximity to the practice being studied inherently leads to matters of distance and autonomy of the researcher (Rhodes et al., 2007). Therefore, researchers need to be open and transparent about how they collect their data (Rhodes et al., 2007), and in this case also where they do that – in a classroom for executive education.

Executive education and interpretative research

Building an educational format around the practical experience and professional dilemmas of public administration practitioners can serve two interesting purposes. Firstly, it can be a building block for teaching executives without having to resort (solely) to scripted teaching cases to bring the ‘real world practice’ into the classroom. Faculty can invite participants to bring their personal experiences into the classroom and stage the classroom as a space for interaction about these personal experiences. This can lead to a good learning experience of the executives that provides them with theoretical perspectives and practical reflection.

Secondly, the personal experiences and practical dilemmas of participants in such an executive education programme can act as a place for interpretative research. Researchers of public administration have long struggled with means to come ‘close’ to where the real practice of public administration is. They have moved into the workplace, sometimes acting as co-workers of practitioners, in order to gather more rich and ‘real’ data about practice. In order to achieve proximity to practice researchers have moved towards – and moved in with – practice. Using the temporal and physical setting of the executive education classroom as a place for research provides a different type of proximity, by practitioners who bring their work practice into the classroom, where they can be ‘studied’ directly by interpretative researchers. This does not provide an image of work-in-action, but does allow for deep reflection about how seasoned practitioners experience the dilemmas of public administration. This can provide a valuable addition to the current spectrum of public administration research methods.

Education in learning ateliers

This paper discusses three examples of executive education programmes that were carried out for three groups of high-level experienced civil servants (see Table 1). The educational format was called the learning atelier; these are executive education programmes tailor-made for experienced professionals and managers. The idea behind the ateliers is that participating practitioners have the opportunity to take their real world challenges into the learning space, and that the faculty – who are also researchers – use the interaction about the participants’ real world issues as assets for learning. The faculty adds theories and reflection to the interaction to analyse and better understand these challenges through sessions in the learning atelier (see Edwards and Ringeling, 2003). This helps participants who bring in cases to better deal with the cases, it helps participants to learn more about theories, concepts and options to deal with dilemmas of public administration and it helps researchers to study practice and develop or test theories and concepts. Therefore, the goal of the learning atelier was threefold.

OPEN IN VIEWER

Table 1.

Three learning ateliers as spaces for learning and research.

	Classroom as space for learning	Classroom as space for research
Learning atelier: Strategic policy advice	Each participant brings his or her micro-narrative about doing public sector strategy in the municipal organization into the learning atelier. This is discussed in class; fellow participants reflect on the micro-narratives. From the individual micro-narratives the faculty and the group distinguished two sets of more general dilemmas. These are apparently professional dilemmas of doing strategy in a public sector organization, and the dilemmas of the particular municipal organization as such. For the participants, the learning atelier then focused on the general dilemmas of doing strategy, and on how these could be dealt with effectively in the context of their municipality. For doing this, participants learnt more about different approaches of public sector strategy, management of expectations and the inherent dilemmas of working ‘in between’ (e.g. theories of boundary work). During the atelier participants applied repertoire in practice and used the sessions to reflect on the effects of their new interventions.	The micro-narratives of the participants provide a first account of a practical dilemma in the work of doing public sector strategy. The group discussions of the dilemmas brought to the fore the perspectives of other participants on these dilemmas and showed what common issues were. The group discussions and narratives provide information not only about personal experiences with doing strategy, but also about the organizational positioning of a strategic unit. What is the best position for such a unit and how can a strategy unit function in between the different branches of the organization? The sessions provided information about what participants considered good intervention strategies for ‘doing strategy’ and for working in between in the organization, Moreover, the first attempts to apply this repertoire were discussed in-class, so this was also available as material for research. It provides a first ‘practical test’ of some of the theories participants developed during the learning atelier.
Learning atelier: Foresight	The project team of the foresight project had reflection sessions with the faculty, to present their approach to the project and to discuss the design and next steps. During the reflection sessions the faculty presented insights from the foresight literature, and shared different examples of linking foresight to the practice of management and policy making. The participants specified their learning question into exploring means to bridge the gap between the long term and the short term in policymaking. Moreover, they wanted to learn more about how to deal with the dilemmas of (1) balancing creativity and expertise, (2) combining internal and external stakeholders in the process and (3) balancing thinking ‘out-of-the-box’ and linking up with the current discourse of the organization – which was important in order for the output of the foresight project to be taken into account by the Board. During the course of the project the participants reflected on their practical experiences and sessions were used to co-design the next step(s) in the process. Eventually, this lead to the idea of the group to experiment with the concept of ‘creative competition’ in doing foresight.	The reflection session directed attention towards the dilemma of doing foresight in a government agency. More specifically, it showed that the team needed to balance several dilemmas in order to do a foresight project that would be relevant for decision-makers. This balancing of dilemmas became the focal point of the study; participants and faculty used the practical experiences of the participants to draw up a more detailed list of dilemmas for doing foresight in a government organization. When the participants had decided they wanted to use ‘creative competition’ as a strategy for organizing the foresight project, that particular method became the primary focus of the research; how did this method play out in the practice of this ‘case’? How did the choice for ‘creative competition’ help to solve the dilemmas of the participants, and did it lead to a better foresight project with more impact?
Learning atelier: Political assistants (PAs)	This atelier consisted of reflection sessions of three different groups, each participating in one session. The political assistants (half of them currently working as PAs, half of them ‘retired’ as PAs and now working in other positions) had never before discussed their work as a ‘profession’, and many of them met each other for the first time in such a context. Each political assistant started with a micro-narrative to introduce himself and present his practical dilemmas; this also included a personal biography. After that, participants discussed each others’ dilemmas and looked for commonalities. They learned about their personal dilemma from the reflections of colleagues. Also, they identified common dilemmas and issues that apparently formed an important and inherent aspect of their profession. Faculty facilitated the discussion, but also presented several theories from similar functions (political civil servants) from other countries, to further fuel the debate about the profession of political assistants in The Netherlands.	Researchers collected the dilemmas of the PAs. During the sessions it became clear that there were several recurring dilemmas. These became the focal point of research; we used the discussions and reflections during the sessions for further exploration of the dilemmas and to make them more specific. Also, political assistants talked about their repertoire to deal with the dilemmas. As a secondary finding, we gathered much information about the personal biography of political assistants: what kind of people they were, how they were drafted, where they had worked before, where their career had taken them after their years as PA, and so on and so forth. We used these narratives and dilemmas to describe the ‘real work’ of PAs, identify the most important elements of that work and also elaborate on the most important dilemmas of the work of Pas. We used that to place the PAs in the political system of The Netherlands in the literature about political appointees in systems in other countries.

Firstly, by analysing and working on real life cases, participants learn, which is the primary objective of the learning atelier. Participants learn from the interaction with their personal experience and the experiences of other participants.

Secondly, through work conducted by the participants, supported by the contributions of academics, most of the real life cases are in a way resolved. Participants are helped in the practical dilemma they face; for example, they better understand the situation, redefine problems and see a new range of options.

Thirdly, through participating in the process, learning from discussions by and with participants and by observing them in interaction about their personal cases, the faculty (who also act as academic researchers) generates data for research.

Teaching and learning

In a learning atelier, each individual brings his or her own expertise to the learning space. In order to facilitate the interaction about these experiences, learning ateliers are staged as spaces for the exchange of knowledge and experience between professionals. Participants bring their individual projects, previous experiences and real world interventions, and introduce them to other participants. They make visible the problems, questions and dilemmas of practice, reflect on underlying dilemmas and describe their positions and roles in their cases. As a group, they use the presented theories and one another’s expertise to explore potential new directions for dealing with the presented dilemmas. As the next step, participants not only talk about possible solutions or analyse the pros and cons of options, but they are also challenged to translate those discussions into a practical solution that can be applied to the real case – by the participant that brought in the case or by a group of participants that worked on the solution as a team.

Lecturers have a threefold role in the learning atelier. Firstly, they fulfil the role of resource. They present new concepts and theories to participants, representing state-of-the-art theory and teaching from their respective fields. They also help participants reflect and see connections between their sometimes very different work practices. Secondly, they also serve the role of warden by overseeing and safeguarding the quality of content produced, and ensuring that discussions lead to concrete results, applicable insights, a workable methodology or towards new repertoires for action. However, the role of the lecturer as warden also needs to take into account that participants are co-developers of their learning process. The lecturer safeguards the process, but at the same time should leave it open enough for significant co-ownership by participants.

Thirdly, the faculty researches the interaction in the classroom. The interaction of participants is utilized as data for interpretative research. The process of interaction is an interesting body of data that can be used to further the understanding of public administration.

Learning atelier: ‘Strategic policy advice’

We used the concept of learning ateliers in an in-company trajectory with three groups of 15 participants from the strategic policy advice unit of a large municipality in The Netherlands. At the time of the programme, the unit was undergoing a transformation towards a new role, from general policy advice to more ‘strategic policy advice’. Participants wanted to learn more about different approaches to public sector strategy and the dilemmas of working ‘in between’ the regular policy units and the cities’ politicians. They wanted to reflect on their personal preferences and explore a new repertoire. They also wanted to exchange ideas on how the new unit should position itself. During the atelier, participants discussed the dilemmas they faced when working in between the ‘policy silos’ and how difficult it was to maintain a more long-term approach in strategy when dealing with politicians. Also, this learning atelier pointed out the importance of ‘managing expectations’ for strategists and strategic units. Because different stakeholders tend expect different things from the strategy unit, it is impossible to satisfy the needs of everyone all the time. Furthermore, participants concluded that because strategic thinking is often also ‘critical thinking’ – going against popular fads in the organization or political ideology of politicians – it is important to keep a close eye on support for the unit. These proved to be useful insights for the repositioning and professionalization of the unit, and were also intriguing research discoveries that increased our understanding of the inner workings of strategy in a policy organization. These insights set the faculty up for a study on the ‘in between’ position of public sector strategic strategists; moreover, faculty members produced research papers about ‘managing expectations’, a concept that participants in the learning atelier had extensively elaborated on as a strategy to work ‘in between’ various organizational contexts.

Learning atelier: ‘Foresight’

The ‘foresight’ learning atelier was a programme for a small group of strategic advisors to the Board of Rijkswaterstaat (RWS) in The Netherlands – the executive agency responsible for the planning and execution of road and water works. Their agency was conducting a foresight study as part of a strategic planning cycle. The topic of the study was the ‘future role’ of the agency, and the Board had already announced it wanted to use the results of the study as a basis for designing its strategic plan for the next six years. With this task in hand, the group set out to conduct its study and, in conjunction with its activities, organized a place for reflection in the learning atelier. The group consisted of the five participants who formed the team that conducted the foresight study. The faculty provided participants with a series of theoretical perspectives and approaches for conducting foresight drawn from the literature; participants brought in their experiences and dilemmas and used the learning atelier as a place to reflect on the ongoing process. The programme ran for the entirety of the foresight process and, thus, allowed participants to discuss follow-up steps, reflect on the outcomes of previous steps and to contemplate overall direction. More specifically, the group was concerned with striking the right balance between creativity and expertise, and wanted to strive to connect truly innovative ideas with running themes and priorities in the organization. The team felt it was important to invite outsiders into the process, in order to think beyond the deeply engrained beliefs of their highly specialized organization; they wanted scenarios to really span the boundaries of thinking. However, they also realized that the scenarios had to be plausible for an audience of experts. Both creativity and content were crucial for the scenarios. They had similar concerns about the willingness of the Board of the organization to accept the scenarios; the Board was committed beforehand, but would probably not accept too ‘far-fetched’ findings. During the programme, the group arrived at a decision to employ a method typically applied in creative thinking and large-group interventions – ‘creative competition’ – and use it in the context of foresight. This represented a new combination in the field that helped to balance the dilemmas of creativity and content, and of innovativeness and connectedness to current agendas. This combination of different contradictory values for conducting foresight in the context of public administration became the topic of the research output of the learning atelier. The process provided insight into a new method for conducting foresight and into the practical struggles of a group of strategic advisors who wanted to enhance the time-horizon of an organization. It provided valuable insight that furthered the field of strategic foresight. This learning atelier resulted in two articles in research journals and a professional publication.

Learning atelier: ‘Political assistants’

The ‘political assistants’ learning atelier consisted of a series of ‘reflection sessions’ with political assistants to political representatives, which form a special group in the public administration ‘landscape’ in The Netherlands. Assistants form a unique group in the civil service in The Netherlands because they are the only political appointees in an entirely merit-based bureaucratic system. Each elected political representative is allowed to appoint one ‘political assistant’, who acts as his personal advisor within the bureaucracy. Often, the ‘PA’ (as they are called) works on the interplay between the political executive power and the Parliament or – in the case of a municipality – City Council. He or she talks to politicians and stays aware of possible deals, worries and opportunities. In addition, the PA is tasked with considering the party political profile of the elected politician. In The Netherlands, government is characterized by the formation of coalitions and collegial decision-making, often resulting in elected political executives ‘losing’ much of their political colouring. They represent the policy and political course of an entire cabinet or coalition, which is inevitably a mixture of political ideas and ideology from various parties. PAs do all that political work in close contact with the civil service. Ideally, civil servants would use that information about the political dynamics to improve their policy advice, but in practice there are often conflicting perspectives between a PA’s purely political advice and the policy proposals put forward by the civil servants.

In the programme, we had intense group discussions with mixed groups of PAs. This alone was rather unique and unusual, as they seldom meet as a group and are each from different political parties. Under conditions of absolute discretion, all participants talked freely about the dilemmas in their work, which variously included reflections on how they dealt with public servants, how they influenced the course of debates in parliament and also thoughts on their often complicated and highly personal relationships with their political boss. PAs almost literally live their boss’ life; they work alongside one another beginning in the morning and continue late into the night. During that entire time, they think and act with only one thing in mind: the well-being of their political boss. They arrange what is needed in the political party, continuously try to imagine what he or she would think of an issue and keep a close eye on other needs. Sometimes it means something as simple as delivering coffee, and other times it requires a more complex role, such as a long night of negotiation with other political groups. Many participants struggled with this and even said they felt as if they were losing their own identity. The sessions helped them to better understand their own profession – which, at the start of the programme, many of them did not even consider a ‘real’ profession – and they found a practical repertoire for dealing with some of their dilemmas. As research, the programme provided exclusive and, as far as we as researchers could tell, unique insight into the unseen realities of work in the political core of our government system. The programme helped individual participants, but also improved our understanding of the interplay between the civil service and the elected political executives at the highest echelons of public organizations. Furthermore, it increased understanding in the civil service of the PAs’ peculiar function and work variety in the bureaucracy. The faculty of the learning atelier used the interactions about the profession of the PAs as material for a professional publication and for a research paper that was presented at two academic conferences; as this profession had hardly been studied before in The Netherlands, this was a very interesting output of the learning atelier.

Data collection

The primary basis for data is the plenary group discussion where participants talk about a set of topics and exchange their ideas and experiences on the selected subjects. During the flow of the discussion, participants introduce new themes they consider important. The discussion can be facilitated not only by the faculty, but also by the group itself. The discussion is not an interview, but opinions, ideas and experiences expressed during conversation can be linked to a topic-list just like a researcher would in order to analyse data from a series of interviews. Researchers typically pre-determine certain topics, but some emerge organically during the session. An important role for the facilitator or the faculty is to ensure that the discussion leads to learning; that is, at the end of the session, points and arguments need to be translated into lessons or repertoire that participants can use in their work. Participants have to take home some sort of insight, idea or reflection from the discussion. The dual nature of the sessions has to remain intact; it is not a group-interview meant to service research, but a learning programme that is also a locus for research.

In learning ateliers, group discussions are often part of a larger and more continuous debate that goes on over the length of several sessions. During the programme, participants have time to revisit ideas and repertoire from previous sessions into their work, experiment with them and discuss their experiences in a future session. The programme is a place for reflection on practical experiments that participants might undertake. The faculty needs to ensure this foundational premise of the programme, which can be especially difficult since different lecturers often conduct sessions. As learning ateliers service a group of high-ranking seasoned practitioners, topics require expert lecturers. Expert lecturers are ‘flown in’ but are not familiar with the overall dynamics of the group or longer-running discussions. That is why most of our learning ateliers have a so-called ‘learning manager’ who oversees and monitors the learning process, not so much in terms of process dynamics, but in terms of intellectual content. All in all, the learning atelier needs to result in a new practical repertoire and often a practically applicable solution to a real problem.

It can also be interesting to open the debate entirely to what participants want to discuss, as opposed to always having facilitated group discussions. In such an open conversation, participants themselves decide what needs to be discussed. They present their own dilemmas, choose where to delve into further detail and when to move on to the next topic. The discussion can be facilitated, but the faculty rarely intervenes in the choice of topics. The motivation behind this format is to encourage practitioners to explicitly identify and address topics, dilemmas and worries implicitly associated with their professions. This often involves long periods of silence and pauses in the discussion. When asked to talk about whatever they consider relevant, most participants remain silent. However, usually, after some time, participants begin talking. Once they do, it is often when the most interesting notions are expressed.

Formulating the central question

In the context of the learning atelier, the formulation of a central question always follows data collection. This is not unusual in interpretative research, where the research focus becomes clear during the process. Formulating a research question means following the process, looking at what was said and by whom, determining where the discussion became passionate or particularly difficult and searching for other relevant points of interest. In addition, as mentioned previously, the learning atelier is not about research in the first place; it is a learning atelier, and there are learning questions as well. These provide the starting point and context within which the discussions take place. They structure what happens in the learning atelier, but the process is not entirely defined or confined by such preliminary learning goals; just like the researcher, participants often only discover what they really want or need to learn after they have already engaged in the process. There are several steps to take into account when formulating the research question in a learning atelier.

Translating local solutions into general solutions or repertoire

In the last phase of the learning atelier, it is necessary to translate local solutions for chosen cases into more generally applicable findings. This is an important step, because although a learning process might be characterized as having been a success if participants learned as individuals and as a group, the learning atelier also intends to provide insights to an overall body of knowledge. Participants learn in and around the practical cases they resolve, but in doing so they develop a repertoire or formulate questions that are relevant outside those cases and beyond the atelier. This process of determining general application does not happen by itself. There is pressure on participants, especially at the end of the atelier, to finalize work on their cases. Participants often have to present their solutions in a real life setting in front of bosses, peers and/or an academic jury, thus exposing their conclusions to critical review. In the weeks leading up to that, they are consumed by preparation for the finale, after which most participants are intellectually exhausted. Their tiredness is not a result of apathy, but instead the simple outcome of having already fatigued their attention span through hard work. Nevertheless, determining general applicability is a crucial step within the learning atelier. In our experience, participants have been significantly more at ease with this element when it is introduced early in the programme, so that awareness of it can permeate discussions of potential solutions from the start. Faculty can helpfully initiate first steps to draw conclusions from discussions, since they are just as much a part of the learning atelier as the participants themselves. At times, groups (or groups of individual participants) have been passionate about generalizing within the context of theory. In one such instance, participants and faculty entirely co-produced the research output from a learning atelier, with a book in Dutch and two international peer-reviewed articles co-authored by faculty and participants in the learning atelier (Van der Steen et al., 2010, 2011).

Outputs of learning ateliers

For participants, the learning atelier is about adding new insights, skills and perspectives to their own personal repertoire of tools for analysis and intervention. As a secondary output, the learning atelier in most cases allows participants to work through a personal case, or to potentially have it even solved.

The research output of the learning atelier consists of a deeper insight into practitioners’ practical dilemmas and professional repertoires. Crucial here is that these insights should be generalized beyond the unique local and temporal context of the learning atelier, and applied towards greater insights, theories and concepts as well. Insights from the learning atelier should be linked to the existing body of knowledge about public administration, a process that works through peer-reviewed publications. Also, it is interesting to use insights from the learning atelier for publications suitable for and catered towards practitioners. Since public administration is an applied field, it is important to make available and directly connect theories and concepts to the practices they describe.