‘The problem-centred expert interview’. Combining qualitative interviewing approaches for investigating implicit expert knowledge

ABSTRACT

Expert interviews are a widely-used qualitative interview method often aiming at gaining information about or exploring a specific field of action. This paper wants to move beyond the focus on explicit expert knowledge by emphasizing the experts’ individual perspectives that affect social practices in a field of action. The paper addresses the investigation of this implicit, interpretative expert knowledge by developing the ‘problem-centred expert interview’. This approach merges the theory-generating expert interview that discusses the social relevance of expert knowledge and the problem-centred interview (PCI) that offers a dialogic-discursive interview procedure investigating individual perspectives. By drawing upon an empirical study in human geography, the paper demonstrates how one could conduct and analyse this methodical combination in research practice. The following discussion shows that, despite some pitfalls, the combination appears fruitful for divulging implicit expert knowledge and understanding the inner logics of decision-making processes. The paper concludes with an outlook on possible fields of application.

KEYWORDS:

• Expert interview
• problem-centred interview
• grounded theory
• qualitative interview
• key agents

Introduction

The expert interview as a method of qualitative empirical research has been a widely-discussed qualitative method in political and social research since the early 1990s. Mainly cited in the European literature (Gläser & Laudel, 2004; Kaiser, 2014; Meuser & Nagel, 1991; Van Audenhove & Donders, 2019), it aims at exploring or collecting data about a specific field of interest. Meuser and Nagel (2009) describe the expert interview as a qualitative interview based on a topical guide, focusing on the knowledge of the expert, which is broadly characterized as specific knowledge in a certain field of action. Social scientists debate intensively about the criteria for being reckoned as an expert, the definition of expert knowledge, or the considerations for being recognised as a good or bad expert (Gläser & Laudel, 2009). Although the term ‘expert interview’ reflects on a plurality of methods, emphasizing different methodological and epistemological aspects, there are some key issues on which most of the authors agree. Experts are considered knowledgeable of a particular subject and are identified by virtue of their specific knowledge, their community position, or their status (Kaiser, 2014).

In general, qualitative interviewing emphasizes the importance of investigating experiences and perspectives of the interviewees for developing a better understanding of social reality (Edwards & Holland, 2013; Flick, 2018). Although the expert interview is methodologically situated in the qualitative paradigm, in practice, individual relevancies of experts tend to be overshadowed by the researchers’ interest in collecting information about a particular social field (Bogner & Menz, 2009; Mattissek et al., 2013). Since individual perceptions and orientations of experts are seen as essential for shaping social practices in a field of action, social scientists argue for a broader understanding of expert knowledge that goes beyond technical data and facts by highlighting the implicit dimension of expert knowledge (Bogner & Menz, 2009; Froschauer & Lueger, 2002; Meuser & Nagel, 2009; Van Audenhove & Donders, 2019). Revealing the implicit dimensions of expert knowledge might be of interest for different strands of research, for example, when experts’ power or influence in a certain field are being investigated or organisational conflicts are being examined.

Although many research questions necessitate a more differentiated view on the dimensions of expert knowledge, the practical questions and overall implication of investigating implicit dimensions of knowledge have not been discussed in much detail yet. Having come across these methodical challenges during my research on the investigation of key agents, I wish to introduce and discuss the problem-centred expert interview. Therefore, the paper combines the theory-generating expert interview by Bogner and Menz (2009), which distinguishes between different kinds of expert knowledge, with the problem-centred interview (PCI) by Witzel (2000), which supports the exposition of individual experiences and opinions by providing specific interview techniques. In order to demonstrate the practical application and analytical potentials of this methodical combination, the contribution draws upon an empirical example from human geography, investigating the role of key agents for socio-spatial change processes.

The remainder of this article is structured as follows: First, methodological and epistemological principles of both interview methods are outlined, and the approach of the problem-centred expert interview is derived. Subsequently, the paper presents an empirical example of the problem-centred expert interview within my PhD research project. After exposing the theoretical background and epistemological interest of my research, experiences of conducting and analysing problem-centred expert interviews are reflected upon. Finally, the paper discusses the challenges and the added value of the methodical combination, before concluding with an outlook on possible fields of application.

Combining theory-generating expert interviews and problem-centred interviews

The epistemological interest in expert knowledge

Based on the existing literature, Bogner and Menz (2009) distinguish three types of expert interviews according to their epistemological functions. The first type is the exploratory expert interview, which is frequently used to gain knowledge and orientation in unknown or hardly known fields. This helps to structure a complex field and to generate first hypotheses. In this context, interviewees can either be part of the field of interest or serve as an external source of knowledge by providing contextual knowledge about the target group. The second type is the systematizing expert interview that is also related to the exploratory expert interview (Bogner & Menz, 2009). This type of expert interview aims at the structured and comprehensive collection of expert knowledge in order to achieve a high level of data comparability (Gläser & Laudel, 2004). Both types are characterized by their focus on technical and processual knowledge. Technical knowledge relates to highly specific knowledge of a field, for instance, on technical applications, information, or data. Bogner and Menz (2009) contrast this to everyday knowledge, describing it as educational knowledge and classifying it as the specific knowledge advantage of experts. In contrast, process knowledge captures knowledge that is based on practical experience and the institutional context of actions. In this sense, process knowledge emerges due to the position of the person in a process and comprises knowledge about interactions, routines, or social practices (Van Audenhove & Donders, 2019). The third type defined by Bogner and Menz (2009) is the theory-generating expert interview that serves as a starting point for the envisaged methodological development of problem-centred expert interviews and is explained in the following.

The theory-generating expert interview

According to Meuser and Nagel (1991), experts can be defined as persons who are responsible for the development, implementation, or control of a solution, or persons who have privileged access to people or decision-making processes. The theory-generating expert interview draws upon these considerations and describes experts as persons with specific knowledge who hold a certain status or exercise a function in decision-making processes in a particular field of action (Bogner & Menz, 2009, 2018). Thus, ‘their action orientations, knowledge and assessments decisively structure, or help to structure, conditions of actions of other actors’ (Bogner & Menz, 2009, p. 54). Following this definition, expert knowledge has a socially relevant dimension as it practically shapes and determines a field of action.

The theory-generating expert interview holds an analytical and interpretative perspective in order to outline interrelations in the empirical data and to develop theoretical approaches. It stresses inductive theory development based on empirical data and thereby aims at revealing interpretative knowledge (Bogner & Menz, 2009, 2018), which is defined as subjective relevancies, viewpoints, or perspectives on which experts draw when enforcing their orientations. Interpretative knowledge is predominantly implicit and comprises, for example, decision-making practices or action orientations. Interpretative knowledge does not exist a priori, but is developed through the ‘act of abstraction and systematization’ of qualitative interview data (Bogner & Menz, 2009, p. 53). Only during analysis does it become evident whether a statement of a respondent has to be interpreted in the professional or private context. Therefore, Bogner and Menz (2009, p. 53) argue for integrating the individual methodologically and highlighting the importance of personal orientations and perceptions, in order to enable a rich investigation of interpretative knowledge. However, in a more recent publication, the authors accentuate that experts are not addressed as ‘private persons’ in the first instance, but rather as representatives of a specific group, e.g., public officials or functionaries (Bogner et al., 2014; Mergel et al., 2019).

The authors do not only debate the definition of experts, but also discuss the interaction structure in expert interviews. They focus on the potential roles and strategies of the interviewers that are shaped by the expectations and perceptions of the interviewees. The interviewer can be seen, for instance, as a co-expert when the expert assumes that the interviewer is familiar with the field and its practical conditions. In contrast, other types are characterized by a more asymmetrical interaction, for instance, when the interviewer is considered a layperson who is naïve about the field of research. On the one hand, this can lead to a broad access of information, as the interviewee is encouraged to explicate the answer. On the other hand, the expert might address issues and provide explanations that are already well known. Furthermore, the researcher might be identified as a potential critic, for instance, because of his or her institutional, organizational, or scientific background. The interviewee may feel criticized by the questions of the interviewer and may doubt his or her objectivity, which can lead to a limited willingness to answer questions (Bogner & Menz, 2009; Van Audenhove & Donders, 2019).

According to Bogner and Menz (2009), the theory-generating expert interview is not linked to a specific interview design or technique. Instead, the authors propose selecting the interview design in accordance with the actual research needs and remaining flexible during the research process. However, they suggest making use of an interview guide that allows for open, but thematically structured interviewing.

The problem-centred interview

The second interview method that is presented here is the problem-centred interview (PCI) that was originally developed by Witzel (1982, 2000). It is a widespread qualitative research method that has been internationally received since its English-language publication ‘The Problem-Centred Interview: Principles and Practice’ in 2012 (Murray, 2016; Shirani, 2015) and has been frequently used in different research disciplines, such as sociology (Vogl et al., 2019), political science (Reiter, 2014) or pedagogics (Leder, 2018). PCI is a qualitative face-to-face interview method that draws upon central principles of qualitative research such as openness, flexibility, and process orientation (Flick, 2009). It is meant to be an egalitarian dialogue between the interviewer and interviewee in which the research question or the ‘problem’ is refined jointly (Murray, 2016). The approach combines two seemingly contradictory sources of knowledge, as it gives equal right to the previously accumulated theoretical and empirical knowledge of the researcher and to the individual knowledge and personal experiences of the respondent. Witzel and Reiter (2012) argue for an iterative process, combining inductive and deductive reasoning, when applying PCI. Based on a sensitising theoretical framework, the researcher should actively encourage participants to tell their stories. By doing so, the method additionally accommodates the preconditions that shape the interviewees’ orientations and opinions (Scheibelhofer, 2008).

PCI presupposes a specific research design and tools¹ for conducting interviews. In general, it draws upon an interview guide that comprises a narrative beginning in the first phase and precise follow-up questions in the second. Its purpose is not to establish a rigid question–answer scheme, but rather to support the researcher with a thematic framework that serves as an orientation to the premeditated research questions. PCI thus facilitates a systematized comparison of the collected data. The interviewer starts with an open-ended question in order to stimulate a narration structured by the individual’s concerns (Scheibelhofer, 2008; Witzel & Reiter, 2012). After the interviewee had finished the narrative episode, the interviewer moves on to general and specific explorations as well as ad hoc questions. Through general explorations, researchers aim to enquire after aspects and details that are of potential research interest but have not been mentioned by the interviewee yet. On the other hand, specific explorations refer directly to the interviewees’ account in order to gain further insights into the interviewees’ opinions. These questions and statements are based on knowledge that was gained beforehand or that emerges during the narrative part. They thus follow a predominantly deductive logic. Witzel and Reiter (2012) distinguish between three specific exploration strategies: mirroring, which strives for cognitive structuring and validation, comprehensive questions, which aim at clarifying meanings or interrelations, and confrontations, addressing the specification of contradictory statements by the interviewees. Additionally, the authors propose making use of ad hoc questions, which comprise additional aspects or keywords, ensuring the comparability of the interviews. As these questions may disrupt the dialogue in terms of content, the authors suggest asking them at the end of the interview. Combining an open narrative beginning with a more structured interview section enables the researcher to remain receptive towards the gathered information, as it encourages the interviewees to describe individual perspectives, to clarify previous statements, and to revise misguided assumptions of the interviewer.

The dialogic-discursive structure of the PCI thus emphasizes the situational co-production of social reality by the interviewer and interviewee in the context of the interview situation. This assumption contrasts PCI with the biographic-narrative interview (Rosenthal & Loch, 2002; Schütze, 1983) that also starts with a narrative section, however presuming that the structure of narratives represents the structure of social reality. These premises are reflected in the communication structure of both approaches: While in narrative interviews the interviewer primarily acts as engaged listener directing the interviewees’ narration as little as possible, in PCI the interviewer gradually changes from listening to intervening in order to guide the interviewee on thematic aspects (Mey, 2000; Scheibelhofer, 2005). By stressing the interviewee’s perspectives, but also enabling the researcher to address specific topics, PCI provides appropriate approaches for reconstructing the implicit dimensions of expert knowledge considering a certain research focus.

The problem-centred expert interview

The methodological and epistemological backgrounds of theory-generating expert interviews and PCI will now be discussed in order to explain the analytical value of combining the two methods. Both methods are based on similar methodological and epistemological premises. Both the theory-generating interview and PCI highlight the perspectives and opinions of the interviewee and aim at formulating new theories by systematizing and interpreting individual statements. Thereby, the objective of the theory-generating expert interview is the ‘communicative opening up and analytic reconstruction of the subjective dimension of expert knowledge’ (Bogner & Menz, 2009, p. 48). The authors emphasize inductive theory building by an interpretative generalization of the qualitative data. In turn, the epistemological logic of PCI argues for developing theories by intertwining deductive and inductive steps when conducting and analysing the interviews. Deductive elements can be found in the interview guide that draws on theoretical concepts and heuristics. On the other hand, PCI follows inductive logics when describing the interviewer as a ‘well-informed traveller’ that stays open-minded to the interviewees’ perspectives and willing to reformulate his or her previous assumptions (Murray, 2016). Although the theory-generating interview argues primarily for an inductive procedure, Bogner and Menz (2009) include the option of combining inductive and deductive elements as was suggested for PCI. Against this background, both methods can be connected to the methodology of Grounded Theory (GT) aiming at (middle-range) theory building based on qualitative interview data (Glaser & Strauss, 2017; Strauss & Corbin, 1997).

Witzel and Reiter (2012, p. 21) refer to the possibility of applying PCI in the context of expert interviews. According to the authors, ‘the PCI corresponds perfectly with the interest in investigating the interpretive dimension of expert knowledge […] as long as it considers the specific role of the respondent in the conversation.’ Conversely, Bogner and Menz (2009, p. 74) mention PCI and state that the two interview methods focus on different epistemological interests. They emphasize that ‘the difference of the procedures lies in the role of the interviewee in the conversation, which is determined by specific epistemological interests and so is more a matter of practical research requirements than of criteria related to method’. The combination of these epistemological perspectives serves as a promising starting point for moving beyond the experts’ role as representatives and taking into account their personal opinions and experiences. As expert interviews represent a very popular research method in various disciplines, a problem-centred approach offers potential for many kinds of research, since it is interested in the implicit dimensions of expert knowledge. The following table outlines selected elements of the theory-generating expert interview and the PCI, which are central for defining the methodological and epistemological background of the problem-centred expert interviews (see Table 1).

Table 1. Elements of the problem-centred expert interview

Theory-generating expert interview	Problem-centred interview (PCI)
Defines and discusses the term ‘expert’	Highlights the individual perspective
Distinguishes different types of expert knowledge	Provides a specific interview design and set of questions
Aims at inductive theory development	Enables comparability of gathered data
	Proposes inductive-deductive theory building

Revealing the interpretative knowledge in expert interviews is a quite elusive research target. By combing theory-generating expert interviews and PCI, we might move a step closer to this objective, since both methods offer useful approaches and instruments for moving beyond explicit expert knowledge. While the theory-generating expert interview contributes to the analysis of the social relevance of implicit expert knowledge, PCI highlights individual perspectives by providing an interview technique that encourages the interviewees to unfold their personal relevancies and perceptions.

In the following section, an empirical example is discussed in order to show the practical application of the problem-centred expert interview. The insights are based on my PhD research in human geography, which focuses on key agents in regional economic development and their influence on socio-spatial change.

An empirical example of the problem-centred expert interview

The idea for combing theory-generating expert interviews and PCI developed from the research interest of my PhD thesis. My thesis focuses on outstanding, entrepreneurial individuals from politics, economy, and administration, but also from civil society, so-called key agents, who take up an exceptional role in socio-spatial change processes in regional economic development.² The overall aim of my research is to develop a local theory in order to contribute to the nascent conceptual discussion of key agents in human geography. Exploring the individual agency of key agents in socio-spatial change processes requests for an in-depth understanding of the social relevancies of involved actors in regional decision-making processes. In order to identify key agents and to understand how and to what extent key agents can influence regional development processes, it seemed necessary to look beyond explicit expert knowledge and to gain understanding of the personal perceptions, orientations, and thoughts that lead to their exceptional influence.

Theorizing key agents

Key agents and their pivotal roles in decision-making and change processes are widely discussed in several disciplines, for instance, in political sciences, sociology, organizational and management studies and, only recently, in human and economic geography. Against the background of the structure-agency debate (Giddens, 1984), several conceptual strands focus on key agents (or agents of change), including, for example, leadership (Normann et al., 2017; Sotarauta, 2016), political entrepreneurship (Battilana et al., 2009; Mintrom & Norman, 2009), and brokerage (Christopoulos & Ingold, 2011; Leick & Gretzinger, 2018).

What connects these various types of theoretical concepts is the fact that they offer an analytical perspective on individuals or groups of individuals that are deemed to make ‘the’ difference within policy, institutional, or organizational change processes. They manage to influence decision-making processes decisively, by having special characteristics, capabilities (e.g., persistence, creativity, persuasiveness), and knowledge (Petridou et al., 2015). While experts are characterized as having the knowledge and the position ‘to structure a particular field of social action in a meaningful way’ (Bogner et al., 2018, p. 655), key agents take an ‘outstanding’ structural and social position that allows their opinions to be heard over those of others. This does not imply that their power is necessarily tied to a professional or official position in decision-making processes. Key agents might also be non-state actors drawing on private resources, ideas, and targets (e.g., engaged persons from a civil movement). They often behave rather subversively in shaping and transforming fields of action beyond established decision-making structures and formal procedures (Battilana et al., 2009; Beer, 2014). Based on these theoretical foundations, it can be concluded that every key agent can be defined as an expert, but not every expert that is interviewed can be identified as a key agent.³

Although it is widely acknowledged that key agents and informal processes in decision-making processes are difficult to grasp empirically (Ayres, 2017), the methodological approach seems to have played a subordinate role in literature dealing with key agents. The studies mention methods, such as (semi-)structured interviews or quantitative surveys, rather incidentally. However, lately a growing number of authors dealing with key agents claim to ‘zoom in also on the “subjective” stories of individuals, and grasp their perceptions, intentions, and change strategies’ (Grillitsch & Sotarauta, 2019, p. 14) which drive their actions.

In order to understand and to conceptualize the emergence and action of key agents, I choose to draw upon the theory-generating expert interview that points to the implicit dimension of expert knowledge as a starting point for theory building. However, the theory-generating expert interview does not provide a specific research design that supports the exploration of these knowledge dimensions. The interview design of PCI, however, enables researchers to approach these implicit dimensions of expert knowledge by accentuating the biographical experiences and personal opinions of interviewees. I now wish to reflect upon and discuss the practical application of this methodological combination.

Research design and theoretical sampling

The research project is based on a comparative case study design contrasting socio-spatial change processes and their key agents in two Austrian small-sized cities. The research design is based on the principles of theoretical sampling (Strauss & Corbin, 1997) including different iterative sampling strategies. In the beginning of my research, combined different methodical approaches compromising a media and documentary analysis and explorative expert interviews (e.g., with representatives of federal state government or regional agencies) (Bogner & Menz, 2009; Kaiser, 2014) in order to explore interesting socio-spatial change processes in regional development and to select the first case study. In the next step, a local ‘gatekeeper’ was contacted in order to acquire further contextual information about ongoing projects and to get access to the local actor network of the first city.

However, in this paper, I want to focus on the problem-centred expert interviews conducted during my fieldwork with experts in one of the small-sized cities. Following the definition of the term ‘expert’ by Bogner and Menz (2009, p. 54), the sample should include persons, who ‘decisively structure, or help to structure’ the social relevancies in a field of action by their action orientation and knowledge. Hence, the interviewees can be distinguished from ‘laypersons’, because they have a specific knowledge that enables them to shape the regional development process, for instance, as initiators, supporters, or preventers, albeit to different degrees (Bogner & Menz, 2018). Initially, the experts were successively recruited by a snowball sampling technique, based on statements of preceding interview partners. The experts provided information about those who were involved in the process and who could provide further insights about it. This was a crucial procedure which was necessary to understand the inner logics of the decision-making process. After the first coding sessions, I additionally decided to contact persons who had been mentioned by the interviewees rather incidentally or who were known from regional media reports, for instance, because they previously had been engaged in the process. This was done to avoid a biased perspective on the process which could arise by talking only to a specific group of experts who might share a collective view (Bogner & Menz, 2018, p. 259). Finally, the interview sample included persons that have or have had responsibilities in decision-making and having specific knowledge that structures the process. It also included the key agents themselves. The sample thus comprised interviewees with different institutional and organisational backgrounds in regional development, e.g., politicians, entrepreneurs, regional development officers, but also involved and engaged citizens, such as a tax consultant, a retired farmer, or an investor.

I repeated this procedure for the second case study. Striving towards local theory-building, I chose to conduct the second part of the fieldwork in a city with similar socio-spatial preconditions and challenges to those of the first. This enabled me to focus on the individual agency of key agents and to contrast their influence in decision-making processes.

Conducting problem-centred expert interviews

Before going into the field, I developed a heuristic framework pointing out to re-search gaps concerning key agents and individual agency in regional development. This ‘sensitising concept’ (Glaser & Strauss, 2017) provided a broad and open perspective on actor constellations and interactions modes in regional development and served as basis for developing the topical guide. Drawing upon the principles of PCI, the interview included an open-ended question in the beginning and follow-up questions, which allow to introduce thematic aspects in the interview course in a flexible way.

Witzel (2000) proposes beginning the PCI with an opening question that will place the focus on the matter of interest and stimulate the participant to relate his or her individual perspective on this issue and to give an account of according experiences. Guided by my research interest, I adapted the interview design presupposed by PCI and started with an individual opening question in order to generate contextual information about each expert’s background. This was done to get an idea of the individual’s formal and/or social position and his or her tasks and responsibilities in the process. This question was particularly important with reference to experts of whom I could only collect vague information beforehand.

In the next step, I continued the problem-centred expert interviews with an open-ended question that invited the interviewees to describe the regional change process from the beginning until the current day. The objective of this question was to prompt the interviewees to reconstruct the economic development process and to mention the roles of process-associated persons by telling a coherent story. In order to support the narrative structure and the comparability of the narrations, a temporal specification was included. Guided by my research interest, I was particularly interested in the interviewees’ perspectives on actor constellations and (in-)formal governance processes that drive the changes in regional development. In order to avoid influence, I phrased the question in an indirect and open way (Meuser & Nagel, 2009):

IV: ‘Could you please tell me how the process […] emerged from the early beginning (or the first idea) until today, and who was involved?’

IP: ‘The topic has a long history, about twenty years ago, when there has been a loss of skilled workers […] Back then, we started with developing a job fair, which was not always a common way […].’

While the narrative beginning at first appeared unconventional to some experts, in most cases it gave impetus to a narrative episode, encouraging the experts to reconstruct the process based on personal explications of actor relations and interactions. By this narrative impetus, the respondents were triggered to include the role of other people (and possible interview partners) in a quite natural and vivid way. As interviewees told their stories chronologically, it was possible to consider the temporal dimension of the process and to retrace the point of appearance of actors. Based on the data thus generated, I obtained first evidence of those who considered themselves to have taken a central position or who was considered to be an important part of the process. This helped me to contact further experts: The interview partners mentioned whom they considered as essential for the process in an explicit way (‘It is the economy, that is to say, Tom Winter, who drives this process’) or in a more implicit way:

‘[…] However, we could not find a national programme that fitted our regional demand and then we had a meeting and the mayor and Tom [entrepreneur] said that they had an idea for establishing a regional innovation hub […]’.

After the experts had finished their narrative sequence, I continued with open-ended questions comprising general and specific explorations. Starting with the technique of general explorations that focused on selected episodes of the narrative section, the objective was to learn more about the interviewee’s position and his or her relationship with and connection to other important actors within the process. The questions aimed at gaining more information about the expert’s involvement or previously mentioned persons involved in the process:

‘Earlier, you mentioned that the other entrepreneurs had been sceptical about your initiative. How did you handle this?’

‘You mentioned Mr. Fisher in the beginning. Can you tell me more about how he had been involved in the implementation of this project?’

The latter question requires proceeding cautiously since asking about ‘third persons’ (and potential interview partners) can be a sensitive issue in the communicative situation, demanding a mutual basis of trust. When interviewers are supposed to be external critics (Bogner & Menz, 2009), talking about other persons could be particularly awkward.

During the interview situation, I also drew upon specific explorations involving comprehensive questions and mirroring.⁴ These techniques facilitate a deeper understanding of interviewees’ statements by yielding to alternative perspectives on previous reports (Mey, 2000; Witzel & Reiter, 2012). Comprehensive questions were used to clarify vague statements and invited the interview partners to explicate their point of view, for instance, about interaction modes between different (groups of) people:

‘So, would you say that entrepreneurs used to interact more competitively about 20 years ago than they do today?’

On the other hand, the technique of mirroring proved to be particularly useful for testing ad hoc hypotheses or hypotheses derived from previous interviews:

IV: ‘So, it seems as if politics and the economy work by common agreement to handle the regional challenges.

IP: ‘The coordination works out very well, in fact. […] However, for me as an entrepreneur it is very important that we as entrepreneurs do not want to be manipulated politically.

The ad hoc questions at the end of the interviews included evaluation questions (Bogner et al., 2014), for instance, asking the experts about their assessment concerning further challenges to the process.

Although the presented type of questions suggested by Witzel (2000), cannot always be clearly separated from each other in practice, the conceptual distinction enables the researcher to draw upon a variety of questioning techniques, including narrative-generating, implicit, or concluding questions and statements. This allows for a flexible and varied application of questions during the interview situation and an interactive creation of interpretative knowledge.

Analysing problem-centred expert interviews

Expert interviews that aim at revealing technical and processual knowledge would usually apply procedures of systematic textual analysis, such as the qualitative content analysis (Mayring, 2000), which aims at reducing and summarizing evident interview data according to predefined research objectives. The theory-generating expert interview however tends rather to follow an interpretative approach that aims at inductive theory building based on the data that was gathered. Its conceptualization builds on considerations of Glaser and Strauss (2017), proposing inductive theory formulation. Similarly, PCI mentions an inductive-deductive approach following the principle of theoretical sampling and aiming at generating theories. Drawing upon on these considerations, my research was oriented towards the principles of GT. Below, I want to focus on the process of open and axial coding to illustrate the analytical value of the data generated by the problem-centred expert interviews. Findings were developed gradually and empirically grounded in a problem-centred dialogue between the research interests on the one side and inductive reasoning of the empirical material on the other.

The open coding process starts during fieldwork and designates the breaking down, examining, and labelling of the data with codes. This is done by asking conditional questions relating to the material, which serves as foundation for further interpretation (Belgrave & Seide, 2018). By scrutinizing the interpretative dimension of expert knowledge, I aimed at gaining a deeper understanding of the experts’ orientations and unveiling why, how and to what extent experts shape the process and respectively might be acknowledged as key agents. As individual perceptions, motifs, and aims guide the interactions of experts, these can be considered as implicit decision-making maxims in a field of action (Bogner & Menz, 2018). The following example illustrates the coding process of implicit expert knowledge (see Table 2). In this context, it must be highlighted, that the distinction between different kinds of expert knowledge represents a result of an analytical process and not a characteristic of the knowledge itself (Bogner & Menz, 2018). Depending on the research focus, the statement could also have been interpreted, for instance, in terms of process knowledge. In order to divulge the implicit dimensions, I coded the sequences line-by-line.

Table 2. Example of coding interpretative knowledge.

Citation

Open coding Interpretative Knowledge

‘We are a network of three people [mayor, entrepreneur, regional chairman]. […] We have not been defining our responsibilities, such as, you have to do that, you have to do this, or something like this, it emerges when we meet each other. We truly regularly meet each other. Sometimes at 7.00 o’clock in the morning, then we drink a coffee in a cafe in the city and then again, we define the next steps.’ [Interview Mayor]

informality and exclusivity of group, self-consciousness of being part of this network recognizes (regional) challenges, mutual trust, equality of members, dissolving hierarchies? need of ‘spatial and cognitive proximity’, continuity, (need of) persistence, voluntary engagement beyond ‘official’ labour time → high personal motivation? meetings are in public and are not hidden → convinced about ‘doing the right’? ‘hands-on’ attitude, taking responsibility

By focusing on interpretative knowledge, one can reveal those relevancies that guide the actions of the expert and structure social practices in the regional development process. The selected statement of the mayor (who could have been identified as a key agent in further research course) shows that he recognized a regional pressure to act. In order to tackle the challenges, he draws upon a flexible and informal, but exclusive group of people based on trust and persistence. The interpretative knowledge proves to be as essential in the further research course in order to explain why some experts emerge as key agents who are able and willing to change or renew a field of action. Thereby, the experts’ actions and experiences must be interpreted against the background of the contextual conditions that influence their actions (Witzel & Reiter, 2012). When talking to experts, this implies consider information about the organisational or institutional background of the interviewees and the (in-)formal procedure of decision- making. The prior documentary analysis or the explicit process knowledge offered during the interviews enables to contextualize the statements in terms of the professional sphere. On the other hand, the statements about personal biographical experiences (that were primarily mentioned in the narrative sequence) turned out to be a very useful source for explaining the experts’ motives, resources and intention.

In the quest of developing a local theory of key agents and their role in regional economic change, I went across the interviews in order to explore similarities and differences and to develop linkages between the data. I contrasted the codes and initial categories to see how the individuals’ orientations and perceptions differ in terms of the regional development process and to work out characteristics and behaviour that might be attributed to key agents. This step is orientated towards the technique of axial coding by Glaser and Strauss (2017). Additionally, I was also able to directly relate the statements of the experts to each other and to interpret the different perspectives on the decision-making process, as the experts talked about the same events. By doing so, I could ascertain, for instance, whether interactions between individuals are perceived in a unilateral or reciprocal way and whether experts share perceptions about forms of interaction. Thus, I could not only merge the reconstructed experiences, but also sketch out the position of the experts in the network. As experts represent ‘a complex interdependence of knowledge and power’ (Bogner & Menz, 2018: 656), it is challenging to reconstruct their position in a field of action. In my research, the experts’ position and the power they exert is less determined by formal authority by virtue of their professional positions but constructed through social interactions (Smith, 2006). By gaining an understanding of the experts’ intentions and interactions, in regional economic development, I could grasp how and to what extent experts structure decision-making processes and gain further evidence of potential key agents.

In the last step, I contrasted the processes and involved actors of the first case study with the interpretations of the second case study in order to strive for a local theory of key agents in regional economic change processes. It thus, became evident, that key agents in regional development accumulate and exert a high degree of relational power, enabling them to increase their individual agency. To reach their targets they rearrange governance arrangements by developing horizontal and vertical axes and altering interaction modes.

Pitfalls and potentials of the problem-centred expert interview

Based on my experience with conducting and analysing problem-centred expert interviews, I want to discuss the pitfalls and potentials of merging the theory-generating expert interview and PCI. Although investigating key agents represents a specific case when conducting problem-centred expert interviews, the lessons learned are of interest for all expert interviews moving beyond the explicit dimensions of expert knowledge.

First, this ambitious style of interviewing may give rise to a number of pitfalls in the interview situation. This might be the case, for example, when the shift from a rather open and narrative style to a more structured interview style becomes complicated or experts grant less time for the interview than was announced beforehand. Switching between different interview styles generally is a thorny situation in PCI, as the ‘communicative roles are not consistent’ (Scheibelhofer, 2008, p. 408). In terms of expert interviews, the respondents might be experienced in talking and presenting their concerns and interests. Some experts offered wide-ranging explications in the narrative phase but made less time available for the interview than was announced beforehand. In this case the interviewer must consider whether to focus on the narrative accounts or to bring the explications to a close and switch from the time-consuming narrative part to the follow-up questions. Before moving into the field, researchers should reflect upon such interviewing priorities in order to be prepared when the interview course does not proceed as intended.

Second, PCI supports the capturing of the implicit dimensions of expert knowledge and its influence in decision-making by reconstructing individual perspectives through the specific mix of questioning techniques. Interviewer and interviewee leave the question-answer scheme (Witzel & Reiter, 2012, p. 20) and the experts are encouraged to recount societal processes. In my interviews, it could be observed that the initial narrative component stimulates the experts not only to describe, but also to reflect upon the process and to include personal relevancies and explanations. Additionally, and in contrast to narrative approaches (Rosenthal & Loch, 2002; Schütze, 1983), the discursive-dialogic procedure of the PCI enables the interviewer to focus on specific aspects of a phenomenon through follow-up questions. These questions allow for an alternative view in order to deepen aspects mentioned by the interviewees or to examine ad-hoc hypotheses. The above-mentioned techniques of explorations, for instance, proved very helpful to elicit personal opinions when experts who were experienced in being interviewed tended to rely upon familiar ‘standard stories’. This specific interview setting offers new perspectives on the dialogic creation of implicit expert knowledge and enables us to reconstruct action orientations and patterns regarding not only professional, but also personal experiences. The problem-centred interview thus provides a useful approach to generate implicit and tacit expert knowledge based on the interview material.

Third, the methodical combination of problem-centred expert interviews contributes to a deeper understanding of a social field of action and thereby supports theory-building. The topic-oriented interview design of the PCI facilitates to compare and to relate interview statements about specific aspects and to investigate the social relevance and political effectiveness of the implicit dimensions of expert knowledge. By contrasting and bridging the reconstructions of the individual perspectives, a ‘multifaceted portrayal’ of local decision-making processes and actor relations emerges. As a result of this, one can detect the inner logics and patterns of decision-making processes and reconstruct the position of an expert in a field of action. Bringing together the analytical knowledge conceptualization of theory-generating expert interviews (Bogner & Menz, 2009) and the interview design of PCI (Witzel, 2000) thus constitutes a promising methodological approach for (local) theory-building and developing typologies.

Conclusion

The reason for combining theory-generating expert interviews and PCI was my specific interest in key agents, which could only be insufficiently addressed by existing interview approaches. In my inquiry, the methodological combination of problem-centred interviews proved itself to offer a fruitful approach for investigating the implicit dimensions of expert knowledge and integrating the experts’ professional as well as personal experiences in the search for explanations. The theory-generating expert interview takes into account the specificity of interviewing experts and the complexity of expert knowledge, whereas the PCI highlights the individual perspective of interviewees and provides a supporting interview design combining an open narrative beginning with a topical guide. Since the two interview approaches show similar methodological premises, it offers the possibility of an inductive, code-based, and theory-generating analysis in accordance with the principles of GT.

Despite its challenges mentioned above, the problem-centred expert interview represents a promising approach for investigating and analysing the increased complexity and informality of decision-making processes (Kaiser, 2014). The combination addresses a constructivist perspective, understanding expert knowledge not only as exclusive knowledge limited to the insights of science or disciplines, but also as practical, local knowledge emerging from professional as well as private experiences (Meuser & Nagel, 2009). Against this background, the methodical combination presented above is recommended whenever the individual agency of experts forms part of an investigation. In this context, it might be of interest in a variety of research fields such as policy analysis, organizational and management studies, or applied social research.

Acknowledgments

I would like to thank Hanneke Friedl, Peter Görgl and two anonymous reviewers for helpful comments on this paper.

Disclosure statement

There are no potential conflicts of interest.

Additional information

Funding

The author received no specific funding for this work.

Notes on contributors

Stefanie Döringer

Stefanie Döringer is a PhD candidate at the Austrian Academy of Sciences, Institute for Urban and Regional Research in Vienna and holds a Bachelor’s Degree in Sociology and a Master’s Degree in Spatial Research and Spatial Planning. She also is an academic associate at the University of Vienna in projects dealing with spatial research. Currently, her research interests lie in local and regional governance, key agents in rural areas, and qualitative interview methods. In her interdisciplinary PhD thesis, she focuses on the agency of entrepreneurial individuals in regional development and their role for socio-spatial change.

Notes

1. Additional to the interview guide, it is proposed that a short questionnaire, recording devices, and a postscript also be used in PCI in order to support the interview procedure.

2. Although scholars increasingly highlight the role of individuals for initiating socio-spatial change processes in regional development, there is still little conceptual knowledge concerning this phenomenon in governance literature, as the agency of individuals has been overshadowed by the focus on institutional and organisational structures.

3. In this context, it must be stated that the definition of the term ‘expert’ as it was proposed, for instance, by Meuser and Nagel (2009) is derived from the sociology of knowledge discussing methodological implications. The term ‘key agent’, however, summarizes a variety of theoretical concepts and approaches from different strands of literature (e.g., sociology, management studies and political science), focusing on different strategies and outcomes of key agents.

4. Confrontations, as suggested by Witzel (2000), were not applied in the expert interviews, since I considered targeted questions and comprehensive questions sufficient. Additionally, such strategies might have compromised the course of conversation, as shifting from an open narrative beginning to confrontative questions can also result in a defensive attitude from respondents (Scheibelhofer, 2008).