The Problem-Centred Interview: Principles and Practice

Introduction In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 1-11 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Introduction People tend to gain and communicate much of their knowledge by means of language. For instance, if we want to know something about the latest play at our favourite theatre, we may hear about it on local TV or radio, we can read about it in local newspapers, or we can ask friends that have been there and they may give us their opinion. The fact that asking somebody else is a common way of obtaining knowledge about something is the reason for the unbroken relevance of ‘interviewing’. As Brinkmann (2008: 471) writes: ‘We can presuppose that humans have interviewed each other in some form or other for as long as they have mastered the use of language.’ In other words, qualitative research interviewing makes use of the ancient human habit of asking and answering questions. It is a well-tried and reliable way of finding out about things and about each other in conversations. Yet there are different ways of doing it. Since the first half of the twentieth century, when interviewing entered the more systematic discussion about social research methods, as Platt (2001) shows in her historical review of the status of interviewing in social research in the USA, many different techniques of interviewing have been developed. They all make specific suggestions about how to collect and construct knowledge. For instance, going back to the scenario about the play, it makes a difference whether we simply invite our friends to tell us how it was, or whether we confront them instead with a list of specific questions regarding the length of the interpretation, the number of people who left during the first break, the availability of beer, or the temperature in the theatre. While such specific questions may be important for us to decide whether we want to attend the performance, they may not help us find out what our friends actually thought about the play. Kvale and Brinkmann (2009: 48–50) introduce two metaphors for describing and contrasting these two ways of obtaining knowledge. They distinguish between the interviewer as a miner and a traveller. The miner-

interviewer has a targeted and well-defined interest in specific informations she considers valuable: she knows what to look for and turns this (re)search into a collection of pure ‘nuggets of knowledge’ (ibid.: 48). Afterwards she will decide what it is actually worth. Alternatively, the traveller-interviewer is openly curious. She ‘wanders through the landscape’ (ibid.) of the area under investigation, involves herself in conversations, and encourages people to tell her about their experiences. With each conversation she may discover new aspects, and develop and modify her opinion. Together with the respondents encountered, she interactively co-constructs the knowledge that she will take home. Both approaches produce useful knowledge but also involve certain disadvantages. On the one hand, chances are low that the miner, who corresponds to a systematic collector of scientific knowledge according to pre-defined standards, will be open to changing her assessment criteria and reflect upon the concept of what (using the mining metaphor) ‘precious metal’ is. On the other hand, it may be difficult for the traveller to come to the end of her journey; she may be overwhelmed by the many new impressions and perspectives she has encountered, and may even forget about her original interest. Metaphors are a useful way of imagining interview research (Alvesson, 2011). This book is about a third Page 2 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

way of collecting knowledge - by involving people more actively into a process of knowledge constitution. In the context of the problem-centred interview (PCI), interviewers take the role and attitude of a well-informed

traveller: they have certain priorities and expectations and start the journey on the basis of background information obtained beforehand. Yet the trip they will finally make, and the story they will tell about it afterwards, depend on the people they meet on the road and on their insider knowledge. By talking to them they are able to refine their assessment of the major sights mentioned in the travel guide. Their guidebook only helped them to outline a preliminary roadmap and frame of reference that remains open to modification and revision on the basis of conversations with the locals. It is through these conversations that they get a better idea about what is relevant and worth seeing. Box 1.1 illustrates this interaction.

Box 1.1: A Short Story About Well-Informed Travelling Imagine the following scenario. You travel to a small mediaeval town in the south of France that is highly recommended in your travel guide. You get up early and visit a small local café for breakfast. One of the locals, a retired teacher, notices you studying a travel guide of his home town and involves you in a conversation. You tell him that you are here to get an impression of how town life may have been organised a few hundred years ago, and what is still left of it. You refer to your travel guide and other preparations you made in advance, and explain that you want to start your exploration at the main square in front of the church. From there, you would march along the bank of the river and then along the remains of the city wall to the old market square and the city hall. The teacher is surprised about your plans and offers to take you on an alternative tour that starts at an old tree outside the former city wall where the main gate used to be. You find out that the story of the old tree and the Roman grave that its shadow covers is crucial for how the town's main roads were organised towards the old market square in a way that they crossed the river where it was narrowest. The old tree also used to be the place where, over hundreds of years, many convicted criminals and ‘witches’ were executed - it is in walking distance to the church, whose crypt served as dungeons. Even today, the significance of the old tree can still be seen. For instance, every festival begins with a small parade starting at the old tree and proceeding to the market square where it is welcomed by the mayor from the balcony of the city hall; agreements with neighbouring towns are symbolically signed under the old tree; and it is popular among local youths who carve hearts and oaths of love into its trunk, although this has long been forbidden by the authorities. With an apologetic smile and putting his arm around the tree like an old friend, the teacher confesses to having done his fair share of carving on the tree. And sometimes he even visits the place with his wife or his grandchildren to talk about the old times and to

Page 3 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

search for fading traces of his very own marks. After three intensive hours of talking to the former teacher, asking him about details and being guided through the town, you understand that the church, river banks, city walls, market square and city hall are all relevant sights and worthwhile exploring, just as your travel guide suggested (it does not mention the multi-purpose function of the crypt though). Yet the way they are arranged can only be understood in the historical perspective represented by the old tree and reproduced in contemporary local habits. With its much broader orientation towards relevant sights, the travel guide does not mention the old tree and its importance in the arrangement of these sights. For that, local knowledge is necessary. As a way of research, well-informed travelling is not about drifting through a (social) space of knowledge and meaning. On the contrary, it requires a lot of preparation in terms of both substance and behaviour. Think about the example above (Box 1.1): local cafés are good places for an ‘explorative’ breakfast as they are usually places where locals hang out; the opportunity to talk to a teacher is gladly taken as teachers are famously knowledgeable and ready to communicate their knowledge; the hunch that the arrangement of significant buildings usually follows certain standards is taken from the travel guide and academic books about urban planning in the middle ages; and a guided tour by a local insider is the perfect way of investigating the contemporary significance of mediaeval urban planning and its practical implications. The search process for this research was purposefully designed and at least the substantive preparation and some of the initial steps were also done on purpose. The technique of the PCI that translates this idea of well-informed travelling into a methodological and practical programme of research could be a good approach for you if: • you identified interviewing as the appropriate way of collecting information regarding a certain issue; • the issue refers to a research question regarding the what, how and why of actions, appraisals and opinions; • you have an idea about people who could provide you with first-hand insights into this topic; and • your interview partners are willing to allow you to collect their extensive knowledge in order to understand their perspective in as much detail as possible. In terms of a preliminary definition, the PCI can be described as a qualitative, discursive-dialogic method of

reconstructing knowledge about relevant problems. This definition involves a few peculiarities. The discursivedialogic character is outlined above in the idea of well-informed travelling and the involvement of interviewers and their knowledge in a dialogue with respondents and their perspectives. The discussion of this dialogue as an epistemological challenge (a task of obtaining knowledge about something) is done in Chapter 2 together with a reflection of the specificity of social (scientific) knowledge. Suggesting a particular way of reconstructing the meaningfulness of this knowledge through interviews, which is the purpose of the whole book, defines Page 4 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

the PCI as a method: it is a ‘stylized way of conducting research that comprise(s) routine and accepted procedures for doing the rigorous side of science’ (see Abbott, 2004: 13). Like other methods of qualitative research, the PCI involves an ‘exchange between real people’ in their own ‘social, cultural, and physical context’. It focuses on meanings and behaviour, which the researcher tries to understand ‘through the eyes and lived experience of the people’ (Schensul, 2008: 521–2).

But What Are the ‘Problems’ in Problem-Centred Interviews? First of all, PCIs do not necessarily deal with issues that are ‘problematic’. That would be a misunderstanding of the term. The French notion of problématique or the German term Problemstellung refers to a specific

research question – this would be a more appropriate meaning for the ‘problem’ in problem-centred interviews. Let us consider the original research puzzle that led to the development and design of the PCI as a distinct technique of qualitative interviewing. When the first author wrote his PhD thesis (Witzel, 1982) in the 1970s he was involved in a research project about occupational socialization of young people (Heinz et al., 1979). The purpose of the study was the investigation of the perspectives of graduates from a (logwer) secondary modern school and their parents regarding the ‘problem’ of finding an occupation. The study wanted to explore these perspectives as authentically as possible and without theoretical preconceptions. The PCI was the genuine method developed for this very purpose on the basis of a review of methodological and methodical discussions available at that time (see below). As research is usually not initiated by people themselves, the first step in addressing the research question of this study in terms of problem centring was taken by researchers. Thus, the starting point here was the identification of asocietal problem with immediate relevance for individuals: the conditions and patterns of transitions of graduates from lower secondary education into the world of work in relation to their familial socialization and other influencing factors. The assumption was that, in the process of occupational orientation (of the child), the researcher would be able to address issues of socialization within families, simply because these issues were relevant to family members. The main challenge and task of the PCI was then to take the perspective of the teenager and his or her parents seriously and to trace their own criteria of assessing and making sense of the problem in this period of their lives, within this rough thematic frame of reference. This first reason for naming the interviewing method according to its orientation towards socially relevant

problems is immediately associated with a key precondition of conducting PCIs: the research question has to correspond to an everyday problem in the perspective of practical knowledge that the respondent can articulate and also has an interest in dealing with. This is an important step towards realising the PCI's endeavour of learning about the real motivations behind actions. In order to bridge the scientific and the practical knowledge without corrupting the respondent's perspective on the problem, the researcher's

perspective needs to be systematised and disclosed. In this way, the term ‘problem-centred’ underlines the method's programmatic opposition to naive empiricism that promotes radical openness and assumes that meaning will emerge only if interviewers restrain themselves (for a critique, see Kelle, 2005). The term also Page 5 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

refers to the practical aspects of the method (see section 2.3). All strategies and activities – ranging from access to the field to forms of communication – are oriented systematically, but flexibly, towards the research

problem, i.e. the object, as well as to the most effective way of disclosing and understanding the respondent's perspective on the problem. Throughout this book we describe the consequences of the PCI's original dialogic perspective on problems on the basis of three examples (cf. Box 1.2 and Table 6.1). They are taken from very different studies that employed PCIs and should explicate the dependence of the choice and implementation of the method on the research object (see the principle of object orientation in section 2.3).

Table 6.1 Three exemplary studies using PCIs

Finally, the aspect of ‘centring’ the problem has caused some confusion in the reception of the technique. It was sometimes interpreted in the sense of a limitation of the topic – it was associated with legitimising strategies of interviewers to bring respondents back on track in case they strayed off topic. Instead, problem centring means that the respondent is encouraged and supported in reconstructing research problems by means of reconstructing practical problems. In the process of a dialogue characterised by mutual trust, the respondent should gradually remember more and more and unfold the overall problem in narrative accounts. This entails the establishment of a focus of the reconstruction of meaning on all crucial aspects

Page 6 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

of the problem involving the breadth and depth that are appropriate for the topic. And there is no reason to expect respondents to stray off topic here.

Box 1.2: Studying Problems: Three Examples Throughout the book we refer to three examples from our own experience in conducting PCIs. They are mainly taken from three very different research contexts introduced below. As we were (leading) researchers in these projects, we can inform about every aspect without reservation and generalise our reflections about particular challenges, pitfalls and mistakes involved. These studies differ in terms of scope, duration, funding and human resources, etc. and are thus able to illustrate the wide range of challenges when using this method (see Table 6.1 for a comparative overview). In the first, STUDY A, PCIs were used in the frame of a longitudinal, mixed-methods study. It investigated the job entry of young adults after completing vocational education in the German apprenticeship system and their successive careers regarding gender and class differences. In the second, STUDY B, PCIs were used for expert interviews in the frame of a short, commissioned and applied investigation of rising costs after the reform of custodianship of adults in the German federal state of Lower Saxony. In a peer research approach, judges were trained to interview judges. STUDY C was a PhD thesis and used PCIs in the frame of investigating meanings of unemployment in the postcommunist context of Lithuania in the perspective of young people in transition to the world of work. The research puzzle and societal problem consisted here in the fact that the transformation from state socialism to market economy brought an end to decades of full employment and introduced mass unemployment as a new problem for both society and individuals. STUDIES A and C had a common interest in the issue of youth transitions, which originally motivated the development of the technique of the PCI. Examples are integrated throughout the book and the reader will learn more and more details about these studies as she moves through the text. The general chapters of the book (3 to 5) include mostly examples from STUDIES A and C. They help to illustrate the basic methodical aspects of the PCI. Chapter 6 complements this general perspective and is dedicated to extensive discussions of examples from STUDIES A and B, which also serve as the basis for the discussion of typical interviewing errors and pitfalls.

Background and Use of the PCI The development of the PCI has its origin in the German tradition of qualitative research and the methods Page 7 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

discourse of the 1970s and 1980s (Mey and Mruck, 2007). In its very first version, the PCI was introduced as a comprehensive mixed-methods approach combining interviewing with case analyses, group discussions and biographical elements (Witzel, 1982). In this book, we will only discuss the interviewing part and refer to it as PCI. The PCI originated in the context of the 1970s revival and re-development of qualitative methods associated with the reception of the interpretive paradigm that Goldthorpe (1973: 449) called a ‘revolutionary “paradigm shift”’. Its development is also a response to the then tenacious status paradox of qualitative interviewing. Open interviewing was also recognised at the time, in influential standard textbooks of social research, as fulfilling several criteria that are taken for granted in empirical research, for instance, the consideration of individual experiences as well as the context of a certain case (e.g. Friedrichs, 1973). Yet, at the same time, the systematic development of qualitative methods of interviewing was largely neglected; they were generally considered as unsophisticated and pre-scientific methods that could not replace the conventional, more formalised techniques. Typically, the qualitative interview continued to have the status of a method applied in the frame of ‘unstructured’ or less structured pre-tests and pilot studies exploring a field of research for the purpose of preparing a ‘proper’, i.e. quantitative-representative, investigation. Alternatively, the material produced with qualitative interviews was used for little more than to enrich and illustrate quantitative analyses with ‘juicy quotes’ that should bring flesh to the bones and colour to the phenomenon under investigation, as Adorno (1961: 8) once put it. After all, as Kohli (1978: 23) maintained, in an influential German publication of the late 1970s, open interviewing was, despite its merits, regarded as too time-and resource-consuming. Somewhat ironically he claims that this also resulted in an arbitrariness of findings that was ‘at least hidden behind an impressive technical apparatus’ in the traditional approach of ‘closed’ methods. The research policy and economy of this period were characterised by hegemonic criteria of generalisation on the basis of representativeness and inference, and by a breadth-before-depth approach. Against this background, the status of qualitative interviewing was, at best, that of a complementary and auxiliary method filling the bythen obvious knowledge gaps that the standardised empirical research (re-)produced. Consequently, the refinement of qualitative methods was until then hardly facilitated. The original design of the PCI, embedded within this methods discourse, tried to establish a distinct qualitative interview technique as a stand-alone method for research that responds to some of this criticism. As indicated in Chapter 2, the principles of object orientation and problem centring constitute alternative perspectives to the instrument-orientation of traditional approaches. The research-economic argument at any rate needs to be rebutted on academic grounds alone at least with regard to the collection and consolidation of qualitative data that is both valuable and valid. Importantly, the PCI is the product of a twofold critique of the social research culture of that time. On the one hand, it criticised the artificial interviewing style suggested by the then hegemonic quantitative paradigm for its fallacy of non-reactivity. Reactivity, meaning the fact that ‘the act of doing the research changes the behaviour of participants’ (McKechnie, 2008: 729), is inevitable as soon as it involves people interacting in face-to-face encounters. While it is commonly understood that ‘people are not machines’, the ‘importance of Page 8 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

interviewing uniformity’ is nevertheless maintained in this approach (Moser and Kalton, 1971: 276). On the other hand, the critique of the PCI addressed radical alternatives from qualitative research, like the narrative interview (Schütze, 1983) for its fallacy of non-intervention. Inherent demands of narrations (Zugzwänge) – i.e. the tendency and capacity of people to unfold, complete and elaborate a story by themselves (and without supposedly ‘contaminating’ contributions of the interviewer) once they began to tell it – are important but rely heavily on quite a few communicative requirements. The rule of not intervening or interrupting accounts is at risk of creating an equally artificial situation that may demand too much of the respondents. As a consequence of this criticism, the PCI instead suggests that interviewers take the role of an agent of active

listening including the stimulation of narrations and thoughts. They dissolve some of the asymmetry inherent in the interview situation by involving the respondent in a process of active understanding that provides the possibility to clarify and deepen meaning and knowledge during the interview. Over the years, the PCI (Witzel, 1982, 1989, 1996) has become recognised as one of the more elaborate approaches within the range of methods for collecting verbal data (e.g. Flick, 2006; Helfferich, 2009: 35–6; Lamnek, 2010: 332–7; Mayring, 2002: 67–72; Reinders, 2005: 116–25). And it has been widely used: the PCI is ‘probably one of the most frequently used types of qualitative interviewing and analysis used in the German social sciences’, as Scheibelhofer (2005: 20) writes. The method is suitable to investigate actions and experiences, their justification and evaluation, as well as individual opinions. It is directed towards topics, objects and their interrelations, which are little explored. Its underlying image of humanity (Menschenbild) considers people as self-reflective and capable of acting and communicating. Overall, the PCI is more of a skill and craft than a (specialised) technique or tool because its appropriateness depends on the concrete object and question of research. The PCI's ‘solution’ of common problems of qualitative interviewing proved to be useful in many research contexts. PCIs were used in a variety of contexts, countries and across many disciplines of the human and social sciences including, for instance: demography (Von der Lippe and Fuhrer, 2004), pedagogics (Szczyrba, 2003), psychology (Kühn, 2004; Mey, 1999; Nentwich, 2008), psychiatry (Roick et al., 2006; Stiglmayr, 2008), sociology (Bolder and Hendrich, 2000; Fritsch, 2006; Reiter, 2003), cross-cultural management (Hajro and Pudelko, 2010), political science (Pregernig, 2007), social work (Schmidt-Grunert, 1999), cultural sciences (Filep, 2009), marketing research (Kurz et al., 2007), medicine (Hasseler et al., 2011), environmental sciences (Medilanski et al., 2006; StollKleemann, 2001), criminology (Strobl, 1998), and food study (Riefer and Hamm, 2011). The PCI can be at the core of interdisciplinary research as it was in the case of the Collaborative Research Centre 186 ‘Status Passages and Risks in the Life Course’ that was carried out at the University of Bremen from 1988 to 2001 (Kluge and Kelle, 2001). More recently, the overall approach of the PCI also constitutes the basis of advances in the method of ‘problem-centred group discussion’ (Kühn and Koschel, 2011). And, as Mey (1999: 148) notes, it has been applied and re-labelled by several scholars for the purpose of their specialised research needs. For instance, Diezinger (1995: 273) calls it a ‘combination of open narrative questions and more precise follow-up questions’; in this way, they try to combine the advantages of an open and non-directive approach with specific stimulation through targeted questions. Lenz (1991: 57) describes Page 9 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

it for his purpose as ‘narratively enlightened guided interview’; and Bock (1991: 161) calls it ‘semi-structured guide-oriented in-depth interview’. Yet apart from a brief introductory text in English (Witzel, 2000), which is available worldwide through the authoritative and recognized online journal ‘Forum: Qualitative Social Research’ (http://www.qualitative-

research.net/index.php/fqs), the technique is so far hardly accessible to the wider international audience. The present book fills this gap by providing an authoritative yet concise and applied introduction to the background and history, scope, technique and application of the method.

Purpose of the Book and Preview of Chapters This book is a comprehensive introduction to the methodology, technique and application of the specific interviewing technique of the PCI (Witzel, 1982, 1989, 1996). It is not a general introduction to qualitative interviewing, and does not intend to replace introductory literature (e.g. Gillham, 2005; King and Horrocks, 2010; Kvale, 2007; Kvale and Brinkmann, 2009; Roulston, 2010; Rubin and Rubin, 2004; Schostak, 2006; Seidman, 2006; Wengraf, 2001). Due to the hands-on character of this textbook we need to neglect relevant specialised discourses regarding methodology and methods that were relevant in the development of the PCI. For instance, the debate about the apparent opposition between qualitative and quantitative approaches as well as their different criteria for research quality is ongoing in the German context although constructive contributions to overcoming it were available 30 years ago (Steinke, 1999; Wilson, 1982/86). We also need to neglect here the discussion of the many approaches of qualitative data analysis that could be applied to interviews collected with PCIs. We can only offer a few practical examples of how it can be done.

Chapter 2, following, starts with an outline of the methodological and epistemological approach of the PCI and indicates how it is distinguished from other established techniques of interviewing. We consider some of the features of interpretive knowledge constitution and explicate the epistemological background of the

well-informed traveller as metaphor for the particular role and attitude of problem-centred interviewers. As a method of interpretive social research, problem-centred interviewing consists essentially in the epistemological challenge of reconstructing problems in a dialogue between interviewers (well-informed travellers equipped with certain forms of prior knowledge) on the one side, and respondents with their practical knowledge from everyday life on the other (see Figure 2.1). They are partners in a temporary and interactive relationship, the quality of which is crucial for the final quality of the interview. After discussing the method's three main principles of problem centring, process orientation and object orientation, we discuss differences of the PCI in relation to other methods of interviewing that share certain similarities.

Chapters 3 to 5 are dedicated to the discussion of the practical steps that usually characterize the process of problem-centred interviewing. These parts constitute a general but dense roadmap for the implementation of PCIs consisting of three steps: preparing, doing and processing PCIs. The flowchart in Figure 3.1 can be taken as the table of contents for these parts.

Page 10 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 3.1 Flowchart of the PCI

Chapter 3, about preparing PCIs’, describes the typical steps before the fieldwork starts. The consolidation of a research interest and its translation into questions go hand-in-hand with the development of a qualitative research design. Then we discuss the crucial status of prior knowledge in PCI research. We distinguish between everyday, contextual and research knowledge and suggest using them in a sensitising way by integrating them into a preliminary framework of research. The development and use of an interview or topical guide is one of the tools that bridge the sensitising framework and the concrete interview situation. Ideally, PCIs are carried out by the principal researcher who is usually better able to handle pre-interpretations developed during the interview. However, this is not always feasible and we therefore also discuss the training

Page 11 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

of interviewers. We conclude this section with some remarks concerning sampling and field access. The main steps of doing PCIs are described in Chapter 4. After some suggestions about the choice of the interview setting, we discuss important aspects regarding the very beginning of the conversation. We then explicate and illustrate the interplay of opening question, opening account and follow-up questions. They constitute the heart of the PCI and form a complex unity of strategies of interaction and communication with the purpose of facilitating a process of discursive-dialogic knowledge production. The collection of information related to the background of the respondent, the optional use of a short questionnaire, and the important moment of debriefing and bringing the conversation and the encounter to a close conclude the discussion in this chapter.

Chapter 5, about processing PCIs’ is dedicated to the typical steps that usually follow the actual interview. The postscript is a self-debriefing tool that helps to capture important information about the conversation and its context which has not been registered otherwise. Transcription and analysis of PCIs do not follow particular rules; they depend on the research design and the research interest. The question of qualitative analysis cannot be discussed exhaustively in this book with its focus on PCI-based data collection. Thus, we restrict ourselves here to providing a general outline of how the principle of problem centring can be considered also in the analysis and interpretation of the interviews. We suggest a series of general steps of a dialogic and problem-centred process of constructing and arriving at ‘findings’. Examples from STUDIES A and C are integrated. Finally, we briefly address a few points associated with the planning of resources for PCI research. Short interviewing examples are integrated into the whole book. Yet in order to do justice to the textbook character of this introduction to problem-centred interviewing, Chapter 6 is exclusively dedicated to providing an in-depth look at practical aspects of the method. We introduce two studies, describe selected aspects of problem-centred interviewing, and provide extensive examples of producing, interpreting and reflecting typical interview passages and how the production of knowledge is contingent upon the interaction of interviewer and respondent. Integrated in this part are exemplary discussions of common interviewing mistakes and pitfalls, which are described and explained with regard to their systematic relation to the basic principles of problemcentred interviewing. Instead of a conclusion, Chapter 7 provides a systematic discussion of interviewing errors. This should not discourage students and researchers from using the PCI but should instead help them to understand the basic principles of the PCI from the negative angle of discussing common interviewing mistakes and pitfalls. Flawless interviews are rare. Neither of us has ever come across a PCI without mistakes, or has ever done one. We think that the idea of being ‘in control’ of the immediacy of interview communication is an illusion. However, the development of the skill of problem-centred interviewing is greatly improved when we are willing to learn from these mistakes. This is what the final chapter wants to facilitate. http://dx.doi.org/10.4135/9781446288030.n1

Page 12 of 12

The Problem-Centred Interview: Principles and Practice

The Programme of the PCI In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 12-34 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The Programme of the PCI As indicated in the introduction, one of the motivations for the development of the PCI during the second half of the 1970s and its particular design was the attempt to firmly establish qualitative interviewing as a recognised alternative within the range of social research methods. A second motivation was the design of a method of interviewing that was more reflected and appropriate for the collection of ‘qualitative’ knowledge about the social world than the ‘semi-structured interview’, which is hardly more than a ‘query tool without theoretical foundation’ (Mey and Mruck, 2010: 423). This chapter introduces some of the characteristics of interpretive knowledge of the social world and discusses its consequences for the overall approach of problem-centred interviewing. In the context of this textbook it introduces a kind of minimal epistemology of interpretive social research (see section 2.1) and how it is translated into the particular challenge of problemcentred interviewing (see section 2.2). Even now, the knowledge issue remains at the core of qualitative research development and reaches into the more recent debate about improving our understanding of social realities by mixing, combining and integrating methods and methodologies of various traditions (e.g. Tashakkori and Teddlie, 1998). The reflection of some of these issues in more detail here, and exploring the way in which they were fundamental for the epistemological and methodological orientation of the specific method of the PCI, are also a reminder of the general principles of qualitative interviewing. On the basis of this discussion, the key features of the original epistemological and methodological concept of the PCI are introduced (section 2.3). Finally, in section 2.4, the PCI is distinguished from selected alternative methods of interviewing.

2.1 What Is Specific About Interpretive Knowledge of the Social World, and How Can We Improve It? The challenge of social research boils down to the following questions which guide the discussion in this chapter:

1

What can we, as researchers, know about the social world?

2

And how can we design our practices of social scientific knowledge collection and production accordingly without:

a

neglecting available social scientific knowledge;

b

inhibiting subjective perspectives; or

c

corrupting the chance to discover novel aspects of certain problems?

The answers to these questions by various research traditions determine how the epistemological problem of knowledge (1) is translated into the challenge of establishing a systematic and reflected methodological programme for the development of an appropriate set of methods, tools and practices of social inquiry (2). There is no single answer to these questions that would be recognised by the different schools in the social Page 2 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

sciences (see Box 2.1). For the purpose of introducing and explicating the programme of the PCI, we want to discuss some relevant aspects related to these questions. In the following we want to unfold our answer (which is necessarily brief here); its essence could be summarised as follows:

Box 2.1: Epistemology, Methodology, Methods - Some Basic Distinctions Like all specialised scientific efforts, qualitative research is characterised by competing paradigms like positivism or constructivism. Each promotes certain world-views and basic assumptions about life and the possibilities and limits of understanding it (e.g. Della Porta and Keating, 2008; King and Horrocks, 2010: Chapter 2; Outhwaite and Turner, 2007). Guba and Lincoln (1994) suggest differentiating research paradigms on the basis of three fundamental and intertwined questions. The ontological question asks for the form and nature of reality. This is the biggest question of the three, and it cannot nearly be addressed here. Suffice it to say that, following the basic understanding of qualitative research, within our common, natural context, social realities are essentially constructed - i.e. their meanings and relevances are products of social interactions. For instance, the importance of money results from the value attributed to it, and not from the material value of the paper or the metal. The epistemological question determines the nature of the relationship between the knower and what can be known. Epistemology is the branch of philosophy that deals with the nature, scope and limits of knowledge. With regard to social research it asks what can be known about social reality and reflects the relationship between researchers and their specific knowledge interest. The methodological question explores how researchers can find out whatever they believe can be known. Methodology is about ‘ways of knowing’ (Moses and Knutsen, 2007). It is a branch of the philosophy of science reflecting the foundations and workings of scientific practices producing knowledge; it is the science of methods. Thus, the fourth question of methods needs to be added. Methods are the actual tools and techniques of social researchers used to obtain knowledge. Typically they correspond - although not exclusively - to certain methodological approaches and help researchers put research interests into practice. Let us come back to our example of finding out about the contemporary relevance of mediaeval urban planning through well-informed travelling (see Box 1.1). The epistemological perspective would suggest that we cannot know how it really used to be a few hundred years ago. We inhabit a different time-space and are culturally different people from those living in the past: we went through a ‘modern’ education system and had our experiences in the twentieth and twenty-first centuries. Yet there are many other details we can find out about the transformation of the city over the centuries and try

Page 3 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

to approximate how city life used to be. For instance, we might be interested in facts regarding the exact outline of the former city and how it determined the transformation of the city-scape. Or, we might want to know how the organisation of life around the city's hardware of buildings and structures changed over the centuries, and how their utilisation was adapted to the requirements of time like the moods of changing sovereigns, or the needs of the people. In methodological terms the search for answers to these alternative knowledge interests requires different research strategies and methods. In the first case we could excavate the remains of original buildings and the city wall after having traced them with infrared technology. Then we could recreate how they were integrated into the architecture that was built on top of it later on. We could also simulate and measure the erosion of parts of the city wall made of sandstone in order to understand the contribution of natural decay, linking their exposure to sun and rainfall with the speed of its disappearance. In the second case we would rather need to reconstruct the social history of the place. We would spend much more time in archives, studying documents describing everyday life, and we would, in order to find out about the current situation, eventually consider living in the city for some time. Then, our understanding could greatly benefit from the guided tour with knowledgeable insiders like the retired teacher whose stories about his parents and grandparents would cover the last century or so. In addition, we could do interviews with experts such as local city historians who might be able to answer the many questions and ideas we developed on the basis of our research. The PCI is a methodical suggestion to reconstruct the interactively constituted knowledge in the social world in an interactive process between interviewer and respondent. It is designed so that the researcher's prior

knowledge defining and structuring the research interest in a preliminary way enters into a discursive dialogue with the respondent's practical everyday knowledge about a relevant issue. The inductive moment of fully considering subjective perspectives complements the deductive moment of building upon available prior knowledge from research in a way that allows novel data to question and revise previous knowledge. The specific management and temporary suspension of prior knowledge should also allow surprising findings by way of abductive inference (Box 2.4). Our metaphor of the well-informed traveller (Box 1.1) describes the associated research attitude. Let us begin with some reflections on the specificity of interpretive knowledge. One aspect that is critical in all qualitative-interpretive approaches is the role and status that is attributed to the researcher's (academic) prior knowledge vis-à-vis that of the practical everyday knowledge of the participants in a certain social setting under investigation. Social scientific paradigms tend to be distinguishable according to the relevance and priority they grant to these two types of knowledge – the first accumulated, systematic, scientific and external, the other emergent, rather spontaneous, practical and internal – in the process of research. Their relationship Page 4 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

needs to be clarified especially for the purpose of doing qualitative interviews, where these two knowledge perspectives are directly confronted and represented in the interaction. Interpretive social science, and with it qualitative research, emerged out of a critique of the then dominant normative position that essentially tries to explain social action on the basis of a deductive reasoning adopted from natural sciences. Following Wilson (1970), two ideal types can be distinguished with regard to the sociological interest in regularities and change in behaviour and action: deductive explanations and interpretive descriptions. Approaches prioritising academic knowledge in the way natural sciences apply it in order to investigate, for instance, the behaviour of non-speaking inhabitants of the animal kingdom, operate mostly in a deductive way on the basis of theories, variables, hypotheses, operationalisation, measurement and testing. Deduction is a form of reasoning that moves from the general (i.e. knowledge, rules or assumptions and hypotheses) to the particular observation, trying to subsume it. Deductive explanations associated with the normative research paradigm try to explain facts by showing that they fit the rule which is known in advance. For instance, a natural scientist investigating animals in a new environment will test all sorts of hypotheses derived from ethology by observation and experiment. The basis consists of strong assumptions about animal behaviour regarding the use of space in relation to other animals or the distribution of food within. The researcher alone has to determine what is relevant because ‘there is nothing at all “meaningful” about nature in the sense of meaningfulness that organises human action’ (Dux, 2011: 19). And in case of doubt, the researcher will probably not try to clarify it by involving animals in social interaction by interviewing, say, a chicken. Such a normative approach that ‘works’ in the natural sciences has also long dominated social research. Social research, however, has to start from completely different presuppositions about the ‘nature’ of the social world and its participants. First of all, the social world is meaningful. The meaning of human action and behaviour is constituted in the interpretive process of social interaction and changes over the course of this interaction. It can only be grasped by means of interpretive description and following the principles of a process called the documentary method of interpretation, which is described in more detail below (section 2.2). The specific ‘nature’ of social realities and their constitution, perception and representation requires an adequate way of investigating them with the appropriate tools and methods. After all, human research subjects, it must be assumed, make sense of the social world and their action in a way that is similar to that of the researchers themselves. What is more, unlike animals, people have the capacity to directly communicate their knowledge and their ways of making sense, a fact that deserves consideration in conceptualising and doing social research. In an inductive process, interpretive descriptions reconstruct meaning according to the subjective relevance of actors and provide the basis for arriving at preliminary knowledge about the social world. For instance, a social scientist investigating human behaviour in social contexts will try to uncover the patterns underlying the observed behaviour by investigating individuals and their social interaction. In case the researcher wants to explore the meaning of their behaviour on the level of their shared common-sense understanding, he has to put his social scientific knowledge aside. In order to approximate this commonsense understanding he will benefit from talking to them even if he may already have some good ideas about it from previous research. Page 5 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

2.2 the Epistemological Challenge of the PCI The process of arriving as close as possible to the people's common-sense understanding is rather complex and requires a change of perspective on the side of the researcher. The readiness to involve oneself with the respondent's perspective is crucial for producing and consolidating knowledge in the interaction between researcher and respondent in the PCI. The following quotation from Alfred Schutz, who was one of the founding epistemologists of the interpretive paradigm, gets to the heart of the task. It begins with a reflection of the fundamental difference between natural and social sciences discussed previously.

The world of nature, as explored by the natural scientist, does not ‘mean’ anything to the molecules, atoms, and electrons therein. The observational field of the social scientist, however, namely the social reality, has a specific meaning and relevance structure for the human beings living, acting, and thinking therein. By a series of common-sense constructs they have pre-selected and preinterpreted this world which they experience as the reality of their daily lives. It is these thought objects of theirs which determine their behaviour by motivating it. The thought objects constructed by the social scientist, in order to grasp this social reality, have to be founded upon the thought objects constructed by the common-sense thinking of men, living their daily life within their social world. Thus, the constructs of the social sciences are, so to speak, constructs of the second degree, namely constructs of the constructs made by the actors on the social scene, whose behaviour the social scientist has to observe and to explain in accordance with the procedural rules of his science. (Schutz, 1954: 266–7) The epistemological challenge for social research – i.e. the challenge of obtaining knowledge about an issue – consists in getting access to the interactively constituted meanings and therefore in collecting and consolidating social scientific knowledge about the issue. The social scientific knowledge is always characterised by remoteness. Unlike first-hand, everyday knowledge (‘common-sense constructs’), research knowledge of the social world consists of categories that are scientific constructs or ‘constructs of the second-

degree’ – social science operates on the basis of ‘constructs of the constructs made by the actors on the social scene’. Even if we take this interactive-inductive approach, the knowledge and experience of the respondent is never fully accessible. As researchers we can only approximate what individuals think about certain issues, i.e. the subjective meaning they have for them. At some point in our example of exploring the small mediaeval city through the eyes of the retired teacher, we get the impression that the old tree also has a very personal meaning for him. He selects and offers this information of his own accord because it is a relevant aspect for him. We try to make sense of what he says. Having had our own experiences carving hearts, we can guess from his hints that the old tree served as a witness to some of his love stories. A question that was never asked could have specifically asked, but would have again provided pre-selected information. For us this glimpse of privacy is a relevant example of how public urban space is meaningfully appropriated and ‘privatised’ by locals. The fact that he takes his

Page 6 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

grandchildren there even turns the tree into a medium of inter-generational knowledge transfer. The concepts ‘appropriating public space’ and ‘inter-generational knowledge transfer’ are part of our knowledge as social scientists. We brought this knowledge to the research situation together with, for instance, our own experience in carving hearts. Yet while the appropriation of public space was one aspect that we wanted to explore from the beginning, the issue of inter-generational transfer and the usefulness of certain memories from our youth came as a surprise. In conceptual terms, these knowledge dimensions are part of our prior knowledge (Meinefeld, 2004), which helps us to ask questions and understand what the teacher emphasises as relevant in his perspective of practical knowledge. The treatment of prior knowledge in qualitative research is a contested issue because it is crucial for the question of whether and how we discover something ‘new’ (see the principle of problem centring in section 2.3 below). The positions range from the complete suspension of prior knowledge in favour of openness, to its full disclosure; even its use in the sense of specific ex ante hypotheses is discussed. The first position wants to avoid the categories of the researcher being imposed on the observations. The second position maintains that it is an ‘epistemological requirement to include prior knowledge in methodological control’ (ibid.: 156). According to the programme of the PCI, prior knowledge is a valuable and indispensable resource that needs to be used in a way which ensures that the principles of qualitative research are not undermined. The particular design of the PCI is an attempt to offer a methodical approach that should help to enhance our interpretive knowledge of the social world by reducing the distance between the researcher's prior knowledge, on the one hand, and the relevance and the practical everyday knowledge of the research subjects on the other. The types of prior knowledge in problem-centred interviewing are discussed in section 3.2, where we suggest ways in which it should be used in the practical context of preparing research. At this point we want to look at how the PCI solves this problem in conceptual terms. In short, the PCI invites respondents to co-construct and reconstruct problems together with interviewers in an interactive and interpretive process of data collection (see Figure 2.1). In a discursive dialogue the researchers’ priorknowledge meets the respondents’ practical knowledge. This exchange provides the chance to develop and refine the researchers’ social scientific constructs (of the second degree) in a dialogue with the respondents’ common-sense constructs (of the first degree). The researchers’ prior knowledge is organised within a sensitising framework in order not to jeopardise the requirement of openness in qualitative research (see below). The function and practical establishment of a sensitising framework for problemcentred research is introduced in section 3.2.

Page 7 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 2.1 The epistemological challenge of the PCI

The problem-centred dialogue involves inductive and deductive moments for the sake of improving our understanding of how and why certain meanings (what) are established with regard to a societal ‘problem’ (see Chapter 1). In this process, the researcher qua well-informed traveller will necessarily ‘operate between two worlds’ and negotiate everyday and scientific conceptions of reality (Denzin, 2009: 9). The feature ‘wellinformed’ that specifies the metaphor of travelling is borrowed from Alfred Schutz's (1964) figure of the ‘wellinformed citizen’ – ‘To be well informed means to him to arrive at reasonably founded opinions’ (ibid.: 122; original emphasis). ‘Well-informed’ refers to a practice that, in order to understand the background of certain issues, appropriates and unfolds relevant knowledge in the perspective of one's interest. This interest serves as the first determination of relevance within ‘an infinite number of possible frames of reference’ (ibid.: 130). In this sense, problem-centred interviewers are reasonably well-prepared and have a solid basis of prior knowledge; their priorities and interests are established, and they keep them in mind when they enter into a nevertheless open dialogue with respondents. Their preliminary roadmap and frame of reference is open to modification and revision on the basis of conversations with the locals about what they consider relevant and worth seeing/noticing. The epistemological programme of the PCI – i.e. the effort to enhance the reconstruction and comprehension of a certain problem in the social world; at the intersection of practical and social scientific prior knowledge; and through dialogic interaction of respondent and researcher – builds upon a range of arguments and Page 8 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

principles developed in the frame of the establishment of the interpretive paradigm and some of its more distinct approaches like ethnomethodology and symbolic interactionism. With the general approach of the PCI in mind they are now briefly introduced.

Basic Requirements of Interpretive Research (a) A first requirement of interpretive research, which came out as one of the consequences of the criticism of the normative-deductive paradigm, is the pervasive consideration of the perspective of the actors. According to the principle of openness, interpretive researchers refrain from formulating explicit hypotheses in advance in order not to mark the boundaries of possible discoveries (Blumer, 1969; Glaser and Strauss, 1967). Instead, the research process is centred on subjective perspectives of the problem on the basis of relatively open theoretical concepts specifying the research question. In other words, it is essential to ‘listen’ to the data and to analyse them in an impartial way by condensing preliminary findings and classifications step by step into more comprehensive concepts. The generation of theory is oriented primarily to the dense everyday lives of individuals and what is relevant to them, as well as to the decoding of the meaning they attribute to their patterns of action and interpretation. Thus, causal explanatory models are rejected. (b) Another feature of the paradigmatic shift is the emphasis on the communicative character of data collection that follows the assumptions of symbolic interactionism regarding the interpretive character of human interaction oriented towards meaning (see Box 2.2). The researcher as a non-disruptive factor in the process of observation and data collection is the ideal in natural sciences - when this was transferred into social research, it was widely criticised: face-to-face interviewing is not a non-reactive method. The introduction of the natural science ideal of non-reactivity can be seen as an attempt to neutralise the influence of interviewers by training them to indifferently ask questions and register answers in an uncommon way that invites a variety of equally uncommon attitudes and responses (Berger, 1974; Cicourel, 1964). Yet the interviewer's influence in face-to-face situations can never be erased, but instead needs to be organised in a way that the research interaction becomes a confiding collaboration providing access to the respondents’ ‘stubborn world’ (Blumer, 1954: 8).

Box 2.2: Symbolic Interactionism and Problem-Centred Interviewing Symbolic interactionism was developed in US social sciences at the beginning of the twentieth century and has become one of the theoretical pillars of interpretive research. According to its ‘root assumptions’, summarised by Herbert Blumer (1969), individuals act toward things on the basis of meanings that these things have for them; these meanings result from a process of social interaction, where they are interpreted (i.e. symbolic), revised and modified (see Denzin, 2004: 82). The fact that people constitute meaning in

Page 9 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

a process of interactive negotiations and mutual role-taking, and thus create the social world as their world, has to be considered in social research. According to Blumer (1969) it results in a research programme that enters into a ‘direct examination of the actual empirical social world‘ (ibid.: 48) and is distinct from ‘elegant logical models’ (ibid.: 24). Blumer suggests a ‘naturalistic’ approach to the empirical social worlds and demands an ‘investigation that is directed to a given empirical world in its natural, ongoing character’ (ibid.: 46). The two phases of exploration and inspection are at the core of such an approach together with sensitising concepts. These concepts are used in the sense of elastic theoretical concepts, which are developed at the beginning of the research process and kept open during it (see below sections 2.3 and 3.2). Blumer's notion of meaning production in social interactions is the basis for emphasising the necessity of dialogue as an important methodological element of the PCI. This dialogue is discursive because it reflects a process of learning: interviewers use these elastic sensitising concepts in order to extend their knowledge and to overcome preconceptions through an examination of the exploration. According to the symbolic interactionist concept of meaning as changeable and situational, problem-centred interviewers also accomplish processes of inference and preliminary conclusion during the course of the interview. These preliminary conclusions (termed pre-interpretations in the terminology of the PCI) constitute a crucial precondition for the fruitfulness of the later analysis and interpretation. The scientific reconstruction of meaning of action and orientation has to consider that meaning can result not only from a real social interaction but also from a dialogue with oneself in the sense of ‘self-interaction’ or as an ‘internalised social process’ (Blumer, 1969: 5). In either case, this generative process of dialogic meaning production may remain hidden in the results of everyday practice. It needs to be reconstructed and disclosed through the process of interaction between interviewer and respondent. This interaction should facilitate memory, create trust and provide the chance to unfold details, and allow statements to be revised and corrected in the course of the conversation. (c) Finally, it is especially the work of ethnomethodologists like Harold Garfinkel (1967/2011) that reminds us of the context dependence of the meaning of articulations, be it single phenomena, or observed action. Situations, actions or statements are not isolated phenomena, but refer to certain backgrounds - i.e. they are documents of underlying patterns. This indexicality, as ethnomethodologists call the referential power of observed phenomena and the ‘mutual determination of appearances and underlying patterns’ (Wilson, 1970: 701), has to be considered in interpretive social research. One way of doing this, as suggested by ethnomethodology, consists of applying the documentary method of interpretation. In a nutshell, Garfinkel describes this method in the following way: Page 10 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The method consists of treating an actual appearance as ‘the document of’, as ‘pointing to’, as ‘standing on behalf of’ a presupposed underlying pattern. Not only is the underlying pattern derived from its individual documentary evidences, but the individual documentary evidences, in their turn, are interpreted on the basis of ‘what is known’ about the underlying pattern. Each is used to elaborate the other. (Garfinkel, 1967/2011: 78) The documentary method of interpretation is a universal (hermeneutic) mode of understanding and meaningmaking; it is used in everyday life as well as by researchers. Once again, just think of the retired teacher admitting to carving into the old tree while embracing it with a smile. His few words, and the gestures accompanying them, point to stories from his youth that are not further divulged but are nevertheless basically understood. The principles of this universal mechanism of understanding are relevant for the programme of the PCI (see Box 2.3).

Box 2.3: Ethnomethodology and Problem-Centred Interviewing Ethnomethodology is a specialist sociological approach dedicated to the investigation of the micro foundations of social order as established by members of a social group. Its goal is to ‘determine the principles and mechanisms by means of which actors, in their action, produce the meaningful structure and ordering of what is happening around them and what they express and do in social interaction with others’ (Bergmann, 2004: 72). Ethnomethodology is not so much interested in the contents of the interaction but in the question of how the process of mutual understanding between individuals in everyday life is possible and practically organised. One basic problem consists in the above-described indexicality of articulations, actions and events. Language is a key factor in this context: each articulation is special in an indexical sense and refers to a certain meaning that can be grasped only by considering the context of its occurrence. The meaning of what is said depends on when it is said. Individuals understand the meaning of expressions through the documentary method of interpretation. It is a universal method of understanding that is equally relevant in everyday as in purposive communication like interviews (see also section 4.6). In order to understand the respondents’ subjective meanings, interviewers need to recognise this method as the precondition of any form of social scientific activity in general and of the technique of questioning in particular. In order to understand the separate articulations of respondents, interviewers try to decode the underlying pattern - the actual meaning. But in turn they also apply interpretive patterns, attribute meaning to expressions, and perceive them as supporting their interpretation. At the same time, they keep these interpretations open in the process of communication: later statements of respondents may, for instance,

Page 11 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

force them to revise interpretations because they do not fit the established pattern. This may force them to also reinterpret earlier statements. If no pattern emerges, they have to collect additional statements in order to develop interpretations. The openness towards a range of interpretations, as well as to their retrospective correction and prospective anticipation, corresponds to the way in which actors in interactions produce and reproduce patterns of reality as a matter of routine. This openness is, on the one hand, transferable to the overall constitution of a research process: it asks for the abolishment of strict rules of conduct and research implementation (see the criterion of process orientation below). On the other hand, it requires giving up the conventional question-answer pattern in interview research (see also sections 3.3 and 4.6). The programme of the PCI is one of the methodological and practical solutions that try to facilitate the step from ‘pre-understanding’ to understanding (Gummesson, 1991: 51–72) and that correspond to these principles of interpretive research. The consequences of the positions of ethnomethodology and symbolic interactionism for the design of the PCI are especially relevant. On the one hand, qualitative interviews need to consider the relevance of the context of articulations in the process of both data collection and interpretation. On the other hand, it is necessary to conceptualise the interview as an interaction process in which interviewers permanently interpret meaning that is constituted as ‘situated, self-organizing and reflexive interaction between the organization of memory, practical reasoning, and talk’ (Cicourel, 1974b: 100). The notion of pre-interpretation (during the interview), which was introduced with the PCI to the methods discourse, is distinct from systematic interpretation in the phase of analysis. It addresses a completely neglected aspect of interview communication and emphasises the achievement of the researcher as interviewer: during the conversation he keeps interpretations open, assesses prospective projects of meaning, and allows for retrospective revision of them, even provoking it. In so doing, interviewers assume a

double role that corresponds to the rules of the documentary method of interpretation: they ‘wait’, in a state of relaxed attentiveness, and they listen, becoming actively involved in understanding and following up. This method reflects everyday activities, and there is nothing mysterious about it (apart from, perhaps, Garfinkel's label for it). Thus, it is easily added to the skill set of interviewers, underlining the closeness of the PCI to everyday life. In the following we explain how the PCI deals with issues of theoretical and conceptual biases, the communicative nature of social research, and the contextuality/indexicality of the meaning of linguistic expressions. For this purpose, the principles of problem centring, process orientation and object orientation are introduced. The capacity of this approach to address the requirements of the object under investigation (and also in multi-methods frameworks) is explicated, as well as its specific profile in comparison to alternative methods of interviewing which are briefly illustrated. Practical issues of the PCI are then discussed in the succeeding chapters. Page 12 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

2.3 Principles of the PCI The Principle of Problem Centring The principle of problem centring is at the core of the methodological programme of the PCI. Its distinct function and importance have not always been fully understood (e.g. Flick, 2006: 161–4). The main purpose of problem centring is the facilitation of a conversation structure that helps to uncover the actual perspectives of individuals on a particular problem in a systematic and dialogical way. In methodological terms, problem centring specifically addresses the non-negligible influence of the researcher at all stages of the investigation; it is directed against both the empiricist bias of normative approaches and the radical openness of its critics. The notion of problem centring refers at first to the starting point of research, i.e. to a socially relevant problem that was identified by the researcher (see Chapter 1). This curiosity of the researcher for a certain issue, and not a different one, constitutes a rather explicit interest that needs to be reflected in the overall process. First of all, the knowledge background of this interest – the researcher's prior knowledge – needs to be disclosed and explicated. This is done by means of a critical discussion and appropriation of relevant theories, concepts, empirical studies and other sources regarding the issue; it can also include first explorations in the field as well as the consideration of expert knowledge and experience. Furthermore, the basic conditions and structural features of the lives of the prospective respondents of the interview study need to be investigated and documented. This is necessary in order to better understand the circumstances on which their behaviour and action depends, and in order to improve the interpretation of their orientations, purposes and intentions. The anticipated attention to crucial (objective) contextual aspects of the respondents benefits the comprehension of the assimilation of societal realities as well as the practical interviewing task of asking follow-up questions and probing. The alternative strategy of locating the entire construction of societal problems within the realm of the subjective constitution of individuals represents an epistemological fallacy: it is misleading to conceive of the researcher as a tabula rasa that quasi-unreservedly absorbs the expressions of individuals. The researcher, in approaching the empirical reality, may succeed in putting aside the conceptual framework he may have in advance. Yet at least when it comes to the practical implementation of the study, the introduction of analytic categories reflecting the second nature of the researcher as theorist seems unavoidable. Metaphorically speaking, the pollution of the field by the researcher's prior knowledge is, after all, inevitable. Not making use of this prior knowledge would essentially deprive us of the learning process and the chance to discover something novel. We suggest calling the dilemma, which can result from this problem in case it is not resolved, ‘Dr Jekyll–Mr

Hyde syndrome’ (Witzel, 1982: 69). It refers to situations where the researcher is torn between impartiality claims on the one hand, and claims to utilise accumulated scientific knowledge on the other. The implicated double nature burdens the research performance at various stages: during the process of data collection the Page 13 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

researcher is expected to empathically assume the perspective of the respondent; the screening and analysis of the collected material, however, should strongly draw on available concepts and theories. Contradictions of what is essentially a ‘double operationalisation’ of the phenomenon may arise. As indicated in Box 2.2, the discussion of this dilemma and suggestions for resolving it are not new. More than 50 years ago, Herbert Blumer proposed the utilisation of ‘sensitising concepts’ for social research (Blumer, 1954). Unlike a rigorous theoretical framework that is established in advance, these concepts are elastic and open to revision because of their general, empirically not contentful nature (e.g. ‘culture’, ‘institutions’, ‘structure’, ‘roles’) (see Kelle, 2005). Practically, the research starts with a preliminary conceptual account of the problem in question that should sensitise and prepare the perception of the researcher. However, this selective preconception, and the associated prior knowledge that any research needs in order to be focused, are kept open towards the empirical observation: the scientific attitude with its specific system of relevances is ‘temporarily dropped in order to be resumed again’, as Schutz (1953: 31) expresses it. In other words, the empirical knowledge assumes a moment of ‘control’ over the theoretical and conceptual knowledge that is revised and consolidated as the process of research evolves. This integration of available and newly identified knowledge also characterises the phases of interpretation and communication as ongoing processes of inquiry of meaning by means of the documentary method of interpretation. The fact that the interviewer already has prior knowledge that is sensitising also increases the chances of abductive insights (see Box 2.4). The practical aspects of regarding and using prior knowledge as sensitising knowledge in PCIs are specified in section 3.2.

Box 2.4: The Sensitising Attitude of Problem-Centred Interviewers and the Chance of Abductive Insight The general attitude of regarding one's prior knowledge, whether from research or other experiences, as sensitising knowledge also increases the chances of arriving at abductive inferences. In contrast to conclusion by induction (i.e. generalisation) or deduction (i.e. subsumption), abductions represent discoveries of new relationships in the data. As Hansen (2008: 457) writes: ‘Abduction results in a tentative and subjective interpretive synthesis among our sensitizing concepts’. Abductive insights are like surprises; they are not accessible through our habitualised routines of analysis and interpretation if we stick to our apparently consolidated knowledge. We have to challenge our knowledge and suspend our certainty about its validity in order for new insights to ‘break through’ and finally improve our understanding. Following Kelle (2007: 127) one can describe the process of ‘learning by failure’ as a practical example of abduction in the context of life-course transitions. For instance, if

Page 14 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

the transition to employment fails and biographical expectations are not realised despite supposedly doing everything ‘right’, the actor is forced to search for an alternative solution. The actor could change his goals and try again in a different way. A young chef finding himself unemployed after completing his apprenticeship training could try to get a job on a cruise ship instead of a restaurant. This would require only a change in the definition of his goals. Alternatively, he can stick to his goals (i.e. to work in a restaurant) and his definition of the situation (i.e. unemployment) and modify his responses. He could, for instance, become self-employed and start his own restaurant. Social innovations like patchwork families, which increasingly are replacing traditional families, are other examples of abductive responses, in this case, to the diversification of relationships. In both cases, the stubborn clinging to traditional solutions (and knowledge) is given up in favour of novel ways of handling the situation. Yet the actors will only afterwards and in retrospect establish whether it (i.e. their abductive response) ‘worked’. The sensitizing attitude of the problem-centred researcher has a similar function. It should keep conclusions and knowledge open to the transformative influence of the empirical data and evidence from the interviews. In qualitative and problem-centred research, knowledge is kept preliminary. As Reichertz (2004b: 163) puts it: ‘Abductive inferencing is, rather, an attitude towards data and towards one's own knowledge: data are to be taken seriously, and the validity of previously developed knowledge is to be queried.’ Against this background the connotation of the principle of problem centring is now twofold. First, it refers to a relevant societal problematic and its theoretical formulation in terms of elastic prior knowledge of the researcher that is part of a sensitising framework of research. It is for this reason that, ideally, the researcher also does the interviews - as a researcher/interviewer. He may need specific training for problem-centred research but he should at least be supported by a kind of supervision. For practical or ethical reasons this is impossible in many research projects; alternatives may even be preferable, as illustrated with one of the examples introduced below (i.e. STUDY B). In these cases, interviewers need to be carefully trained, prepared and involved in the research process in a comprehensive way in order to understand the research questions and become fully competent and independent co-researchers. Only then does the interview provide the chance that prior knowledge, where it exists, can be replaced with discovery (see the principle of process orientation below). Second, the principle of problem centring aims at research strategies that are able to optimise the respondents’ possibilities to explicate themselves. Their points of view regarding the specific problem must have the chance to be brought to bear even if they are opposed to the original interpretations of the researcher or any assumptions possibly underlying the interview or single questions. This point is particularly important considering the fact that many respondents reflect upon their lives and their contexts for the first time in a more systematic way. The principle of problem centring thus triggers an interactive reflection of Page 15 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

both the self and its relationship to circumstances: as the interview unfolds, respondents link aspects of their own lives to other people and institutions relevant to the problem (i.e. relational self-reflection). Thus, through problem centring, respondents can establish a network of meaningful interrelationships between relevant aspects according to their own criteria and priorities. Importantly, this avoids the unreflecting analysis of isolated ‘variables’ on the basis of (questionable) assumptions concerning causality relationships between independent and dependent clusters of variables; this would be typical of scientific reasoning.

The Principle of Process Orientation The idea of operating on the basis of previously introduced sensitising concepts points to another basic principle of the PCI – the principle of process orientation. This calls for the stepwise and flexible production and analysis of data. The qualities of single aspects of the problem as well as their interrelationships emerge slowly and in permanent reflection of their relation to the methods used. The principle of process orientation applies to the overall course of research as well as to the development of the communicative exchange between interviewer and respondent that is part of the process of reconstruction and comprehension. Finally, the controlled consolidation and extension of interpretation in the scientific context is also a process. This principle closely follows the research programme of the Grounded Theory approach as it is described, for instance, in Strauss and Corbin (1990). The use of qualitative methods is here suggested in order to generate theories in the course of a process of data collection and analysis, rather than to presuppose theories. Yet, as the above discussion of prior knowledge indicates, research can hardly be free of preconceptions of some kind. The research process, thus, always accommodates deductive moments that need to be dealt with in a systematic way. Following the Grounded Theory approach these deductive moments need to be channelled into a procedure of discovery that allows overcoming possible biases from prior knowledge. The basic idea is to conceive of the research process as an iteration of inductive and deductive steps. First, the collection, organisation and interpretation of data lead to categories and theoretical concepts. This is followed by the selective use of these core categories/generated theories in organising a more structured search for new ideas and relations in the data. In this way, deduction and induction enrich each other mutually. Deductive qualitative research is closely linked to theoretical sampling. Unlike statistical sampling, which may consider possible combinations of properties of certain variables in the process of selecting single cases randomly, theoretical sampling operates on the basis of purposefully defined comparison groups. The criteria and rules for defining groups and subgroups included in the process of data collection are thereby established on the basis of the development of theoretical knowledge about the problem; they are modified and adapted as the research progresses. Theoretical sampling is an integral part of theory generation in the sense of a parallel process of data collection, coding and analysis. In this way, following the Grounded Theory approach, the emerging theory produces its own logic of selectivity, defining both direction and depth of its development. With regard to the interview as a particular event between interviewer and respondent, process orientation aims at constituting a chance to produce and revise pre-interpretations at the moment of data collection. Page 16 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Pre-interpretations represent the subjective level of knowledge and understanding of individual researchers/ interviewers as it evolves during the conversation. They are interpretations that are developed and revised in the process of interviewing; they are further refined, systematised and controlled in the process of analysis and interpretation after the interview. If the communication process is focused well enough on the reconstruction of relevant orientations and actions, the respondents react with trust and open up. This trust relationship promotes their capability to remember, and it motivates self-reflection. As respondents continue to explain their view of a problem openly and in cooperation with interviewers, new perspectives on the issues are produced again and again in the course of the conversation. This may also lead to alternative interpretations of the same topic as well as to a revision of earlier statements or to redundancies and contradictions; both are welcome in PCIs. Redundancies are valuable as they reformulate perspectives and facilitate interpretation. Contradictions express individual ambivalences and indicate indecision that should be addressed; they may be the result of misunderstandings on the part of interviewers or an error or lack of memory among respondents. Contradictions can and should be clarified in the process of interviewing through follow-up questions, probing and confrontations (see Box 4.5). Taking a more interactive approach to the PCI means that the interviewer is able to overcome communicative barriers or poor articulation which may hamper the exchange. The specific communication strategies of the PCI producing both interview material and comprehension are devised to facilitate this process of iterative knowledge consolidation; they are introduced in more detail in section 4.6. In particular, the more narrative-oriented approaches in problemcentred interviewing allow the reconstruction of the historical relevance of a certain topic to a respondent. The typically ahistorical character of structured interviews is thus overcome and replaced by (micro-) historical accounts.

The Principle of Object Orientation A third criterion of problem-centred interviewing concerns the appropriateness of methodical and practical approaches to the research issue as well as to the research subjects; it originates from the discourse about and the criticism of conventional, normative-deductive approaches. They are characterised by being ‘method oriented’, i.e. having a clear preference for certain methods independently of the research topic. And normative approaches tend to investigate social phenomena by atomising theoretical assumptions into measurable variables, a process that contributes little to the clarification of the way in which individuals actually perceive their realities. The principle of object orientation is directed against the automatic application of somehow recognised and established methods (for instance, in textbooks) without reflecting their appropriateness. This applies also to possible preferences for qualitative techniques: for particular research questions standardised methods can be the better choice. In the meantime, related issues are also discussed under the label of the ‘principle of appropriateness’ (Steinke, 1999: 38–40). The design of PCIs for a specific research purpose and the emphasis on certain practical aspects always need to consider the best way of establishing access to the investigation of the individuals’ behaviour and reflections. In short, the practical realisation of the PCI entirely depends on the peculiarities of the object of research. For instance, the analysis of life courses requires an emphasis on biographical aspects and needs Page 17 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

to give priority to narrative elements over the involvement of interviewers in structuring the conversation. However, the analysis of interpretive patterns with regard to a societal and biographical problem (e.g. coping with the requirements of working life) rather asks for a more dialogical way of interviewing. Here, the possibilities for detailed follow-up questions are more important. In both of these examples the practical design and the process of doing the PCI need to be adapted. The principle of object orientation should not be mistaken for a plea against quantitative methods. In fact, it is rather an invitation to overcome the opposition between methodical approaches for the sake of improving the potential of social research. The identification of large-scale regularities will continue to need numbers-based generalisations as much as the substantiation of frequencies in terms of lived experience cannot do without qualitative analyses of people acting, interacting and interpreting the social world.

2.4 the PCI and Other Interviewing Techniques The PCI is one of many distinct techniques within the landscape of qualitative interviewing (see Flick, 2006 for a comparative overview). Especially in Germany, the invention and labelling of specialised interviewing techniques can be quite confusing (see Mey, 2005 for a discussion). The PCI is an integrated suggestion to overcoming the apparent contradiction of deductive and inductive aspects of qualitative research. It systematically deals with the tension between prior knowledge and empirical material in the research process without trying to get rid of it. The method produces a dialogue between research (i.e. conceptual and theoretical interest in and knowledge about a certain issue) and social reality (i.e. knowledge of individuals, experts in the field). The explicit interest of interviewers in certain topics and their aspects alongside issues emerging as relevant for respondents constitutes an important commonality of the PCI with the active interview (Holstein and Gubrium, 1995). The PCI, like the active interview, acknowledges and reflects the interviewers’ active involvement in the process of meaning production. Yet in addition to the how (process) and the what (substance) of meaning-making that the active interview emphasises, the PCI has an explicit interest in

why (reasons) there is the prevalence of certain meanings, actions and opinions. For the PCI, the subject and knowledge behind the interviewer is as important as the ‘subject behind the respondent’ (ibid.). This is the reason why the PCI suggests applying and explicating both research interest and prior knowledge of interviewers in terms of ‘sensitising concepts’. They facilitate pre-interpretations that are specified, modified and substantiated interactively during the interview. The how of the interactive constitution of meaning (what) could then be reconstructed by various types of sequential analysis of the material. In our introductory remarks we mentioned that this discursive-dialogic approach of the PCI distinguishes it from methodical alternatives that were more strongly in opposition to the quantitative paradigm and its fallacy of non-reactivity. For instance, the radical tabula rasa attitude of the narrative interview designed by Schütze (1983; Jovchelovitch and Bauer, 2000) idealises the subjective perspective with paradox consequences: the exaggeration of the principles of openness and non-intervention, and the bracketing out of prior knowledge, Page 18 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

reintroduce features of the artificially distanced and ‘neutral’ observer. While techniques like the narrative interview or the life story interview (Atkinson, 1998) have a nearly exclusive focus on first-person narratives and their extensive reconstruction, the PCI considers narrative elements alongside other sorts of text. In particular, the narrative with its strong foundation in narrative theory follows a strict choreography. The initial opening question is followed by an uninterrupted main narration. When the latter is concluded, ‘immanent’ follow-up questions carefully deepen certain issues. Further so-called ‘exmanent’ questions introducing additional topics can complement the interaction but are strictly separated from the rest of the interview (Jovchelovitch and Bauer, 2000: 62). The narrative interview heavily relies on the ‘power’ of narratives and the ‘inherent demands of the narration’ (Zugzwänge) (Jovchelovitch and Bauer, 2000: 60) to produce a detailed, complete and ‘uncontaminated’ account of what is relevant to the respondent. In order for narratives to ‘work’ they have to assume (and realise in practice) an extremely asymmetrical situation: at least in the first and most important part of the narrative interview, the burden of maintaining the interview by providing an extensive narration is mainly on the shoulders of the respondent; interviewers are required to abstain from more far-reaching interventions and interruptions. Problem-centred interviewers are much more flexible in becoming actively involved in the production and clarification of meaning than narrative interviewers. This dialogical process also underlines the flexibility of the method to meet the communicative capacity of the respondent. In other words, narrative competence of the respondent is not a methodological precondition of the PCI; and its application is thus less restricted. Furthermore, as we will show, ad-hoc questions and forms of specific probing, even confrontations, can be used in order to fully explore the phenomenon of interest more directly. The PCI uses narratives differently (see Mey, 1999: 138–50; 2000). In a methodologically reflected way, the PCI encourages and utilises narratives as prime sources for further exploring the given ‘problem’ in a dialogic way. Sometimes this is done on the basis of an interview guide but the specification of the dimensions of the problem unfolds essentially from within the perspective of the respondent. The appropriateness of using more symmetrical-dialogic or more asymmetrical-narrative forms of communication depends first of all on the development of the interview interaction. Unlike the narrative interview, the PCI does not suggest a strict adherence to certain stages where one or the other is ‘applied’ (see also Box 4.4). Also the separation of narrative and argumentative passages in the sense of experienced versus reflected knowledge – which, for instance – Nohl (2009: 22–3) in the perspective of the German school of documentary method demands is not compatible with the problem-centred approach. Nohl (ibid.: 22) claims that one cannot assume that, beyond solving certain action problems, ‘respondents know what they know, that they can thus simply explicate their knowledge’. He maintains that respondents are not able to establish a cognitive distance to their thoughts and to reflect them. The assumption is that especial knowledge that instructs action is knowledge ‘that has to be explicated first by the researchers’ (ibid.: 23). Thus respondents should not be ‘urged’ into argumentative statements during the interview where they would establish the difference between implicit and explicit knowledge themselves. The PCI does not establish such strong assumptions. The separation between narration and argumentation Page 19 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

certainly is analytically relevant. Yet in the reality and practice of accounts, these two aspects of knowledge are not separated – there is no need to urge respondents to change from the level of telling about experiences to the level of reflecting on them. First of all, people are perfectly able and used to reflecting the preconditions and anticipating the results of their action against the background of preferences, expectations and projects. What else is the establishment of meaning in everyday life? In the logic of symbolic interactionism, an actor can, as Blumer (1969: 5) notes, even establish meaning ‘in a process of communication with himself’. Meaning is not just there; actors have to work for it: ‘The actor selects, checks, suspends, regroups, and transforms the meanings in the light of the situation in which he is placed and the direction of his action’ (ibid.). It is exactly a process that involves a ‘reflexive interaction between the organisation of memory, practical reasoning, and talk’ (Cicourel, 1974b: 100; emphasis added). The establishment of meaningful knowledge about one's action can be done in interaction with others or in a silent dialogue with oneself. The PCI takes this capability of the actor seriously and translates it into the design of an interview interaction that invites the respondent into a joint process of reconstructing his knowledge. It is the interview's double role to stimulate and support this process of sorting thoughts and talking about it. In this way both become part of an iterative process of understanding. In the end, the interviewer's contributions, and especially the follow-up questions, can help to explicate the respondent's implicit knowledge and make it more systematic for later self-reflection. During the interview, respondents are as much involved in the interpretive work of meaning-making as interviewers. For the PCI, the argumentative reflection of experiences is not the privilege of the researcher. After the interview, and according to the respective approaches of data analysis, researchers can of course continue the process of interpretation and explication without the direct participation of the respondent. Like many other types of interviewing, the PCI also makes use of the tool of an interview guide. Yet this does

not make a conventional semi-structured or guided interview. The interview or topical guide does not have the purpose of establishing a question–answer scheme for working off a list of issues with open questions. Instead, in the sense of a road map, the interview guide should establish a thematic frame supporting the researcher/interviewer qua well-informed traveller in his purpose of problem centring (see section 3.3). In the PCI, the question–answer pattern is replaced with dialogic interaction emphasising the quality of the encounter as a (temporary) social relationship. The PCI is also not simply a combination of different techniques or interviewing styles, e.g. the narrative and the guided one. It is, as we want to show in this monograph, an original and integrated method of interviewing. By emphasising different aspects, it should become clear that all of these interview techniques offer alternative solutions to some of the problems of interpretive knowledge constitution. They all have features that may make them preferable methods for certain research questions; yet none can provide all-inclusive solutions to the challenge of qualitative collection of verbal data. Finally, as distinct methods, they can still be applied for various research purposes. For instance, all of these techniques provide alternative ways of doing expert interviews; and they can equally be used for ethnographic research. In any case: the technical aspects and the practical design require adaptation to the specific research purpose; while sharing commonalities of Page 20 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

the interpretive approach, the texture of findings produced by them can vary; and their application will likely yield different possibilities and limits for generalisation.

The PCI as Expert Interview Due to its flexibility and its object orientation, the PCI can be used for a variety of purposes including interviewing in the context of biographical and life-course research. Furthermore, the PCI is a suitable way of doing expert interviews in accordance with their main functions of exploring, systematising or generating theoretical and insider knowledge about certain issues (Bogner and Menz, 2009). According to the more traditional understanding, expert interviewing is mainly concerned with the exclusive knowledge that officials represent owing to their position, thereby neglecting the person ‘behind’ his role. This attitude has changed, and more recent contributions to expert interviewing promote an alternative view which instead considers interpretation as a necessary second layer of knowledge. This approach emphasises the problem of separating technical knowledge and factual information from interpretive knowledge associated with personal appraisals and interests. In this alternative perspective, the personal experience of experts and the resulting specific opinions and aspirations are taken into account; they are recognised as being part of the experts’ process of meaningmaking in which knowledge is always embedded. The point is, like other participants inside social contexts, experts make use of leeway or circumvent and evade rules or guidelines that may be considered, from the outside, as static and constitutive of organisations. Also expert knowledge has a subjective and pragmatic dimension that can be reconstructed in addition to the substantive aspect (see Box 3.1). Correspondingly, for Bogner and Menz (2009: 54) experts are ‘people who, on the basis of specific knowledge that is derived from practice or experience and which relates to a clearly demarcated range of problems, have created a situation where it is possible for their interpretations to structure the concrete field of action in a way that is meaningful and guides action’. An expert interview is first of all defined by its particular respondents in the above definition; it does not constitute a distinct method as such (see also Trinczek, 2009). For instance, the theory-generating expert interview is ‘a qualitative interview with a particular social group’ (Bogner and Menz, 2009: 55). However, after having characterised the specificity of expert knowledge and its constitution, Meuser and Nagel (2009) suggest that the appropriate way of collecting it is in an ‘open interview based on a topic-guide’. It should avoid ‘closed questions and a prefixed guideline’, and ‘narrative passages are not excluded’ (ibid.: 31–2). The PCI corresponds perfectly well both with the interest in investigating the interpretive dimension of

expert knowledge (see Box 3.1) and with these altogether rough criteria for carrying it out. Owing to its epistemological foundation in inductive and deductive logics, the PCI is suitable as long as it considers the specific role of the respondent in the conversation (Bogner and Menz, 2009: 74). Depending on its purpose, expert interviews may require a high level of specialised and technical prior knowledge on the side of the interviewer. In this case, interviewers may act in the role of co-experts. Furthermore, expert interviews may Page 21 of 22

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

force respondents into loyalty conflicts with their organisation that need to be considered. Such conflicts may result from an inconsistency between individual experiences and interpretations on the one hand, and the public image put forward by the organisation on the other (see section 6.2, discussing STUDY B, for an example of using PCIs for expert interviews). Evasive answers during the interview are indicators of conflicts of loyalty and should be treated carefully. In case of doubt, interviewers should ask respondents to sign a written agreement for the interview to be used. http://dx.doi.org/10.4135/9781446288030.n2

Page 22 of 22

The Problem-Centred Interview: Principles and Practice

Preparing PCIs In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 35-63 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Preparing PCIs After having discussed key epistemological issues informing the design of the PCI, Chapters 3 to 5 explicate how the related methodological challenges are put into practice. They provide an overview of the main features and practical elements of the method in the overall process from conceptualising the interview according to research interest to analysing and interpreting it. These parts constitute a general and rather dense roadmap for the implementation of PCIs. These chapters should show how some of the stages in interviewing, which are also common to other techniques, get the specific PCI ‘touch’. The process of problem-centred interviewing falls roughly into three main stages: the preparation of the overall conceptual and practical approach; doing the interview; and the processing of the material collected. The flowchart of the interview in Figure 3.1 provides a schematic overview of the single steps that need to be considered when organising an interview project. The highlighted part refers to those steps typically involved in doing a single interview. Depending on the overall design of the study, some of the steps may be recurrent, for instance, when the analysis of interviews informs further sampling and case selection, the process of preparing an interview will start again, now on the basis of the findings from the analysis. If the PCI is part of a multi-method study, the sequence will likely need to be adapted or complemented.

3.1 Research Interest, Questions and Design What do I want to know, and why? And how can I find out about it? This is one of the key questions at the beginning of any research project, even though sometimes the driving force towards systematic knowledge production may not even materialise in the form of concrete questions. However vague, some kind of interest (why?) in a real life puzzle (what?) will have motivated the researcher to engage in the discussion and investigation of a problem and in the search for a way to investigate it (how?). For instance, Maxwell (2005: Chapter 2) distinguishes between personal, practical and intellectual goals; all three can be legitimate aspects of research. Making yourself aware of precisely what inspired the involvement and what you want to find out is important for many reasons. For instance, the design of a study depends upon it as well as the decisions of other people to support your efforts or not (e.g. your supervisor, funding institutions, the ‘field’, your partner, etc.); interests need to correspond to the personal

capability and resources in order to actually realise certain ideas. PCIs as part of dissertations and theses on various academic levels draw on different resources than, say, commercial research or multi-method projects, longitudinal studies or cross-country comparisons. Especially when an academic career is among the goals of the enterprise, much of your prospects depend on how you are able to explain why a certain research interest and question is worth pursuing and how it fits into the academic programmes of departments or faculties. Finally, during the darker moments of writing a thesis, when the initial enthusiasm slackens off, recalling the original, perhaps personal, driving forces may well be the last resort of keeping oneself going. We know of many students who have reached this point, and believe it is worthwhile anticipating and preparing for it. Page 2 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

For some, the idea of designing, organising and structuring qualitative research is often misunderstood as contradicting the principle of openness. Yet like each process driven by human decision, qualitative research consists of a series of choices which can be anticipated and connected in a meaningful way. ‘Research designs’, Creswell (2009: 3) writes, ‘are plans and the procedures for research that span the decisions from broad assumptions to detailed methods of data collection and analysis.’ As plans, research designs make visible which decisions and steps are necessary and how they are interconnected. They are informed by experience and, in general, are driven by a pragmatic interest in finding answers to questions in ways that satisfy standards of scientific inquiry. Compared to quantitative design, qualitative research design is characterised by more flexibility and an interaction of the various elements rather than a linear structure. According to Maxwell (2005: 3), qualitative research design is ‘an ongoing process that involves “tacking” back and forth between the different components of the design, assessing the implications of goals, theories, research questions, methods, and validity threats for one another’. Let us take a brief look at these components in respect to PCI research. The goals of research depend on the specific interest in certain issues and/or their general relevance – ‘Why

is your study worth doing?’ (Maxwell, 2005: 4). The reasons for research can be very diverse and may range from mere curiosity to the basic necessity to save money; as they influence the research process in many ways they can and should be identified and explicated. Often enough the goals of research, and whether they are recognised by others, decide whether research is supported and funded, or not. Also influential is the worldview and research paradigms to which a researcher subscribes, as well as the selection of certain professional goals of social research. For instance, Ragin (1994: Chapter 2) distinguishes seven main goals of social research; each of them is equally legitimate:

1

Identifying general patterns and relationships

2

Testing and refining theories

3

Making predictions

4

Interpreting culturally or historically significant phenomena

5

Exploring diversity

6

Giving voice

7

Advancing new theories.

In order to embed the research project within a broader theoretical or empirical environment a conceptual

framework establishes a preliminary perspective on the issue under investigation. It is informed by theories, concepts, prior research or other relevant experiences – ‘What do you think is going on with the issues,

settings, or people you plan to study?’ (Maxwell, 2005: 4). As discussed below, PCIs operate on the basis of forms of prior knowledge that should first of all sensitise the researcher for relevant aspects of the problem. One of the most important steps in the research process consists in moving from a vague interest and idea at the beginning to formulating and specifying research questions around a certain issue (e.g. Creswell, 2009; Flick, 2007: Chapter 2; Maxwell, 2005: Chapter 4). Research questions specify the exact research focus that Page 3 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

instructs and limits the investigation – ‘What, specifically, do you want to understand by doing this study?’ (Maxwell, 2005: 4). Ideally, one overarching main research question is identified. In addition, it may be useful to try to find an alternative formulation of the main research question that puts the issue in the most general way possible. This increases or even establishes the researcher's sensitivity to the broader relevance of the topic. Usually, research questions need to be modified and further focused as the study moves on. Yet trying to formulate them as precisely as possible already at an early stage greatly facilitates the whole process. Other research aspects, like the choice of appropriate research methods, depend on the research questions. Methods are tools for collecting and analysing information that, after having been interpreted, helps to provide answers to research questions – ‘What will you actually do in conducting this study?’ (Maxwell, 2005: 4). This is the stage at which the PCI needs to be recognised and introduced as an appropriate method for data collection in comparison with alternative qualitative methods than can be considered (see, for example, Bryman, 2008; Flick, 2006; Potter and Hepburn, 2005; Silverman, 2006). This is the case when one wants to investigate actual problems of individuals in a way that develops interpretations and directly clarifies open questions in the process of data collection, i.e. in dialogue with individual respondents. In other words, the PCI and its specific advantages of dialogic knowledge production are being identified as an appropriate way to collect data, either as the only method or as part of a multi- or mixed-methods project. In order to make sure that the quality of research findings is recognised by others, we need to comply with common standards of validity – ‘Why should we believe your results?’ (Maxwell, 2005: 4). This question is typically addressed in qualitative research at different levels and explicated by showing how it is actually done. Finally, the consideration of resources – in terms of money, time, skills, etc. – is an important aspect that can restrict and determine many of the other research design choices (Flick, 2007: Chapter 5). Qualitative research may not be expensive but it is a laborious and time-consuming enterprise, a fact that is often underestimated in preparing for it. PCIs, in the context of single author dissertations – for instance as MA or PhD research – need a different design than PCIs which are part of a multi-method project involving several researchers and teams. Also, PCIs in longitudinal studies or cross-country comparisons, often implicating the translation of original material for the purpose of exchange, require different amounts of preparation and different designs (see section 5.4). In the language of the PCI, at this stage of research, the reasons, knowledge resources and goals for embarking on a research journey need to be explicated, for they shape the researchers’ problem orientation (see Chapter 2). However preliminary they may be, interest, motivation and (personal, practical and intellectual) goals together with concepts, theories and knowledge of relevant research make up the investigator's research capital. This capital, which is part of prior knowledge and can bias and dilute results, is confronted and aligned with findings produced in the course of research. Therefore it needs to be reflected in the course of preparing a study. Once work is in progress, a research diary or the postscript (see section 5.1), a sort of interview protocol and key instrument of the PCI, are important ways of collecting and storing associations that connect oneself to the topic as well as first interpretations resulting from this connection. Page 4 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

3.2 Forms of Prior Knowledge and the Use of a Sensitising Framework What should I know, as a well-informed traveller, before I start interviewing, and how should I organise this knowledge? According to a modern myth Native Americans were unable to ‘see’ Columbus's ships because floating objects of that size and shape were outside their experience. While this is likely to be little more than a legend, it points to an important problem in research: the way we perceive things, and whether we consider them as relevant at all, tend to depend on what we already know about them. The kind of information available in advance (i.e. prior knowledge) often determines the quality of the journey of the traveller, for better or worse. The role of concepts in knowledge production and the relationship between (pre-)conceptions and objects of knowledge have been the subject of many methodological debates (e.g. Erzberger, 1998; Kelle, 2007; Moses and Knutsen, 2007). For instance, Piaget and Inhelder (1971) write in the introduction to their study about the mental imagery of the child:

Knowledge is an assimilative process – the object can be known only by being conceptualized to varying degrees. (Piaget and Inhelder, 1971: xix) What is crucial here for the preparation of interviewing is the fact that any form of (scientific) understanding is possible only on the basis of prior knowledge and in one's own categories, whether they derive from everyday or specialised research knowledge. All research, quantitative and qualitative, faces the same fundamental problem:

We have to accept the fundamental restriction that every observation only takes on meaning in respect of one's own meaning schemata, and so prior knowledge inevitably gives structure to our observations and must therefore be seen as the foundation of all research. (Meinefeld, 2004: 156) As soon as qualitative interviewing by means of the PCI is identified as an appropriate technique for data collection, the principles of problem centring and object orientation instruct the substantive part of the preparation of research before the fieldwork can begin. In short, the texture of the problem defines necessary scope and depth of prior knowledge, while the object of research establishes how and when certain shares of prior knowledge become relevant during the interview. At this stage answers need to be found to guiding questions like these: • What do we already know about the problem in question (from previous research, media, discussions, everyday experience, etc.)? • What are the key dimensions of the problem? Which concepts and theories could be used to frame the issue; which are typically used; and which could provide alternative perspectives? • What do I want to know in addition to all this; and what do I want to find out in the course of my Page 5 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

research? If the research is to follow academic standards, a review of the relevant literature and theories is required in preparation. Applied research may do without this theoretical preparation and instead rely on other sources of knowledge. Prior knowledge of the researcher as a well-informed traveller is stratified. We want to distinguish four types: everyday, contextual, research and sensitising knowledge. All of them are critical in order for the PCI to ‘work’. They contribute to informing the traveller/researcher before he hits the road: preparation helps him to get started, to survey the journey, and it may indicate the way in case he gets lost. Yet, prior knowledge is always ambivalent; it can also easily turn into preconception and constrain the research experience when it is imposed. Each of these types of prior knowledge entails a particular risk in this respect, and the right balance needs to be found.

Everyday Knowledge First of all, any social research builds on everyday knowledge that researchers have about the topic. For instance, researchers of STUDY A, who once were school graduates and are now perhaps parents themselves, have certain ideas about the transition to employment in Germany; and the representation of post-communist unemployment in the media as well as personal observations in the country motivated STUDY C. Everyday knowledge can and should be explicated and disclosed inasmuch as it motivates the general research interest and constitutes the research puzzle. Full advantage should be taken of the superiority of qualitative over quantitative research to recognise the significance of everyday knowledge and not let it become part of some kind of implicit ‘shadow methodology’ (Kelle, 2007: 103). Everyday knowledge entails particular risks being imposed on the research process. Due to the fact that it is omnipresent and usually ‘subtle’ it is often not reflected in its different appearances. For instance, researchers, like everybody else, are equipped with values, convictions and aversions, etc., which can influence and bias their research interests as well as their pre-interpretations during the interview. Also, mere interest or the pursuit of certain research priorities and specialisations, often considered a virtue in the competitive academic world, can bias the interviewer's attention and approach. For instance, think of a family researcher with a personal history of multiple divorces, or a social worker with a personal history of poverty. In all these cases, their (valuable!) everyday knowledge and experience can be a source of bias that needs to be reflected. See Pitfall 3.1 for an example. In addition to ‘genuine’ everyday knowledge, three forms or levels of ‘synthetic’, processed prior knowledge can be distinguished and need to be collected to prepare the interviews – contextual knowledge, research knowledge and sensitising knowledge. The distinction of ‘genuine’ knowledge on the one side and ‘synthetic’ on the other is analytic here; its purpose is similar to the distinction of ‘technical’ and ‘non-technical’ literature introduced by Strauss and Corbin (1990: 4–60). The balance of the knowledge gap between the researcher and the respondent that results from the accumulation of this additional synthetic knowledge on the side of the researcher is then the main challenge of doing the PCI itself (see communication strategies, section 4.6). We discussed the related dilemma of the researcher under the label of Dr Jekyll–Mr Hyde Page 6 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

syndrome in the context of introducing the PCI's main principle of problem centring (see section 2.3).

Contextual Knowledge The object-oriented design of the interview starts from the research puzzle, the research purpose and the circumstances of research. There is no standard solution. Yet whatever the concrete research issue or the units of analysis may be – individuals, interpretive patterns, policies, etc. – the establishment of familiarity with the subject matter needs to be based on basic information about the objective context of the problem. The availability of this most general form of prior knowledge constitutes the minimum requirement

of problem-centred research. Knowledge about living conditions of respondents that determine their options; the history and quantitative aspects of policy reforms; or details about legislation relevant to the topic, are but three examples of background information that researchers have to investigate beforehand. If PCIs are done with experts, contextual knowledge needs to include organisational rules, statutes, etc. of the expert's environment. Such a preliminary exploration of facts and circumstances enhances the understanding of interview statements; it helps to discover and clarify stereotypical or contradictory accounts during the interview; and it provides interviewers with a better chance during the interview to be perceived as competent by respondents. Already Merton et al. (1990) emphasise the advantageous position of interviewers who have done their homework and are prepared. For the purpose of the particular technique of the focused interview, they write: ‘Equipped in advance with an analysis of the situation, the interviewer can readily distinguish the objective facts of the case from the subjective definitions of the situation’ (ibid.: 4). Sometimes, contextual knowledge may be the precondition for being able to make sense of what is said. But even in interviews where this kind of basic expertise does not appear to be necessary it helps to link statements directly to relevant background information already in the course of the interview. Only in this way can the context tentatively be drawn into the conversation and its relevance examined. The extent of the accumulation of this second type of prior knowledge depends on the concrete research question. In any case, we also need to be aware that contextual knowledge is at risk of being imposed on the research process. It influences the final research design and may suggest certain directions of research. For instance, STUDY A (looking at the status passages of young adults to employment) required an in-depth knowledge of the institutional framework of school-to-work transitions in Germany as well as the options available in the system at each stage of the transition. In this way individual transition experiences could be assessed on the basis of knowledge about the stratified and highly standardised German system. The social scientific STUDY B (looking at the cost explosion in the new custodianship of adults) could be designed and carried out only on the basis of a thorough review and comparison of the previous and the current legislation. Basic knowledge about the involvement of additional institutions and parties as well as the broadening of the circle of beneficiaries was necessary in order to specify the research questions. STUDY C (investigating meanings of unemployment in post-communism) had to establish knowledge about the status of work and unemployment in the different societal contexts before and after the collapse of the communist system. Page 7 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Knowledge about the scope of the communist taboo against unemployment and the constitutional right and obligation to labour was able to put current accounts and experiences of unemployment in perspective.

Research Knowledge Basic contextual knowledge is always a filter for delimitating and focusing the review of available studies

and research literature as the third level and type of prior knowledge. It is the balance and utilisation of this kind of knowledge that is most delicate when it comes to problem-centred interviewing: research knowledge provides the researcher/interviewer with a kind of specialised knowledge that both facilitates and biases his or her perception and interest; it strongly influences the preparation of the interview guide as well as the establishment of pre-interpretations of the problem (see Chapter 2). In this sense, research prior knowledge is closer to the phenomenon under investigation. It is usually, but not always, a form of concrete, empirically

contentful knowledge (Kelle, 2005) – i.e. it serves specific knowledge interests and is typically based on findings from research in relevant sub-fields of the researcher's discipline. For instance, an investigation of the political socialisation of party members will be prepared and done differently by political scientists, sociologists or educational scientists. They will also find different answers to similar questions. Their distinctive traditions and approaches not only shape the way in which research is framed and focused; they also largely determine the kind of empirical evidence consulted to build theories within the discipline. One of the risks associated with research knowledge is the danger of the dependence of collected data on theoretical approaches. In order to escape this dependence, and following the principle of openness, one may want to work independently of these theories, or do completely new research into the issue. One might then even be able to fill blank spots of the available research knowledge. In this case, the available research prior knowledge would not be closer to the phenomenon but may even be more remote from it. Research knowledge for the preparation of studies using PCIs is usually collected through the review of the foremost literature of the discipline regarding the topic. At least in academic research, this step of reviewing the relevant literature is indispensable. This applies also to qualitative research where the ‘inductivist self misunderstanding’ together with the ‘naive empiricism of the emergence talk’ (Kelle, 2005: paragraph 24) introduced by certain interpretations of Grounded Theory methodology is largely overcome by now. Knowing one's field of research is widely recognised as an advantage: state-of-the-art knowledge focuses research and improves its quality; and it connects the new research to the evolving scientific discourse. For instance, before the actual collection of data for STUDY A, the literature on professional socialisation and transitions to employment in Germany was surveyed. Applied and commissioned research, like STUDY B, may follow different standards altogether as the greater time pressure does not allow for lengthy literature reviews. Also here, reference is made mainly to contextual and everyday knowledge only. When research is pioneering and exploring novel phenomena, like in STUDY C about unemployment in post-communism, research knowledge is hardly available and thus less important compared to the other types of prior knowledge. In any case, prior knowledge from available research must not obstruct the well-informed traveller in the aim of theory generation. Page 8 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Box 3.1: Expert Knowledge When PCIs are done as expert interviews, the additional type of expert knowledge needs to be considered, which is the very ‘subject matter of the expert interview’, as Meuser and Nagel (2009: 24) maintain. Expert knowledge first of all refers to specific technical and process-related aspects of rules, routines and operations. Usually little is known about it, but some of it may be part of contextual knowledge that researchers can explore in advance. Depending on the background of expertise, experts may have some or even more research knowledge than interviewers; in fact, this may be one of the reasons for interviewing them. In any case, interviewing experts requires thorough preparation because of the higher risk of being seen as an incompetent interviewer. Expert knowledge has, like all knowledge, a subjective dimension. Experts are as much ‘private persons’ as interviewers and as private persons they engage in an interactive process of knowledge constitution and exchange. Their interpretative knowledge of certain aspects of the research issue is as valuable as their technical and process knowledge. Thus, it is necessary to integrate the subjective dimension of experts methodologically. Especially in the frame of theory-generating expert interviews, Bogner and Menz (2009) suggest reconsidering the ‘ideal of neutrality’ that often underlies qualitative interviewing. Instead, the authors recommend strategically making use of a range of possible roles that interviewers can assume during the interaction with an expert. For instance, interviewers can be co-experts, lay-people, authorities, accomplices or critics. These roles go hand in hand with different interviewing styles and may well produce different results; and, furthermore, the way prior contextual and research knowledge is applied also depends on them. Finally, these roles determine the ‘bargaining position’ of the researcher vis-à-vis powerful experts (Obelene, 2009a).

Sensitising Knowledge The specificity of research knowledge, however, may get researchers stuck in the detail: as much as it can facilitate the perception of relevant aspects it can blind for new perspectives on the problem. The latter runs counter to the idea of qualitative research as it bears the risk of undermining the principle of openness. In order not to corrupt the basic ideas of qualitative research, the crucial question for the specific technique of the PCI is how prior knowledge becomes relevant and is played out especially in the course of the interview and during the analysis. This is why we introduce a fourth category of prior knowledge that does not compete with everyday knowledge of the respondent. Following Blumer's (1954) notion of the sensitising concept already introduced above (see section 2.3), we call this form of prior knowledge ‘sensitising knowledge’. Its status and ‘function’ are essential Page 9 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

for problem-centred interviewing.

According to the PCI's principle of problem centring (see section 2.3), prior knowledge must not impede the respondents’ chance to enter the dialogic reconstruction of the problem. What Herbert Blumer writes about sensitizing concepts also programmatically applies to our notion of sensitising prior knowledge: A sensitizing concept … gives the user a general sense of reference and guidance in approaching empirical instances. Whereas definitive concepts provide prescriptions of what to see, sensitizing concepts merely suggest directions along which to look. Hundreds of our concepts – like culture, institutions, structure, mores, and personality – are not definitive concepts but are sensitizing in nature. They lack social precise reference and have no benchmarks, which allow a clean-cut identification of a specific instance and of its content. Instead, they rest on a general sense of what is relevant. (Blumer, 1954: 7) Sensitising prior knowledge is distinguished by its twofold quality. On the one hand, it is of general relevance

to the problem and can thus include more abstract concepts available in the theoretical and conceptual toolbox of particular disciplines. Compared to definitive concepts associated with the kind of research knowledge discussed above, sensitising prior knowledge is characterised by a higher distance to the problem. In other words, it is more abstract and empirically less contentful – i.e. it is open to being filled in with empirical substance regarding the issue under investigation by the respondents’ perspectives. ‘Leadership’, ‘communication patterns’, ‘organisational culture’ and ‘degree of integration’ are all examples of sensitising concepts in the area of evaluation research in organisations (Patton, 2002: 280). On the other hand, the sensitising quality of prior knowledge depends first of all on how it is used. In a heuristic way, sensitizing knowledge suggests directions of research and encourages the progressive refinement and/ or revision of our understanding and pre-interpretations on the basis of problem-centred dialogues with respondents. Possibilities for discovery are maintained. In the following quotation, Denzin describes Erving Goffman's (1963) treatment of the concept of stigma as an example for approaching a research topic and using concepts and scientific prior knowledge in a sensitising way:

He began with a rather vague and loose definition of stigma that he claimed was ‘an attribute that is deeply discrediting’. Three types of this attribute were designated: abominations of the body or physical deformities, blemishes of character (mental disorder, homosexuality, addiction, alcoholism), and last, tribal stigma of race, nation, and religion. Moving beyond classification, he analyzed data collected in such traditional sociological specialities as social problems, ethnic relations, social disorganization, criminology, and deviance. From these areas, relevant commonalities were organized around the stigma theme. (Denzin, 2009: 14–15) The utilisation of sensitising concepts is recommended in order to delimitate the initial contours of the problem without determining its contents. Usually, and especially when research interests are well-defined, concepts for sensitising frameworks originate from within the boundaries of certain disciplines. This can be different

Page 10 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

when the PCI is part of interdisciplinary research where various knowledge domains need to be integrated (Repko, 2008). Also within mixed-methods designs the status of concepts for qualitative research can vary (Creswell et al., 2003; Mertens, 2003). In all cases, this preparatory step of the research process roughly indicates the direction in which the investigation will evolve. The overall aim is to arrive at an attitude of

impartial expertise that is maintained by the interviewer qua well-informed traveller throughout the process of data collection in order to keep it open (see Box 2.4). In the context of PCIs, all prior knowledge – and also everyday, contextual and research knowledge – should be used in such a sensitising way. The main goal of preparing for problem-centred interviewing is indeed to convert all prior knowledge, i.e. all research capital available in advance, into sensitising knowledge (see

Figure 3.2). Figure 3.2 Sensitising prior knowledge

Prior knowledge should then be used in an attitude of theoretical sensitivity as it is propagated in Grounded Page 11 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Theory (Strauss and Corbin, 1990). Importantly, related prior assumptions that introduce a deductive moment into the conversations need to have the chance to be overcome in the process of interviewing. In other words, the sensitising approach should guarantee that the everyday, contextual and, in case of expert interviews, research knowledge of respondents has the chance to enter into a dialogue with that of the interviewers.

Sensitising Framework In big research projects, sensitising the research capital available and required can be quite extensive. In this case, one way of organising prior knowledge in a systematic way involves the establishment of a sensitising framework for research – i.e. a preliminary conceptual and analytic frame of reference consisting of tentative research hypotheses that identify some of the contours of the phenomenon but without addressing their specific contents. A sensitising framework constitutes an additional outer layer of knowledge framing the researcher's perspective. It embraces his or her prior knowledge that specifies the epistemological challenge of problem-centred interviewing illustrated in Figure 2.1. Coming back to our metaphor of the problem-centred interviewer as a well-informed traveller (section 2.2), the sensitising framework fulfils the function of a preliminary roadmap that roughly sets out where the journey goes within a certain territory. Especially in explorative research where little is known about the phenomenon and the risk of getting lost is particularly high, such a roadmap is useful for getting research started and for guiding fieldwork. A sensitising framework can facilitate many of the steps in the research process including the preparation of the contents of the interviews and the interview guide. However, it may not always be necessary, or it may need to be even narrower in more strongly-focused research. In principle, the ‘place’ of emphasising the use of theory in the process of research is variable, at least in qualitative and mixed-methods research designs. Creswell (2009: 65–9) maintains that conceptual models at the beginning of research are particularly appropriate for studies with a cultural theme or a theoretical lens, while an emphasis on developing theory postpones the introduction of theory and concepts. Maxwell (2005) suggests conceptualising qualitative research design as an interaction of the following design components: research questions, goals, conceptual framework, methods, and validity. In this understanding, design in qualitative research is an ‘iterative process’ where also the use of theory and the consideration of concepts are subject to the necessity to move back and forth between these elements. Wengraf's general advice concerning the introduction of the ‘right’ amount of theory at the beginning of research is useful here:

[O]ne should start with a coherent and potentially productive Conceptual Framework with a Central Research Question and subsidiary Theory-Questions that do not foreclose the further theorizing and

conceptual innovation that they will, one hopes, trigger off. (2001: 93; original emphasis) Box 3.2 explicates all these levels of prior knowledge on the basis of the sensitising framework of STUDY C. In other cases this might be less complex; yet prior knowledge concerning context as well as available research and relevant concepts is in any case necessary for the design of PCIs. Page 12 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Box 3.2: Main Research Question and Sensitising Framework (Study C) STUDY C used PCIs in the frame of a PhD thesis investigating meanings of unemployment in the post-communist context of Lithuania from the perspective of young people in transition to the world of work (e.g. Reiter, 2008a). The research puzzle and societal problem occurred here owing to the fact that the transformation from communism to market democracy brought an end to decades of full employment. After the collapse of communism, (mass) unemployment became a novel problem in society and was first observed by the second author in the perspective of everyday knowledge. The introduction of the societal figure of the unemployed also changed the criteria of citizenship and social integration and exclusion. Furthermore, for young people, unemployment became a new option and a source of uncertainty in their transitions to employment, which was previously rather linear during communism. Owing to the novelty of this phenomenon the study was explorative. It proceeded mostly in an inductive way along two axes of analysis which were introduced at the very beginning in the sense of a rough limitation of the field (see below). In addition to contextual knowledge about Lithuania before and after the collapse of the Soviet Union, the introduction of sensitising concepts (SC) was particularly important to frame and orient the research. The study was done in the frame of a structured PhD programme, and the development of an extended research proposal during the first months of the project was required (and supported!). The proposal included a sensitising framework that is briefly introduced. At first, the general interest of STUDY C in post-communist societal transformation and the emergence of the problem of unemployment during the country's transformation to a capitalist market economy did not suggest any specific direction of research. The value of everyday knowledge and experiences was very limited because of the researcher's rather high degree of unfamiliarity with the context. Due to the historical distinctiveness of the research puzzle and the uniqueness of the event, research knowledge was not really available or of little help. Most of the social scientific research into post-communism was dedicated to the transformation of institutions like the labour market, the welfare system, democratic representation and the like. Thus, much of the available literature was written in the perspective of political economy and democratisation research. But the abolishment of full employment and the ‘rapid rise of unemployment from zero to double-digit rates’ was not only ‘one of the most fundamental economic phenomena of the twentieth century’, as Svejnar (1999: 2852–3) puts it in the perspective of labour economics. Against the background of a very general research question - i.e. ‘How is the new phenomenon of unemployment accommodated within the accounts and reflections of

Page 13 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

young people in transition to the world of work?’ - at least two other aspects were identified as relevant, for which research knowledge was missing. On the one hand, the societal transformation also fundamentally shifted the biographical perspectives of citizens in these countries: unemployment became a new and probable life-course risk. In the sense of a working hypothesis on the biographical level, one could assume that the novel threat of unemployment puts young people who are on their way to the world of work into a situation of ‘biographical uncertainty’ (SC) (Wohlrab-Sahr, 1992). The lack of guidelines for orientation in an increasingly complex societal environment, the awareness of this complexity, and the related feeling of uncertainty concerning self-status make individual lives and life plans essentially unpredictable. School-to-work transitions have the character of ‘fateful moments’ (SC) that Giddens (1991: 243) defines as ‘moments at which consequential decisions have to be taken or courses of action initiated’. On the other hand, the emergence of mass unemployment redefined the relationships between citizens and the state and between people themselves; it introduced new criteria of social inclusion and exclusion. In terms of a working hypothesis on the normative level, one could expect that the idea of the employment-centred life course - but without employment guarantees, which distinguishes the new market regime from the former communist system - is now becoming relevant as a reference category for the orientations of young people. It introduces the participation in productive, formal employment as the precondition for social citizenship and ‘recognition’ (SC) (Honneth, 1996). In this respect, the transition to the world of work constitutes a ‘status passage’ (SC) (van Gennep, 1981); it links ‘institutions and actors by defining time tables and entry as well as exit markers for transitions between social status configurations’ (Heinz, 1996: 58–9). These research hypotheses and sensitising concepts were combined into an analytic and conceptual frame of reference. It was developed before the actual fieldwork started and constituted the luggage and roadmap of the well-informed traveller. The framework is visualised in Figure 3.3. Only the lighter grey area was addressed on the basis of the empirical data that was collected with PCIs. The idea was to observe societal transformation indirectly by means of investigating normative and retrospective biographical aspects of youth transitions to the world of work. All the elements of the framework had the status of general, abstract and empirically not contentful concepts that needed to be brought to life empirically.

Page 14 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 3.3 Sensitising framework (STUDY C)

The PCI was chosen for data collection because it supported limiting and reducing the otherwise very broad topic. The topical guide (see Figure 3.4) included aspects relevant to these two levels of analysis. The interviews explored the respondents’ own experiences, allowing reconstruction of their biographies and transitions; and they addressed their opinion about unemployment and unemployed people, allowing reconstruction in related interpretive schemes. Furthermore, as the PCI is a technique that is closer to everyday conversation than, for instance, the narrative interview, its use facilitated the preparation and training of the interviewers (see Box 3.3).

Page 15 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 3.4 Topical guide (STUDY C)

3.3 Interview Guide How do I prepare questions and topics for the actual process of interviewing? As soon as the direction of research is consolidated, research questions are sufficiently clear, and problemcentred interviewing identified as (one of) the suitable method(s) of data collection, then the more applied part can start. At this stage the right balance needs to be found between the ambitions of the project and the available resources. In particular, the development of the interview guide is immediately related to the sensitising framework and thus the most important bridge between the researcher's interests and the field. The PCI should not be mistaken for a conventional guided interview. Aaron Cicourel's remarks, given in an interview, are instructive here when he points to the ‘drawbacks of using fixed-choice and open-ended questions that were not reflected in the respondents’ daily life experiences, including their understanding of each question and the kinds of knowledge domains each question presupposed’ (Witzel and Mey, 2004: paragraph 53). Falling back into the conventional question-answer pattern is a possible consequence of inappropriate questions (see Pitfall 4.1). The problem-centred approach to using an interview guide and the attitude of the interviewer as well-informed traveller require the consideration of a few specific aspects. The main function of the interview guide in PCIs is that of securing both the problem centring of the interview and the comparability of the individual interviews. The interview guide thematically organises the researcher's background knowledge and interest. It should facilitate a controlled approach to the object of research in Page 16 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

the course of the face-to-face interview and establish a certain comparability of the resulting interviews. Especially for beginners, the interview guide is also an aid to memory that provides the conversation with some structure. It can serve as a backup tool in case communication is difficult and, if applied carefully, it helps the respondent to develop narrative sequences. The interview guide should cover all aspects of the problem in the form of single thematic fields that are further differentiated by keywords and categories or in terms of exemplary pre-formulated questions. However, the structure and inner logic of the interview guide as well as the order of topics and suggested directions of introducing questions are supposed to provide only preliminary guidance. They are based on prior knowledge, research questions and, eventually, experience from pilot interviewing. Following the principle of problem centring, the actual conversation should then rather be ‘guided’ and structured by the topical priorities that the respondent establishes in the course of the conversation. In other words, ideally it is the topical guide and thread suggested by the respondent that is the centre of attention; the interview guide is the researcher's tool that accompanies and facilitates this process in the background. Figure 3.4 is an example of a topical guide from STUDY C. Similarly, Kurz et al. (2007: 471) suggest visualising the interview guide by means of mind-mapping. In practice, in the course of the conversation, interviewers register the issues brought up by respondents that overlap with the interview guide. They tick them off in the mind's eye inasmuch as they provide answers to underlying research questions. In this way, it is possible to supervise how individual elements are brought up and worked through in the course of the PCI. In case the conversation is unproductive, slows down or comes to a halt – which is very likely to happen at some point during the interview – the interview guide is a source of possible stimuli as regards contents that can be brought up and formulated ad hoc and depending on the situation. In this way, interviewers are able to scan relevant topics in addition and in relation to the narrative thread developed by the respondent. This interaction should allow detection of fruitful topics for further exploration and a check of their relevance for respondents. For interviewers this can become a very challenging task. On the one hand, they need to keep track of the respondent's thread of argumentation and identify and collect appropriate follow-up questions. On the other hand, they need to decide on the appropriate moments, during the interview, for certain issues to be differentiated (with regard to the problem-centred research interest) by introducing follow-up questions (see section 4.6 for the discussion of follow-up questions). This twofold pressure to decide for (or against) following up, and introducing additional topics while the conversation is evolving, is the source of many interviewing mistakes. Stepping back from the immediacy of the ongoing interaction and exchange of information while being involved in producing it is difficult and requires experience. In order to improve one's sensitivity for the overall process of problem-centred conversations, less practised interviewers in particular benefit from intensive interview training, which is also a means of appropriating the interview guide (see below). However, one should abstain from formulating standardised questions or from using exemplary pre-formulated questions in a standardised way. ‘Adherence to fixed questions’, as Merton et al. (1990: 47) call this specific fallacy in interviewing, may unnecessarily limit the extent to which relevant Page 17 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

data is collected; and it may corrupt the development of the interview in general. Nonetheless, there is nothing wrong with learning from experience and collecting clever formulations of key questions. In ethical terms, it may even be advisable to prepare the wording of questions, introducing particularly sensitive issues into the conversation in advance. They can then be adapted and modified as required in the situation. In any case, this form of standardisation – of the question, not of the answers! – should be kept to the absolute minimum and reserved for these occasions. Apart from topics and questions, interview guides can also include sample introductory statements as well as brief instructions for interviewers (for examples from STUDIES A and B see Appendices 1 and 2). Limiting the interview guide to directions of questions, rather than including specific pre-formulated questions, altogether helps to reduce the artificial character of the conversation. Besides, the suggested sequence of gradually clarifying the problem in conceptual terms, which is implied in these directives, hardly coincides with the respondent's perspective on the problem. This gap can be used in a productive way and support the full flexibility of the researcher/interviewer in addressing and following-up issues in the frame of a problemcentred dialogue. In other words, this flexibility helps, on the one hand, to immanently link clarifications and specifications to the material already provided. On the other hand, according to the situation, additional topics can be introduced through additional follow up questions by referring to the problem. The interview guide should be short, not overloaded, and well-arranged. One of the most common ‘performance errors’ in qualitative interviewing is the tendency to dismiss opportunities to reconstruct issues dialogically with the respondent, by superficially and mechanically ‘executing’ the agenda of the interview guide and ‘ticking off’ questions. This problem, referred to as ‘interview guide bureaucracy’ (Hopf, 1978 and 2004: 207), is more common when the interview guide actually includes pre-formulated questions. Interviewer training helps to reduce this error but it is best avoided altogether by refraining from providing such questions. Instead, the main topics that should be covered during the interview can be organised in a topical guide. Lacking experience, unexpected turns during the conversation, or stress are among the reasons that can cause interviewers to fall back into this conventional question-answer scheme. Pitfall 4.1, which is introduced after we describe the PCI's communication strategies (section 4.6), takes a more extensive look at how adherence to points in the interview guide can become problematic. The actual categories of research have to be adapted to the everyday language of the respondents. This general requirement of translating categories and questions for the interview needs to be addressed as early as possible in the preparation. Remember:

Research questions are not interview questions! Research questions are ‘theory-questions’ as Wengraf (2001: Part II) calls them in his extensive discussion on preparing appropriate interview questions for structured in-depth interviewing. PCIs will likely fail when the translation of research questions and categories is omitted and the issues are not addressed in common language. Interviewers that force the discussion of certain issues, and don't address the topics in the language of the respondents but instead reproduce categories of research knowledge, will be unable Page 18 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

to collect relevant information (see Pitfall 3.1). Furthermore, the communication between interviewer and respondent, and the conversation they have, will probably not flow as it should because the respondent may not be able to relate to the researcher's expressions. This is a very common error at all levels, especially among researchers who do not usually establish their arguments on the basis of ‘talking’ to people. Wengraf (2001: 156) further suggests, in the sense of a general rule, that the ‘real’ research purpose should not be communicated in advance for it will ‘contaminate’ the interview. This may be a slight exaggeration, but an information strategy or policy towards the respondents should indeed be part of the preparation for an interview.

Pitfall 3.1: Direct Use of a Research Question as an Interview Question The following passage is taken from one of the interviews for STUDY A. Soon after the respondent talks about the situation in the apprenticeship market for hairdressers, the interviewer forces the gender topic on the respondent.

I: Yes, yes, uhm … did you experience the fact that you are a woman in an advantageous way for the search [of an apprenticeship place], or do you think … that was … that did not play a role, gender so to say? R: No why? Well … I: No? R: No, well, I do not quite understand this question now. The interviewer is a committed women's advocate and she gives priority to gender issues throughout the rest of the interview. In this passage she tries to introduce this important aspect of the interview guide at the very beginning of the interview by using the related research question as an interview question. Unsurprisingly, the respondent does not quite know how to deal with the question.

3.4 Interviewer Training, the Problem of Un/Familiarity, and Peer Research How, and why, should I prepare and train other interviewers? The importance of prior knowledge and its management in the methodological foundations of the PCI emphasises the critical role of the interviewer. One feature of the PCI's core principle of problem centring

Page 19 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

is the ideal of the researcher/interviewer - i.e. the researcher who is familiar with all aspects of the study and especially the sensitising framework is also doing the interviews. However, often this is not possible. For reasons of research ethics, expertise, social distance, academic culture, lacking resources or a crossculturally comparative research design, this ideal does not always survive the practical and pragmatic requirements of research. For instance, interviews with traumatised victims of violence must be done by interviewers experienced in dealing with emotional situations. Certain topics that involve a high degree of specialised knowledge are better studied by means of peer

research (see below). This may also be appropriate for the study of rather closed social groups like criminals or youth cultures. Academic cultures may intervene when, like in Germany, professors are hardly ever involved in doing interviews (or research) themselves, even if they are officially the main researchers. Instead, as there are no permanent research staff employed by universities, research is mostly conducted by temporary researchers preparing their MA or PhD theses in the frame of the project. Often, perhaps for reasons of time, simply not all of the interviews can be done by one person. Finally, with Europe opening up and systems of higher education being standardised, ideas and research questions move across borders together with students and funding opportunities. For research, this means that issues of culture and language add to the complexity of investigating research puzzles in largely unfamiliar contexts. Cooperating with local research partners and teams has become standard practice, and interviewer training has become a basic requirement. In most of these cases where the researcher is not doing (all of) the interviews, various degrees of

unfamiliarity regarding either the sensitising framework or ‘the field’ need to be dealt with. This has important implications for research in general. For the purpose of designing PCIs that intend to reconstruct societal problems in a dialogical way, the leading question is:

How should the interviews be prepared when the researcher/interviewer is not familiar with the topic – i.e. when the perspectives of the respondents are particularly vulnerable to being misunderstood and misrepresented? One way is the management of prior knowledge, already discussed above, by means of using empirically less contentful, abstract concepts and sensitising frameworks. This can be a good solution even for cross-cultural studies. Such studies usually deal with the highest degree of unfamiliarity, and perhaps the lowest degree of concept suitability, which is taken for granted in the context of our professional socialisation. Regardless of where we actually live, it seems, the ‘state of the art’ of our research is very likely dominated by Western concepts and findings. There is little discussion about the general validity of ‘our’ results, however ‘weird’ they may be (Henrich et al., 2010). We are most familiar with our own research environment – the consideration of contributions in languages other than our mother tongue may require huge additional efforts; or, they may be completely out of sight. Even experienced researchers are increasingly unable to keep track of the abundance and diversity of academic publications in their field of expertise. A bias in research topics and concepts can also be due to research funding policies suggesting a canon of references to be considered, or due to global cultural homogenisation dictating priority perspectives. Just think of the catchwords that Page 20 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

have dominated European research during the last few years: social exclusion/inclusion, active citizenship, sustainability, civil society, participation, quality of life, governance, etc. If this isn't the case, how is it that the regional government of Podkarpackie in south-east Poland is suddenly working on a regional development strategy on the basis of SWOT-analyses and other contemporary world-cultural consulting tools (Büttner, 2012)? Many of our concepts and terms to address phenomena are unquestioned and taken for granted; they can strongly influence and blur our research perspective and our pre-interpretations. In order to comply with the principle of openness we need to deal with the fact that our research interests are shaped and limited by these factors and our very specific academic culture. Sensitising frameworks, if applied in the right way, are sufficiently general and leave space for cultural particularities to come to the fore. Unfamiliarity regarding the conceptual frame of reference, for instance when external interviewers are involved, is best reduced by intensive interviewer training. As a general rule for PCIs, external interviewers must become familiar with the particular interviewing technique as well as the details of the overall approach and the research capital it involves. They need to understand that interviewing is always part of the process of interpretation for it highlights certain aspects and neglects others. Interviewers need to be sensitised to the fact that they are never instruments of data collection only and that interviewing is, in the sense of the PCI, an interactive, socially-embedded enterprise. When research is carried out by a group of people, members of a research team need to agree largely on their approach to the field of study. In this way they are able to reduce doubts concerning the relevance of certain issues and the intensity of their exploration. Only fully knowledgeable interviewers are able to make competent and independent decisions during the interviews. In terms of interviewing skills, workshops and other forms of practical training should be organised, and there is a broad range of literature giving a general introduction to qualitative interviewing (e.g. Fielding, 2002; Gubrium and Holstein, 2001). Kvale and Brinkmann (2009: 92–3) suggest learning interviewing by witnessing others interviewing, by practising interviewing and listening to the interviews, and by being part of a community of interviewers. Initial lack of experience can be reduced by using co-interviewers during the first interviewing encounters. And experienced researchers should give advice, support and control throughout the initial stage of the data collection. On the basis of transcribed interviews, methodical issues can be analysed in relation to the development of the problem in the course of the interview. Interviewer training should also include trying out various introductions to thematic fields based on hypothetical situations. As research moves on there will be experience with questions and formulations that work or do not work with certain groups of people. This experience is most valuable for interviewer training and should be used. Experienced interviewers should also, in any case, be made familiar with the particularities of problemcentred interviewing, which can also be refreshed in the course of such forms of training. In terms of contents, interviewers for PCIs should at best co-participate in the design and preparation of the study; at least, they need to be introduced to the sensitising framework and the necessary background knowledge distinguished Page 21 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

above. In order to use this prior knowledge competently, training for problem-centred interviewing must go well beyond learning the general practical skills of qualitative interviewing. The training for interviewing is at the same time a training of interpretation: interviewers need to understand and become aware of their preconceptions; how they may influence the flow of the conversation; and how they can be used in a controlled and productive way. In workshops and on the basis of examples from their own interviews, interviewers should discuss and criticise the impact of preconceptions in each other's interviews. Ideally, training and reflection continue until research is concluded and help interviewers to improve their overall methodical sensitivity and their capacity to focus on the topic in particular. In cases where the unfamiliarity is in the field rather than the conceptual background (or it is in both), elements of peer research can be integrated. Peer research employs members of the field in the process of data collection (a strategy that may be particularly interesting for expert interviews like in STUDY B – see section 6.2); think of a hospital nurse interviewing colleagues, or a young person interviewing peers in a night club. Peer research also has its downsides; for instance, think of a businessman hired to interview competitors in the field. There are many practical and ethical issues to be considered when using peers for research, for instance, remuneration, power and hierarchies, risks to anonymity, etc. They cannot be addressed in detail here but see Burns and Schubotz, 2009; Coar and Sim, 2006; Elliott et al., 2002; Hockey, 1993; and Platt, 1981. However, for our purpose of discussing how to best prepare and use prior knowledge in PCIs, peer research is a particularly interesting option. Peers represent a form of insider knowledge and ‘natural’ expertise that can be utilised beneficially; they are also familiar with the conventions of communication and hardly have problems getting access to the research environment. In case there is little time to prepare, for instance, due to time pressure in commissioned research like in STUDY B, involving well-trained peer interviewers can also be a shortcut through a lot of preparatory reading and can simplify the sometimes complicated access to the field, including getting in touch with individual respondents. By definition, peer research involves (external) interviewers that do not usually have any interviewing experience. Thus, intensive interviewer training, including content as well as practical skills, constitutes the base on which specific PCI training can be built. The crucial questions are:

How can ‘too much’ (or too specific) prior knowledge, which peer interviewers may have, be controlled? Or, how can it at least be moderated so that there is a chance that pre-interpretations are challenged? In the frame of the terminology developed above, what is specific when a peer (in the role of interviewer) enters into a PCI dialogue with a fellow? First of all, the respondent's practical knowledge will be more familiar to the interviewer than in a non-peer PCI, and the knowledge asymmetry will thus be less distinct (see Figure 2.1). The large overlap in their everyday and contextual knowledge will facilitate the ease of communication between them and the ease of arriving at a common interpretation of the problem and its dimensions. However, the research knowledge of peer interviewers will be rather limited (Figure 3.2); and Page 22 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

they may be biased towards preconceptions that can result in simply not asking certain questions. In other words, what may qualify as findings for the researcher not involved in the interview may be commonsensical and everyday knowledge for the peer interviewer and thus remain implicit and unexplored. When experts of the same field are employed as peer interviewers, like in STUDY B, their expert knowledge, whether hard (e.g. facts) or soft (e.g. codes of conduct), should be utilised carefully. The main problem consists in getting the peer interviewers to use their prior knowledge in a sensitising attitude. This is precisely what the training and preparation of peers for PCI interviewing needs to focus on in order to make the most of their familiarity with the issue. In practical terms and as well as getting familiar with research questions, design and technical equipment, peer interviewers need: • to learn how to handle their insider knowledge according to the requirements of the PCI; • to reflect and control their privileged ability to influence the conversation; • to be trained in using the auxiliary means of interviewing, including the interview guide and other guidelines. Providing peer interviewers with additional research knowledge, however, is of secondary importance and may hardly be feasible in case of commissioned research. In this case, the substantive and scientific prior knowledge of peer interviewers necessarily remains limited.

Box 3.3: Interviewer Training and the Problem of Un/Familiarity (Study C) The PCIs in STUDY C were conducted by two Lithuanian native speakers, both at that time MA students of sociology. Native speakers were chosen in order to be sure that the interviewers were able to follow the details of the accounts without restrictions, and because they were closer to the cultural and linguistic environment of the young respondents. Aspects of peer research were relevant here as well. The familiarity of the interviewers was important: in this way, the already-limited willingness of the young Lithuanians to participate in the interviews and to openly recount their lives was not further undermined by cultural differences or the burden of potential misunderstandings. The interviewers were familiar with qualitative interviewing techniques and experienced from their studies. A very first preparatory meeting took place some six weeks before the actual interviewing workshop; it introduced the purpose of the study and a cooperation agreement was established. Afterwards the interviewers were provided with relevant parts of the extended project outline and a short summary article about the method of the PCI.

Page 23 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The interviewing workshop then took place immediately before the first PCIs were done. It consisted of four main parts. The first part was an introduction and discussion of the study, its detailed research questions, and its conceptual background. The second part covered technical and communicative aspects of problem-centred interviewing, the use of the interview guide, and issues related to doing the interview. In the third part, practical issues were discussed, such as access to the institutions that had already been contacted in advance, the use of tape-recorders, and transcription conventions. Finally, the complete interview guide and all stages of the interview were discussed. The basis was a flow chart similar to the one in Figure 3.1. Special attention was dedicated to the interview guide for the main part of the conversation, i.e. the opening question and the follow-up. In the case of this study, it had the form of a minimal topical guide (see Figure 3.4). The flow chart translated the conceptual framework as well as the research interest into more concrete yet still abstract issues, which were extensively discussed during the interviewing workshop. The main issues for follow-up were organised around the initial contribution of the respondents. In this case, the opening question invited the young respondents to talk about the last five years of their life including their (linear or non-linear) transition to employment (see the example in Box 4.2). Follow-up questions should first give priority to the main issues addressed during the initial account - for example, the last years at school, the transition to university, or the struggle to find a job - and explore their potential for contributing answers to the research puzzle. This was complemented by questions dealing with other issues of the topical guide. The order of addressing them was secondary; all were equally important. The discussion of the structure of the interview guide and exemplary questions related to each of the topics resulted in the modification and revision of these questions and their wording. A further moment of ‘appropriation’ of the task by the interviewers was their spontaneous involvement in the revision of a first translation of items included as a short questionnaire at the end of the interview (see Box 4.7). An additional working meeting was held where software for transcription and for converting tapes into digital format was introduced. The first interviews were conducted in the days immediately following the preparatory workshop. During this initial period, the contacts with relevant institutions and key people were established by the researcher who negotiated access and introduced the project and the interviewing team. After each of these first interviews, the researcher met with the interviewers to discuss experiences, potential problems and possibilities for improvement.

Page 24 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The cultural closeness of the interviewers to the respondents and their familiarity with the overall living conditions required the reflection of their own preconceptions concerning certain issues of coping with the societal transformation. For instance, several of the young respondents openly talked about their intention and reasons to leave the country. The reasons could range from lacking opportunities and disappointed aspirations to anger and resignation regarding the current situation. The interviewers, on the other hand, had their own, sometimes strong opinions about emigration, which was altogether a very controversial issue in the country at that time. This ‘conflict of interest’ was then discussed in the team in order to reduce possible interviewer biases. A second aspect of context familiarity also needed to be addressed; it concerned the fact that, at least to the ear of the foreign researcher who was unfamiliar with the context, some important issues remained unexplored. The overlap in familiarity of the local circumstances for both interviewer and respondent sometimes resulted in insufficient detail. In other words, the interviewers had occasionally forgotten that they were collecting the information for somebody else, who did not know the context so well. What was taken for granted and was everyday knowledge to them was precious information to the researcher. It needed to be explicated and documented during the interviews, otherwise, the researcher would have very limited material with which to reconstruct the problem afterwards.

3.5 Sampling, Access Who should I interview, and how should I approach my respondents? The PCI is best combined with some form of selective or purposive sampling on the basis of empirical advance information, for instance from expert interviews. The selection of respondents can be done on the basis of a preliminary sampling plan around rather global categories identifying important sub-groups in the population. For most questions, gender, age and social background are among the standard distinguishing categories. Depending on the research focus, others may be more relevant (e.g. ethnicity, place of residence, organisational position, expertise, etc.). For instance, in mixed-methods projects, sampling can be informed by statistical analysis of secondary data, or even an original quantitative study preceding a PCI-based indepth investigation of, possibly, a sub-sample of the quantitative study. Alternatively, and in line with Grounded Theory, theoretical sampling (Strauss and Corbin, 1990) based on theoretical assumptions and empirical findings can be applied. Theoretical sampling is the ideal way of selecting respondents as it best corresponds to the PCI principle of process orientation. Rather than just saturating given categories empirically, theoretical sampling aims at discovering categories that are relevant to the topic as well as their interconnectedness. Thus, it may be preferable to base the selection Page 25 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

of respondents on different situations or social configurations rather than certain categories. Situations and interactions associated with them (e.g. lone motherhood, winning on the lottery, incontinence) can draw other people or social networks (e.g. parents, peers, partners, colleagues, friends, media, etc.) into the sampling process. In case of theoretical sampling, the deductive moment at the beginning of the research process does not automatically establish an arbitrary cut-off point in terms of data collection. Instead, data collection is kept open: later sampling steps are prepared and informed by earlier ones and by preliminary data analysis; findings and emerging (grounded) concepts and theories are gradually consolidated, saturated and validated in the course of research. Such a research process is able to integrate sampling, analysis and interpretation; these are not isolated steps on the way to ‘findings’ crystallising only at the end of the process. The aim is to differentiate and saturate the categories identified on the way. Furthermore, the requirements and direction of the analysis feed back into and co-determine the selection. Similarly to the ideal of the researcher/interviewer discussed above, concessions to the practical reality of research are sometimes necessary in respect to the selection of respondents. For instance, the final sample usually depends on available resources for data collection, transcription, possibly translation, and analysis (see section 5.4 for a discussion of resource-sensitive planning of PCI research). The considerate composition of a sampling plan can help to narrow down the horizon of selecting cases from the beginning. This provides possibilities for sampling for analysis within the original sample and for strategic data analysis that utilises the full potential of the realised sample. Sampling is intertwined with practical and methodological issues of field access. In particular, getting in touch with respondents is a delicate, often neglected and rarely reflected moment in qualitative research or specific methods. In methodological terms, contacting potential respondents is also part of the epistemological challenge of the PCI (illustrated in Figure 2.1) for several reasons. First, researchers and respondents bring along very heterogeneous expectations about the actual interview/ research interaction. Usually, both inhabit separate life-worlds that confront each other during the conversation. Second, researchers and respondents have different interests and expectations. During an interview, both take these interests into a temporary relationship, which is clearly identified as a research relationship (with all its consequences) only by the interviewer. The respondent, on the other hand, is likely to have no idea of what research is, and may define the situation as an event ranging from a casual encounter to a tense and disturbing exam. The researcher should clarify the purpose of the interview on several occasions and make sure that the respondent arrives at a good understanding of the situation, however incomplete it may remain. The respondent needs to understand why and how she can actually contribute to the interviewer's research interest; she is invited ‘to “work at” her responses’ as Cicourel (1974a: 198) puts it, and to elaborate on them. Especially during the phase of getting to know each other and warming up (see section 4.3) the interviewer can find out about the definition of the situation by the respondent. Sometimes the researchers themselves Page 26 of 27

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

have multiple roles and conflicts of interest (Bell and Nutt, 2002). For instance, in peer research, interviewers constantly have to negotiate their various perspectives on the issue. The richness and validity of the interview depend upon the degree of social desirability biasing the interaction as well as on the ability of the participants to manage their possible multiple roles. Third, when access to respondents is mediated by a third party – e.g. parents, teachers, schools, bosses, employers, etc. – the status of this party, as well as its relationship to the respondent, needs to be considered. For instance, access to young people, which depends on the consent of parents, teachers or social workers, requires the consideration of different issues than, say, interviews with employees. The former may cause misunderstandings with regard to authority, while the latter may involve conflicts of loyalty. Fourth, the moment of getting in touch with respondents is the first occasion to actualise the problemcentring character of the interview. For the first time, research questions need to be translated into issues connecting to the everyday life of the respondents. The success of the interview is contingent upon whether the respondent is able to relate to the researcher's interest in a way that is relevant for her i.e. the ‘problem’ needs to be a common one. In practice, at the early stage of the interaction, researchers need to succeed in establishing respondents as ‘experts’ who are motivated to reconstruct the topic in their perspective. Interviewers assume only a supporting role. The ‘expertise’ of respondents is interactively established through the way interviewers relate to the respondents. This aspect of expertise on the part of the respondent and strangeness on the part of the interviewer (that remains an important part of the whole interview) is particularly critical for the introductory explanation and the opening question: for the first time, the space of ‘the problem’, whose meaning is negotiated in the course of the interview that follows, is outlined more concretely (see later). http://dx.doi.org/10.4135/9781446288030.n3

Page 27 of 27

The Problem-Centred Interview: Principles and Practice

Doing PCIs In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 64-94 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Doing PCIs 4.1 Interview Situation, Setting Where should I do the interviews? The choice of the location of the interview and the arrangement of the setting should contribute to facilitating intensive dialogue. It is part of establishing a framework within which the reconstruction of the problem can unfold. The PCI's principle of object orientation becomes particularly relevant here. What kind of context does the investigation of a specific issue (from the perspective of specific respondents) require? Apart from avoiding noisy and busy places, other requirements may need to be considered depending on the interview topic as well the respondents. For instance, certain ‘problems’ with a high degree of personal and emotional involvement ask for, among other things, particular sensitivity regarding the location. In such cases, institutional structures of support that continue beyond the interview (e.g. social work organisations) are preferable settings, even at the cost of other unavoidable disadvantages and biases necessarily attached to the dynamic of organisations. Experts often have offices at their disposal, the homeless will not; yet the latter may feel more comfortable in a public space. In both cases, the qualities of the site pervades the interview; thus, the interview context becomes part of the investigation. Finding out what suits respondents best and still allows for decent recording is a basic part of the process when confronting the features of the two lifeworlds of the researcher and the respondent. Often, ethical issues, perhaps in general involving managing the visibility of the research, need to be considered at this stage. Who should, or should not, know about the interview, let alone overhear some of its contents? This can be a very sensitive matter in certain situations, for instance, in organisations as well as family networks. Box 6.3 discusses some impressions related to the place of interviewing in STUDY A. Offering respondents a number of alternative locations for interviewing can help in the selection of an appropriate setting. For instance, interviews in the private flat or house can provoke the involvement and control of partners, family members or parents. Respondents who chose such a private setting may well be aware of this. On the other hand, interviews at the workplace may prove hard to record but can provide unique insights into the everyday life-world of respondents. For instance, during an interview in the evening in his shop a self-employed shoemaker continued to fix shoes after business hours. He had to be asked to consider the tape recorder and to perform tasks, like gluing soles, which did not produce noise. At the same time, the researcher/interviewer had the opportunity to witness typical aspects of how self-employment in a small enterprise can deliver scant reward.

4.2 Introductory Explanation and Briefing What should I say about the project and the interview? Page 2 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

In case no prior exchange has taken place, briefing the respondent about the background, purpose and nature of the interview can be done, or repeated, at the beginning of the interview. A good introduction captures the respondent's attention, and it is the most important source of his commitment to the topic and the interaction. Apart from introducing oneself and the institutional context of the study, some other organisational or substantive details may be revealed. This is also the place where the ‘knowledge problem’ on the part of the researcher (quasi as a stranger to the field) can be introduced as the reason for conducting the interview, with the respondent as the (everyday) expert. At this moment it is useful to point to the very asymmetrical communication patterns that interviewing usually involves. The specificity of the PCI requires that the respondents understand the importance of extensive narratives and that the conversation does not simply follow the regular question–answer pattern of conventional interviewing. However, interview partners should not be confronted with original research questions or the conceptual framework – this must not be the respondent's burden to carry! Pitfall 3.1 describes such a situation. In terms of communication, the introductory explanation is the moment where respondent and interviewer agree on a ‘contract’ for the content of the interview as well as their respective roles in the interaction. The clearer the introductory explanation is, the higher the respondent's resilience to interviewing mistakes will be. It may even help to compensate for inappropriate questions (see Pitfall 6.9). Ethical standards of qualitative research and of qualitative interviewing are increasing in importance (Kvale and Brinkmann, 2009: 61–79). For good reason, project evaluation committees usually check proposals for potentially problematic aspects regarding research ethics; often, proposal forms for funding applications explicitly ask how ethical issues are addressed. While ethical issues apply to the whole process of research and interviewing, they are central to this introductory stage. In case they are neglected or introduced and handled carelessly, consequences may be severe and range from uncertainty about how to use the material to open conflicts and lawsuits between respondents and researchers. The introduction includes: matters of privacy confidentiality voluntary participation and anonymity; practical issues like recording, perhaps video recording; questions of what may be consequences of participation, who will have access to the material afterwards, or how the contents of the interview will be represented in reports, books or papers. At this stage, it is important to give the respondents the opportunity to ask questions concerning all aspects of the interview and the research. Informed consent at the end of this exchange may be ratified verbally, or by signing a letter of agreement. Authorisation through signing a written agreement is advisable, especially when conflicts of loyalty with organisations (employers, etc.) are expected. If the interview is stored in an archive or database with public access, additional measures are necessary. If the ‘contract’ is verbal, permission to record should be captured on tape, and repeated once consent is given. In any case, recording should start as soon as it is agreed. The following example of an introductory explanation is taken from STUDY C (see Box 4.1). The nativespeaking interviewers were asked to formulate the beginning of the conversation following this example.

Page 3 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Box 4.1: Example of an Introductory Explanation (Study C) ‘First let me tell you a little bit about me and the study. My name is XXX. I am a student from Vilnius University. I want to interview you for a study about young people growing up and making the step from education to work here in Lithuania. Many studies are done about this topic but most of them are so-called questionnaire studies, where young people give answers to a long list of questions. We want to do it differently. We decided to meet young people personally and talk to them, especially about their life and what they think about topics like work and unemployment and finding a job. We want to give them the opportunity to talk about themselves and to talk about things that are important to them. I am doing this interview for a colleague who cannot speak Lithuanian, so it is really important that you tell me everything in a way that you would explain it to somebody who has no idea about Lithuania. The interview is confidential and anonymous; that means that I do not even want to know your name. I will just ask you to choose a pseudonym that will be used instead of your name - a pseudonym is some kind of name that you can choose and that will be used instead of your real name. It is important to do this because you have to be sure that you can talk about everything here. There are no right or wrong topics; everything that is interesting for you is interesting for us. Of course, your participation is entirely voluntary and you can always refuse to answer my questions. In order to be able to fully concentrate on what you say I would like to use a recorder during our conversation. It is small [take it out of the bag now!] and I'll just put it somewhere and switch it on as soon as you agree. I'd like to use it so I don't have to write down everything you say. Afterwards I'll transcribe the things you said, which means I'll write down what is recorded. This text will then only be used for scientific purposes. Parts of this interview may be used in the final research report and in publications, but your real name will of course not be included. I will not even know it. Now, is this OK for you? Or do you have some questions that you want to ask before we start? Do you want to know anything else? If not, you can always ask later.’

4.3 Warming Up Page 4 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

How do I get the conversation started? Before the actual interview starts, both interviewer and respondent should have the opportunity to get familiar with each other. This stage should contribute to moving the relationship of trust, established formally in the preceding step, to the more informal level. A good interview requires interviewer and respondent to resonate with each other. Depending on the social and cultural distance of the two, as well as the setting, warming up or breaking the ice could be used to describe this process metaphorically. Essentially, warming up is about small talk and informal chat that should be unrelated to the topics of the interview. The main purpose is to provide the respondents with the possibility to get used to speaking before the actual interview and to give them an opportunity to hear themselves (i.e. literally ‘to give them a voice’). First impressions about the style of speaking, readiness to talk, the clarity of voice and expression, and the articulateness of the respondent can be valuable for interviewers in order to understand, for instance, how much asymmetry in communication the respondents may be able to handle. The interviewer can find out how the situation is actually defined by respondents - e.g. as casual conversation or tense, exam-like situation - and consequently, how the roles of the two participants are likely to be distributed in the interaction. In the logic of the PCI, this step once again helps to establish the roles of the two parties in relation to the reconstruction of the problem, now with regard to their communicative competence and contributions. Warming up is perhaps comparable to the sound-check before concerts; it allows for fine tuning of the communication and the contributions of the participants in terms of input and output. The tuning of communication actually continues well into the initial interview but at least a first step is made. Usually, this initial negotiation immediately follows the introductory explanation with its standardised content. However, depending on the respondent and the situation, breaking the ice and familiarising oneself with the interviewee may also precede the rather formal step of introducing the study. For instance, it can be part of walking together to the site of the interview; or an informal telephone conversation before the interview may have a similar effect. The order of what comes first, the introductory explanation or the warming-up, is not as important as the fact that it is done per se, and before the main part of the interview is initiated by the interviewer with the opening question.

4.4 Opening Question How (and what) should I ask at the beginning of the interview? The beginning of the problem-centred conversation, the core of the interview, is a particularly critical moment for the unfolding and quality of the dialogue that follows. The initial phase of the interview, starting with the interviewer's opening question, has the function of breaking up the stereotypical question–answer structure which is the usual, socially-shared image of what characterises an ‘interview’. Instead, the aim is to facilitate the establishment of a narrative conversational structure evolving largely according to the priorities of the respondent. In other words, and following the PCI's principle of object orientation, content (the individual Page 5 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

perspective on the problem at hand) and form of the account (the way of articulating and assimilating the issue) should depend, for the most part, on the respondent. Conceptually, in the terminology of the PCI, the opening question is a tool of general exploration (see below). It aims at encouraging narratives and in this way producing material in the sense of starting points for further explication. The opening question should, on the one hand, emphasise once more the significance of narrative accounts and extensive storytelling on the part of the respondent. Interviewers should also point out that they will sit back and listen; they will not interfere except for clarifications, and will instead take notes for later follow-up. Importantly, this announces and anticipates the ‘return’ in communication to a less asymmetrical dialogue afterwards. On the other hand, the opening question should be relatively general and non-directive in order to create space to be filled in a narrative way; drawing the respondent's attention to specific aspects of the problem should be avoided. Opening questions of the PCI should combine the quality to provide a ‘blank page to be filled in by the interviewee’, as Merton et al. (1990: 15) write about ‘unstructured questions’, with the establishment of a problem-centred frame of reference. Within the latter the respondent should be free ‘to indicate the foci of attention’ (ibid.). The narrative orientation of PCIs and, correspondingly, the interviewing style, as well as the opening question, may vary according to the issue at stake and the personality of the respondent. As a rule of thumb, an initial narrative question combined with an altogether ‘receptive strategy’ of interviewing, as Wengraf (2001: 154–6) calls it, is preferable when the respondent is expressive and communicative. A more ‘assertive strategy’ (ibid.) throughout the interview - attempting to activate the respondent's stock of knowledge - is more appropriate if he is uncertain, shy and inarticulate. In any case, the richer the opening account triggered by the initial question, the better for the dialogical reconstruction of the problem that follows. Box 4.2 includes examples from STUDIES A and C that illustrate the range of opening questions that can be applied depending on the topic and situation. A third example is included in the interview guide used in STUDY B (see Appendix 2).

Box 4.2: Examples of Opening Questions The first example is taken from STUDY A; it is very short and to the point. It links the opening question to information about professional choice available from a survey done before. The communication following this example is analysed in more detail in section 6.2.

‘You learnt the profession of … (e.g. hairdresser). How did this actually come about? Please, tell me about it.’ The second example is taken from STUDY C; it is much longer and goes more in the direction of an opening question for a narrative interview. It tries to introduce the

Page 6 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

respondent more slowly to the topic of investigation and mainly invites him to talk about a time frame of five years that is likely to cover all biographical experiences that are relevant here. The meaning of work is also briefly addressed but not further specified here; a few other issues could have been added. The advantage of such a longer opening question is that it introduces several issues that can be picked up later on. Even if the respondent's opening account does not cover all of them, the interviewer establishes ‘claims of interest’ by mentioning them in the opening question - i.e. he can come back to them later in the conversation. If these follow-up questions are introduced properly - e.g. ‘In the beginning I also mentioned that … Could you tell me about it?’ - the respondent will understand the ‘reason’ why they are asked. A lengthy introduction like this also has the advantage of giving the respondent time to prepare the answer. Ideally, the cognitive process of preparing a response will be triggered after the first two sentences, which express the main interest of the interview.

‘As I said, I want to talk to you about two things - first about you and your life and then about what work means for you. So, at the beginning I want to ask you to tell me about your life. Try to think, for example, of the last five years of your life before you came here [i.e. the institution where they work, study, or spend their free time]. Tell me about your experiences and how these five years passed. I will say very little and only take some notes, and if I ask you any questions it will mainly be about something not clear to me, if I don't understand something. So try to think back and tell me everything step-by-step that is important to you. There are no wrong or right things. Everything is interesting for me. Just tell me YOUR story of the last five years.’

4.5 Opening Account Why is the opening account so central for the PCI? The narrative that follows the opening question - although it doesn't always follow it! - is methodologically crucial in two ways: it provides the substantive base, i.e. a pool of topics, as well as ‘reasons’ for following up and probing. Differently from the biographical-narrative interview, however, the PCI is usually not interested in the analysis of ‘narratives’ as such (including the analysis of textual types; see section 2.4 and Rosenthal, 2006). Narrative accounts are important means of generating knowledge that corresponds closely to the way experiences were originally made. They are rich sources for follow-up and dialogic probing; the detail produced through storytelling deepens the thematic analysis. The opening account first of all produces an initial view of the respondent on the problem, a view that is largely

‘uncontaminated’ by the interviewer. The overlap of the content of this opening account with the interviewer's Page 7 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

precise interests and expectations is secondary to the emerging contours of the issue at stake from the perspective of the respondent. In epistemological terms, the respondent's (everyday and practical) knowledge constitution has absolute priority here. Often, this initial statement can be the most revealing source for both follow-up questions and later analysis. Even if the opening account may turn out to be meagre in terms of length, it may contain the ‘essence’ of the case, which is only differentiated and further substantiated in the course of the interview. Thus the interviewer needs to be alert to detail and take mental or written notes of key aspects that may otherwise get lost during the interview and resurface only afterwards in the course of analysis. In case the stimulation of an opening account was not successful - and the reasons for this may be quite diverse - it should be tried at a later stage and after some more dialogic and casual conversation around the issue. The second crucial function of the opening account concerns how it delivers the communicative legitimation of the interviewer's curiosity in the follow-up process. The idea is that respondents react cooperatively to questions that seem obvious and ‘natural’. In general, and especially in interviews about sensitive issues or those potentially exposing private information, every question needs a reason, from the perspective of the respondent, in order not to be irritating. The initial narrative therefore provides not only the most valuable information but also it is a recognisably incomplete source of information - i.e. the respondents consider the follow-up questions, which complement the initial narrative, to be legitimate. Thus, the opening question may include additional points of reference for follow-up (see the longer example in Box 4.2). Follow-up questions are most acceptable if the respondent understands that they are based on the interviewer's curiosity in relation to the first account and to the answers to the opening question. When the opening question invites the reflection of a (biographical) process or experience over time (e.g. ‘the last five years’), the initial narrative also allows the interviewer to assess the respondent's way of dealing with an open time frame (Reiter, 2003). The interaction and coordination of opening question, opening account and follow-up is absolutely crucial for successful PCIs, as the following example in Box 4.3 illustrates. In Chapter 6 we deal with these issues more extensively.

Box 4.3: The Interplay of Opening Question, Opening Account and FollowUp - an Example (Study A) In order to illustrate the importance of providing an open space for the respondent to fill in by formulating an appropriate opening question, we want to use the example of a young bank clerk with the pseudonym ‘Emil Bank’. He was interviewed shortly after finishing his apprenticeship education. This is how the interview started.

I: Ah, you can simply start there, how it happened that you decided in favour of becoming a bank clerk. We can continue from there.

Page 8 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

E: Yes, of course. Ah well, it actually came more or less through my brother. He also did bank clerk. And I did commercial school, and it was actually bank clerk or industrial clerk, it depends. And as my brother did bank clerk and rather liked it, and as I also learned a bit about how it works, I actually also came to the bank. At that time I sent five applications, I think, and got three positive answers, well, two definite acceptances and the bank was also interested in me, I got it then … I: You got three positive answers? E: Yes, well, two definite acceptances, and one was interested, but then I already had the other job. And, well, then I simply cancelled it there. I: And was it already like, ‘I will now go to commercial school, this is clear to me. I'll go in the direction of banking or in the direction of something businesslike.’ How did it happen? E: It was actually already, when I went to commercial school, yes, somehow into business, yes. Because otherwise I wouldn't have done commercial school, otherwise I'd have done normal secondary school. But attending commercial school means my apprenticeship is six months shorter. I: Ah, I see, yes. E: Commercial school is two years, secondary school two-and-a-half years. And I mean, also the chances are, I was told, a bit better to get from commercial school into the bank than from secondary school. Let's take a look at the opening question, the response and the follow-up. The respondent's original construction of orientations and actions, both retrospectively and prospectively, is followed by three levels of reconstruction involved in the whole process. First, the interviewer reconstructs the constructions of the respondent by formulating follow-up questions (reconstruction I). Second, the interpreter reconstructs both the constructions of the respondents and the reasons and decisions involved in the formulation of the questions by the interviewer (reconstruction II). The reader of the resulting report or article would be involved in an additional level of reconstruction (reconstruction III).

How does the interviewer formulate the opening question? The transcript does not include the preliminary talk. It starts with the colloquial formulation

Page 9 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

of the opening question that continues the natural everyday communication of the preliminary talk. The introductory question aims at stimulating a narrative account related to one of the key issues in the interview guide. By weaving in the expressions ‘simply start there’ and ‘then continue’ the interviewer communicates his substantive interest (what) and his expectation of the response to have a narrative structure (how). In substantive terms the question targets the level of action. This has two advantages. First, it is easier for the respondent to reconstruct the level of action and it facilitates the reply on a very concrete level. In this way, respondents will likely combine the description of their actions with reference to related circumstances, decisions, orientations, etc. as well as evaluations of the results of these actions. Second, this additional ‘soft’ background information about the situational and subjective embeddedness of action facilitates the interviewer's pre-interpretation in the course of the interview as well as the systematic analysis afterwards.

What is included in the first narrative sequence of the answer? In this case, the initial narrative sequence is rather short, which is typical for this group of young people. In general, the extent and form of the initially offered answer to the opening question depends on several aspects, such as: • the extent of biographisation - i.e. the respondents’ reflection of their (professional) biographies that enables them to produce a coherent and detailed narrative; • the respondents’ role taking - i.e. the readiness and capability to accept their role as respondents, to disengage temporarily from the taken-for-granted contextuality of all explications, and to support the interviewer's efforts of understanding them; • the persistence of the initial unawareness of the extent of detail required by the interviewer - i.e. the readiness to give up the general communicative norm of concise answers to questions. The answer sequence consists of four thematically distinct segments that can be attributed to the topics of ‘occupational options’ and ‘search for apprenticeship training’ from the interview guide (see Appendix 1):

1 2 3

Page 10 of 31

The important role of the brother for the occupational choice mentioned in the first sentence. The importance of the type or level of school as a resource for the occupational choice and the two related professional options. The specification of the role of the brother, who likes this profession. Emil

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

4

thinks that he could also establish a similarly positive relation to this profession. It remains open to what extent the expression ‘I also came to the bank’ refers, either to a rational decision on the basis of Emil's similar appraisal, or to the spontaneous realisation of an opportunity. However, the option of industrial clerk has already been discounted. The realisation of professional interests. His effort consisted, probably, in writing five applications. He first talks about three positive answers, but corrects it to two. It is not clear how he continues his explanation but he seems to qualify one of the positive answers, saying that one bank only indicated interest. His concluding statement ‘I got it then’, finally, does not specify which option he managed to realise.

This opening narrative statement is concluded with a clear break (transcribed as ‘…’), a coda, that is not interrupted by the interviewer.

How does the interviewer follow up? The interviewer could continue by exploring details related to any of the discussed issues. All of them could be starting points for further narrative sequences which would then allow the reconstruction of links and relations within the data. Considering the affinity of the present example to life-course sociology one could, for instance, explore the temporal order and chronology of the activities mentioned. The reconstruction of the sequence indicates that the choice of the fourth topic is an attempt to clarify the confusion regarding the number of applications, a choice that turned out to be unfavourable, as will be shown in the following. It would have been better to postpone the clarification of this detail as it could be addressed later on. Yet the interviewer opts for the immediate clarification by simply repeating the uncorrected version in terms of a question (‘You got three positive answers?’). The brief reply indicates that this was the wrong choice: Emil's answer revises his statement but is equally difficult to understand. Obviously he wanted to say that the third interview at the bank was not important for him anymore because he already had two positive answers and had probably already signed the apprenticeship contract with one of these two banks. Short opening narrations are particularly difficult to handle in terms of interpreting them (reconstruction I) and formulating follow-up questions. That's why it is advisable to follow the thread offered by the respondent in the initial narration. Topics established therein should be extended by means of narration-generating follow-up questions; they help to increase the wealth of the material for later analysis. In methodological terms, the

Page 11 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

process of following up provides the possibility to solve problems related to short initial narrative sequences described above. Respondents can be supported in their efforts to reconstruct a period of their life and in recalling associated details (see biographisation). In addition, they can be encouraged to present their actions and orientations in a way that is understandable for the interviewer (see role-taking). In thematic terms, the stepwise unfolding of the overall issue of the transition to employment allows us to relate detailed descriptions of certain steps to the overall shape of the biography. This ‘slow’ procedure has two advantages for the respondent: the specification enhances his feeling of being taken seriously; and he understands that it is logically and thematically relevant to reconstruct the whole process from his vocational orientation to the realisation of a certain professional option. In this way it is comprehensible for Emil that the interviewer elaborates on issues that remained open, for instance: • • • • • •

Which bank did he choose as his employer, and why? Did he have a signed contract with one of the banks? Did he get the other positive answers too late to be considered? Why did he commit himself that early to one of the banks? Did he actually commit himself in terms of a contract? How does he evaluate the result of his action in retrospect?

Finding answers to these questions is important because it addresses a crucial theoretical question: his responses indicate that Emil did not organise his search for an apprenticeship place according to the model of rational choice. He did not wait until he had received all replies from the banks in order to arrive at a considered ‘choice’. Instead it seems he accepted the first available option. This hypothesis is underlined by the fact that, according to our context knowledge, at the time of the interview, the job market was particularly tight.

How is the interview continued? The interviewer understands the dead-end situation created through the first follow-up question and reconnects to the thread available in the initial narrative of the respondent. He comes back to the issue of attending commercial school; this was the starting point for Emil's professional decisions and he uses it also as a starting point for another narrative account. In a somewhat clumsy way he formulates the next specifying question that should encourage a narration about whether Emil had already started commercial school with the aim of working in business or in the bank sector. Emil recognises that the interviewer

Page 12 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

understands him and explains that he had already planned to take a job in business when he took the decisions concerning secondary school. The choice of commercial school had been ‘instrumental’; and he was able to save six months of apprenticeship.

4.6 Follow-Up and PCI Communication Strategies How do I make sure that all relevant interview topics are addressed in the right way? Once the opening account is concluded, a more or less extensive interactive and dialogical reconstruction of the problem follows. As explained above, the opening account is the main hub for exploring the issue. If possible, follow-up, probing and complementary questions search to establish links with aspects addressed in the initial narrative or statement. This part of the interview in particular, and the way it is done following the principle of problem centring, distinguish the PCI as a method of dialogic and interpretive data collection from other techniques of interviewing. The dialogue with the respondent is the best opportunity for the researcher to understand the respondent's perspective. The aim is the generation of both a rich and comprehensive body

of knowledge about the issue by means of clarifying pre-interpretations that the interviewer develops during the conversation. In qualitative social research, interviews constitute joint efforts of interpretation that involve both interviewer and respondent. It is not a symmetrical situation and the exploration of the respondent's perspective has priority. Yet, in ethnomethodological terms (see Garfinkel, 1967/2011), the respondent's statements are

indexical – i.e. they refer to a certain context of production (see section 2.2). Therefore, in order to understand their meaning, the interviewer needs to consider the context and learn more about it. For instance, the statement ‘It is raining’ has different meanings for people living in the Sahara to those living in a rainforest. People living in either environment are likely to understand the meaning on the basis of shared contextual knowledge that usually remains implicit. The fact that they are part of a certain cultural context enables them to identify the meaning and the pattern that underly each other's statements, actions and events. Everyday realities operate on the basis of such hermeneutic processes that enable common and coordinated action. In the example, inhabitants of the Sahara understand (and agree!) that something special is happening when it rains, while inhabitants of a rainforest may not even pay attention to this statement. The mechanism that people use in everyday life in order to understand each other is the same that the researcher can apply to interpret and understand indexical statements. Garfinkel (1966) calls it the

‘documentary method of interpretation’ (see Box 2.3). Expressions like ‘It is raining’ are ‘documents’ representing a certain underlying pattern and context of understanding. Researchers are usually, at least to some extent, alien to the cultural context and the worlds of knowledge that the respondents inhabit. In order to understand what the respondents tell them, they need to find ways of gaining access to this meaning

context in which statements are embedded. In the example, they could, for instance, ask the respondents (i.e. inhabitants of the Sahara or a rainforest) for the recollection of their first experience of rain. The length Page 13 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

and detail of the answer to this question, alone, probably tells the researcher a lot about the meaning of the event of rain. In other words, he starts developing pre-interpretations of the meaning of rain for this person. In the course of the interview he then has the possibility to ask for specifications associated with the experience of rain. And in the course of the conversation he may even change his mind about it – i.e. revise his pre-interpretation. Researchers can investigate structures of meanings by hermeneutic understanding and combining deductive and inductive elements: the deductive attribution, by the interviewer, of single aspects of the interviews to available patterns of interpreting meaning (i.e. prior knowledge) is complemented by the inductive search for new patterns in the accounts of the respondent. The design of the PCI includes a variety of special communication strategies (see Table 4.1) that can be applied to extend the context knowledge of the researcher and allow him to reconstruct and understand what the respondent wants to say. These communication strategies should: • translate the basic idea of this documentary method of interpretation into questions that help to explain the contextual meaning of statements; • gradually get the respondents to understand that their statements are charged with implicit knowledge so that they start to explain it on their own accord by means of specifications, narratives, descriptions, etc.; and • consider the process character of the documentary method of interpretation: meanings are revised over time through reinterpretations.

Table 4.1 PCI communication strategies General exploration Beginning

of

conversation,

generating opening question Detailing questions Examples from experience

Generating material narration-Open invitation to explicate the respondent's view, focusing on problem Specification of themes, problems, background; stimulation of memory Stimulation

of

memory,

reconstruction

of

Completion of information, gaining comparability

Repeated thematic comparison

Conceptual clarification, thematic differentiation

Specific exploration

Generating comprehension

Comprehension questions Confrontation

and

establishing links to structural environment

Ad-hoc questions

Mirroring

context

Cognitive

structuring

for

interviewer

and

respondent,

communicative validation Clarifying common-sense structures, lacking or unclear terms and facts Clarifying or inviting further specification of contradictory statements

PCI communication strategies represent different ways of involving the respondent in a process of clarifying meanings and revising pre-interpretations. They either serve the purpose of general exploration regarding Page 14 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

the issue; then they generate material. Or, they deepen the knowledge about the problem in an effort of

specific exploration; then they support the generation of comprehension. These communication strategies were already part of the original design of the PCI at the beginning of the 1980s; in the meantime, some of them have become standard tools in qualitative interviewing. In the following, we introduce them in an overview and illustrate them with examples (see Tables 4.1 and 4.2). More extensive examples can then be found in Chapter 6.

Table 4.2 Examples of PCI communication strategies General exploration Generating material Narrationgenerating opening question

‘You would like to become a hairdresser. How did this come about? Please tell me how it all happened.’

Detailing questions ‘Could you tell me in detail what happened there?’ Examples

from

experience Ad-hoc questions Repeated thematic comparison

‘Could you give me an example for … (e.g. a daily routine or biographical episode)?’ Ad-hoc formulations regarding topics from the interview guide at suitable moments Contrasting typical/untypical cases or past/present comparison (see section 6.2)

Specific exploration Generating comprehension Mirroring Comprehension questions Confrontation

Summarising and rephrasing, provoking contradiction, inviting comments – ‘As far as I understood …’ Ideas from the interview – ‘You just talked about … I did not understand this.’ Summarising of contradictory statements and careful demanding of clarifications – ‘Before you said that … But right now you say … Did I not understand you correctly?’

The effort of generating material (i.e. narrations and storytelling) usually starts with the opening question (see section 4.4). In order to broaden the narrative base, storytelling is invited through various forms of general probing (e.g. specifying questions, examples, thematic comparison) and ad-hoc questions – they explore knowledge and meanings indirectly. In the terminology of the PCI, these elements are part of the general

exploration of the issue aiming at gradually disclosing the respondent's view of the problem in an open and inductive way. Thereby, the interviewer picks up topics of the initial narration and invites the respondent to further specify his view, again preferably in the form of additional extensive narrative accounts. Asking for examples of experiences or biographical episodes stimulates the respondent's memory, helps clarify lacking or unclear issues, and gradually differentiates and thickens the respondent's perspective. Usually, the opening account was exhausted as a source of follow-up before all the relevant issues of the topical interview guide are covered. In this case, complementary ad-hoc questions address these issues in order to guarantee a certain comparability of the interviews.

Page 15 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Differently from the narrative interview (Jovchelovitch and Bauer, 2000), the PCI does not suggest strictly separating ‘immanent’ and ‘exmanent’ follow-up questions; they are not exclusively part of certain interview phases. This would create a very artificial interaction; instead, the problem-centred interaction approximates everyday communication. The above-introduced idea of communicative legitimation of follow-up questions out of the initial narrative (see section 4.5) can be approximated here by simply introducing additional questions that refer to something that has not already been discussed in the course of the conversation. Also, neglected aspects introduced in the frame of the opening question can be addressed. In this way the PCI consists of an ‘interplay of material and comprehension generating questions’ (Mey, 2000: 143). Unlike the narrative interview, the PCI integrates dialogic and narrative forms of communication throughout the whole interview communication. They are not separate or reserved for certain interview phases. The alteration of narratives and dialogues depends on the interview interaction (see Box 4.4). In particular, our extensive discussion of examples in Chapter 6 illustrates the essentially interactive and dialogic (re-)construction of knowledge in PCIs.

Box 4.4: Dialogue and Narration: Different Interviewing Styles or Two Related Forms of Communication? The narrative and the dialogue are integrated elements of problem-centred communication. Interpretations that take narrative and dialogue for two separate forms of communication misunderstand their role in the PCI. They assume that these forms are applied at different stages, and they tend to interpret the dialogic form to follow the conventional question-answer pattern. In the PCI, there are no designated transitions where there is a change ‘from a narrative to a semi-structured part’, as Scheibelhofer (2008: 408) writes. At which point during the interview and to which extent dialogues or narratives predominate depend first of all on how the interview interaction unfolds. Interviews are always special situations ‘that the interviewer and interviewee co-construct and from which the interview content emerges as a contingent product’ (Lee and Roth, 2004: paragraph 37). The interviewer's part consists in stimulating the production of interview material by inviting narration and storytelling at the beginning of the interview and later with regard to certain thematic aspects and sub-topics. This is necessary in order to be able to formulate meaningful follow-up questions at a later stage where he can ask for further details that can help to deepen or revise pre-interpretations. For instance, the interviewer can ask the respondent to describe concrete experiences and events or situations and episodes. Depending on the extent and comprehensibility of the respondent's account, the interviewer chooses questions that generate either comprehension or narration.

Page 16 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The respondent's contribution to the construction of the interaction depends on whether the interviewer's interventions stimulate short answers or narrations. Some respondents do not need much encouragement to provide extensive narrations, which, at first, do not require follow-up. Yet it is more appropriate to expect respondents to provide short answers, for instance because … … short answers better correspond to social conventions; … respondents might still be uncertain about whether the interviewer really accepts their points of view in case they deviate from social norms; … it is difficult for respondents to reconstruct problems, owing to fragmented memory or their lacking language competencies; …. at the beginning of an interview respondents do not understand the desired level of detail or the chances and benefits of biographical reflection; thus they may at first be happy with superficial banalities. Interviewers need skills and experience to use the two communication forms of dialogue and narration in an appropriate way. As soon as they manage to anticipate possible reasons for short answers they can try to balance it on the level of the interviewing relationship. The consequences for the actual length of answers, however, always remain uncertain. The length and narrative quality (or not) of answers depend on the interaction process between the interview partners. They are involved in a constant process of communicative negotiation. In the logic of the PCI it is the quality of this process, and not the design of various narrative or dialogic ‘phases’, that decides on the extent to which issues are deepened and explicated. Techniques of generating and enhancing comprehension of what has been said use specific probing questions (e.g. mirroring, comprehension questions) as well as confrontations in order to clarify contradictions and inconsistencies in the account. These techniques are part of the specific exploration of the issue and introduce deductive elements into the conversation. In practice, similar to communication techniques used in non-directive conversational therapy, the interviewer invites the respondent to elaborate and reflect upon some of his statements. This provides the opportunity for the respondents to defend their views against possible misunderstandings, or to modify them as the dialogue evolves. In any case, the revision of their opinions is welcome and only contributes to the consolidation and validation of knowledge.

Paraphrasing something that has been said, or mirroring it back to the respondent in one's own words, is the most moderate form of doing this. In addition, direct comprehension questions can be used in case certain earlier responses need clarification. Differently from the narrative interview (Schütze, 1983; Jovchelovitch and Bauer, 2000), the PCI suggests using comprehension questions within their original context of relevance in the account, instead of postponing it. In this way, respondents perceive the interviewer's active interest in Page 17 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

what they have to say; they understand that they are taken seriously and are motivated to provide detailed answers. Finally, confrontations may be used carefully in order to clear misunderstandings and to invite further detailing. This particular form of specific exploration needs to be applied with utmost care for it may require a solid relationship of trust between respondent and interviewer (see Box 4.5). Also the introduction of humour as a means of communicating doubt and misunderstanding may be helpful.

Box 4.5: The Importance of Confrontations Confrontations constitute the strongest form of asking respondents to reflect certain issues. Typically, they either challenge contradictory or controversial statements, or they address aspects of reality that are likely to be part of the respondents’ everyday life. This form of specific exploration is the most difficult to handle. Some interviewing techniques such as the discursive interview (Ullrich, 1999) use it systematically to provoke interpretive patterns. The purpose and use of confrontations has been misunderstood in the reception of the PCI. For instance, Scheibelhofer (2008: 409) criticises their use and maintains that they ‘endanger the interview setting as a whole’ because the respondent may ‘feel uneasy’ when confronted with inconsistencies or ‘compelled to defend himself or herself’. Confrontations in PCIs are greatly misunderstood when they are associated with revealing lies or exposing the respondent's possible self-deception. This must not be the purpose of a problem-centred research conversation. The problem-centred interviewer is not a control freak who judges the respondent's capability and readiness to be consistent; and there is certainly no need to formulate a confrontation in an unfriendly way. As well-informed traveller and attentive conversation partner, the interviewer is eager to understand and reconstruct the meaning of the problem in the perspective of the respondent. It is particularly important that the reasons for confrontations are plausible from within the interview conversation and the texture of the topic (i.e. communicative legitimation). Interviewers have to emphasise their substantive interest in the issue and maintain a good atmosphere. There are various reasons that can make this effort difficult. Statements can be redundant or inconclusive, time lines can be confused, or specificities of language or dialect can produce inconsistencies. For social scientists, individual action is always multidimensional - i.e. apart from a psychological dimension it has the additional practical and social aspect referring to the additional capability to act under conditions of ambiguity (Smelser, 1998). Qualitative interviews provide exactly the chance to uncover inconsistencies, their reasons

Page 18 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

and meaning structures. In PCIs, forms of specific exploration like confrontations are at the core of the interviewer's pre-interpretations of the account. They invite respondents to participate in a reflexive meta-communication about what has been said. Once interviewers are familiar with the respondents’ perspective they are able to address specific aspects as embedded within the overall context of the conversation. In a relationship of trust, where the respondent can talk freely and without qualification, confrontations are not perceived as breaching an implicit contract. On the contrary, confrontations articulate and underline the sincere interest of the interviewer. Rather, not clarifying a misunderstanding would be inappropriate because, within this implicit contract, the respondent has the right to be understood. The following example is taken from STUDY C; it is part of an interview with a 17-yearold school dropout of Russian origin. The interviewer follows up on a remark indicating that the respondent thinks that the situation is worse now than it was during communism; even without first-hand experience he has a strong opinion about it. In the example, the confrontation (‘So you see, they would take you to the reformatory …’) improves the understanding of the respondent's situation and exposes the uncertainty (‘I would also be afraid not to go to school’) involved in negotiating the validity of former assessment criteria within the society's normative framework after the collapse of communism. Furthermore, the passage includes examples of mirroring, and it underlines the importance of familiarity in order to understand what the respondent wants to say (see section 3.4). The availability of (peer) interviewers and translators with the same cultural background greatly facilitated comprehension. I: You said that now everything is worse. Has anybody told you, for instance parents, about the time of the Soviet Union? (V: Uhm.) Do you think it was better then than now?

R: Well, from all the stories so I think that it was better then. I: And what from these stories do you think was better? R: Better? (3) Well, what do I know, earlier everything was cheap and somehow, and somehow I heard that earlier, for example, to work, everybody had to work. Somehow if you don't work, they can put you to prison or something more. Simply, the state itself gives work to people, it makes them work itself. I: But if one doesn't want to? To force one to work … R: I don't know, well … well, simply … what do I know? After all I didn't live at that

Page 19 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

time, I don't know. I: No, I'm asking for your opinion. R: If one doesn't want to? Well, what do I know? I: So people are put in jail? R: Well, I don't know, maybe. Maybe they are put in prison, maybe they get a few days in a cell, or something else … There is a Russian pogovorka (saying; interviewer): ‘You don't want - we will teach (you)’ (translated by the interviewer; authors). Or even: ‘You don't want - we will force (you)’ (translated by the interviewer; authors). Or something like that, I don't know, I don't remember now. Well, you don't want, you don't want it, you will sit in prison for a week and you yourself will scream to be released, that they give you this work. I: So you think that even now it would be better that they give (it to) everybody, everyone would be told to work, whether they want to or not? R: Whether they want to or not, anyway people were ready then (in the Soviet Union; authors), and there was work. Everybody worked. Young and old. There were no such (people) like me, probably. Such (people) like me, if they don't study, they go straight to some kind of reformatory (jail for minors; translator) or somewhere else …

I: So you see, they would take you to the reformatory … R: Such (people) were even afraid not to study. If these times were now, I would also be afraid not to go to school … (laughing) (emphasis added; authors). In epistemological terms, and following the principle of problem centring, the aim of the process of clarification by means of specific exploration is that of a discursive-dialogic reconstruction and validation of knowledge about the problem. The shift from general to specific exploration and from one style of communication to another follows exactly the documentary method of interpretation introduced above. In practice, the interviewer facilitates narratives through general exploration in order for patterns to emerge out of the individual statements. The shift towards different techniques of specific exploration, on the other hand, can result in new patterns of understanding meaning, or in the revision and correction of available patterns through additional specification. In other words, in the problem-centred communication process, the researcher's prior knowledge and the respondent's everyday knowledge enter into a corrective relationship. Importantly, the pre-interpretations that the interviewer may have developed during the interview get the Page 20 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

chance to be revised or corrected. Due to the complexity of this process and the skills necessary to realise it, the researcher should also do the interviews (i.e. as a researcher/interviewer). The usually small number of respondents on the basis of purposive or theoretical sampling makes this possible. Alternatively, interviewers need to be trained intensively to fully understand the dynamic of this principle of exchanging and revising knowledge. Another advantage of this unity of researcher and interviewer and their joint effort of interactive knowledge production, which is only mentioned here and dealt with in more detail in section 2.3, concerns the management of sensitive moments in the course of research. They may require particularly high communication skills similar to those for guiding biographical-narrative conversations that can even have the character of interventions (Rosenthal, 2003). Now that we described the various communication strategies of the PCI, we want to come back to one of the most common and problematic interviewing mistakes (Pitfall 4.1). It results from an unfortunate and contradictory combination of some of these strategies at the beginning of an interview that leads to the establishment of an unproductive question-answer scheme with little chance to generate either sufficient material or in-depth understanding. A problem-centred interaction cannot develop.

Pitfall 4.1: Domination of the Question-Answer Scheme How the question-answer scheme emerges in an interview … The question-answer scheme starts to dominate a conversation when interviewers fall back into a socially accepted form of communication with a clear-cut distribution of roles: as the subject and driving force of the conversation, the interviewer structures the conversation by asking clearly defined questions; as the object of the conversation, the respondent interprets these questions, partly out of convenience, without further efforts to enter the sometimes difficult reconstruction and evaluation of experiences. In cases where the roles are established as described above, interviewers will fail to act according to the basic (qualitative) attitude which values the respondent as the subject of the description of actions and orientation. Instead of encouraging, supporting and actively listening, they confront respondents with a series of questions and deprive them of the chance to actively shape the flow of the conversation. Short answers to precise questions are characteristic of the question-answer scheme. Because of the lack of substance it is difficult to formulate specific follow-up questions. Respondents, on the other hand, feel that they are not taken seriously, and a relationship of confidence is unlikely to develop. The respondents’ readiness to invest in the reconstruction of their perspective is lost and the inadequate interviewing style further reinforced: the interviewer cannot help the short answers and is more and more drawn into a circle of formulating equally isolated follow-up

Page 21 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

questions. Lack of experience is the main reason for this regression into a traditional form of interviewing that ignores the basic principles of the PCI. Often, interviewers are overwhelmed by the requirements of active attention necessary in this intensive confrontation with another person. They cannot ‘sit back’ and relax because they feel challenged as the driving force of the interview; and they do not manage to overcome the respondents’ initial expectations towards a ‘typical’ interviewer, i.e. as somebody who asks one question after another.

… and how it undermines the chance to reconstruct meaning Once the interviewer is established as the person in charge of the interaction, he has to respond to each short answer with a new question. In turn, this forces the respondent to provide answers to isolated questions and makes it difficult even for him, supposedly the authority of knowledge about himself, to comprehend the connection between them. These accounts necessarily fall short of providing insight into the interconnectedness of different aspects of a topic that is usually established on the basis of narratives and by means of typical rhetorical conjunctions like ‘and’ (summative), ‘and then’ (temporal), ‘because of (causal), etc. In the end, the interviewer fails to collect the detailed basis for deepening his understanding and for controlling preliminary interpretations; he cannot arrive at empirically-grounded answers to his research questions. Also for the respondents the narration is an important source and base of constructing meaning. It provides them with the chance to reflect on issues in a practical perspective and to articulate their views in original ways. They would consequently follow their subjective logic of how events and actions develop, and also formulate expectations and appraisals within this structure. When the question-answer scheme predominates a conversation, this genuine structure of unfolding experiences is scrambled into pieces, and linked elements are missing. This fragmentation of knowledge constitutes a serious interpretive challenge as it is difficult to assemble these knowledge pieces produced through ever-new sub-questions. In other words, it is difficult to examine them for how they conceptually correspond to the meaning of the terms used in the main question. Ideally, according to the process-related character of interviews, respondents would establish and gradually differentiate their universe of meaning in their own words. Statements and opinions are corrected and revised by respondents as the narration evolves, from rough outlines in the beginning to rich details at the end. They can only remember, specify, and, if need be, correct details in the course of the interview. Redundancies and contradictions are usually part of such extensive accounts; and they are beneficial for the process of interpretation.

Page 22 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

An example from STUDY A This example of an interview with Kuno, a young sales clerk, illustrates some of the problems related to the domination of the question-answer pattern. The first sequence is taken from the very beginning of the interview, the second one from a somewhat later stage.

First sequence 1

2 3 4

I: Well, you do not need to give specific answers to questions, but I would love it if you could just talk about what comes to your mind, what kind of experiences you have, and how you explain for yourself what became of them, and such. Let's start with how you got into your profession. When did you start thinking about what you would like to become? A: When I was still small, actually. As my father does that, I actually also wanted to, something commercial. And then I went to … I: And what kind of commercial does your father do? A: He is a commercial employee at X (company), services …

Second sequence 30 31 32 33

34 35 36 37

I: How was that, when did you start searching for an apprenticeship? A: After the eighth grade. I: Did you then … well, how did you do that? A: First, well … my brother works there in the company, and I did an internship there, that's why I thought, I'll also go there. And in fact, there you get simply … that's a delicatessen shop, there are several departments, that's where you get, and then also into the office, purchase and so, one gets to know it all. I: You said internship. What internship was that? A: That was an internship of 14 days, and then one could (go) to this company, and one simply assisted there, and so. I: Still at school, yes? A: Yes.

A closer look at the first sequence (1–4) shows that already the very first lines of the interview establish a question-answer pattern with consequences for the rest of the conversation. The interviewer starts by outlining his alternative approach to interviewing and then continues by formulating a narration-generating question. He complements this

Page 23 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

by a detailing question concerning the respondent's first thoughts about the issue of professional choice (1). As is typically the case, Kuno at first responds to the second and perhaps also simpler detailing question by specifying when he was first preoccupied with issues of professional choice. But then he obviously remembers the first question and initiates a narrative by pointing to an option that is oriented towards his father's profession (2).

What went wrong … The interviewer's well-intentioned methodical explanation at the beginning and then the first question - an invitation to recount the respondent's professional background - is contrasted by a second, detailing question. Like in a conventional guided interview, Kuno replies to this last specific question by providing a specific answer. In this way, the first, more thematically-oriented stimulus is also subsumed under the question-answer scheme. Thus, the response to this general invitation to storytelling remains brief and is restricted to a short explanation of his orientation towards the professional choice of his father. The interviewer probably was not aware of the contradiction between the two different questions - a combination that should in any case be avoided! He probably thought of the second question as a means of assisting the open narration; it should have merely only helped the respondent to find a possible starting point for unfolding the overall background of entering the world of work. Kuno, however, is not able to recognise this connection and therefore answers the two questions separately, and follows the standard question-answer pattern. In fact, at one point Kuno seems to remember the open invitation and the interviewer's emphasis on the alternative way of communicating: he makes one attempt to describe how he realised this position (‘And then I went to …’) but does not complete it. The abrupt discontinuation of this thought suggests that Kuno is uncertain about his possibility to shape the interview. At that point in the conversation, he is not able to resolve the contradiction between the problem-centred proposal of an open narration and the detailed question that immediately follows. The interviewer does not seem to recognise Kuno's attempt to start a narrative account and does not make us of it. Instead, he turns his interest from the interview guide to explore the precise profession of his father, to a detailing question (3). This comes unexpectedly for Kuno and results in another short answer, which again does not open up any possibilities for a narrative reconstruction of the respondent's professional history (4).

Page 24 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

… and what are the consequences? The consequences of this interaction - and they are evident from the second interview sequence (30–37) discussed below - are unfortunate. Contrary to his announcement at the beginning, the interviewer's actual questions and behaviour indicate that he wants to remain the subject of the interview. He essentially establishes a question-answer scheme that does not invite extensive responses but can be satisfied with short answers. Only a respondent that is highly competent in biographical reflection could have compensated for the interviewer's mistakes here. After an inadequate attempt to explore the reasons behind the professional choice, the interviewer, who keeps navigating the conversation schematically, now has to find his own way of addressing the issue of realising the professional choice (another point of the interview guide). In fact, the first part of the related question seems to go in the direction of a narration-generating question regarding the search for an apprenticeship: ‘How was that, …’. However, the interviewer again immediately corrects this open approach in favour of the inquiry of specific information about the moment of Kuno's efforts (‘… when did you start …’) (30). With this abrupt change in the questioning style, the interviewer repeats his earlier mistake of formulating an inappropriate question. Once again, he does not stimulate a description of the respondent's history as it is envisaged in the narrative invitation at the beginning of the conversation. A more extensive description would have included information about the respondent's temporal structuring of aspirations and of efforts of realising his ideas and their results. Instead, the interviewer now has to struggle for the temporal and biographical localisation of each new issue he introduces. By failing to stimulate a narrative biographical account, the interviewer lost the chance of getting a genuine temporal order of experiences and events that would unfold in such an open narration.

The chance to get it right … The respondent reacts appropriately and provides the requested information in a concise way (31). Then he falls silent and, corresponding to the conventional question-response pattern, awaits further questions from the interviewer. This time, the interviewer manages to keep the communication going by asking a narration-generating question, which opens up the event space of searching for an apprenticeship place in a more adequate way (‘… how did you do that?’) (32). This question cannot be answered with a simple ‘yes’ or ‘no’; it is formulated in an open way with the purpose of stimulating a narration at the level of action. This approach is now rewarded with a narrative and a more substantive answer: with the support of his brother, Kuno could participate in an internship and managed to

Page 25 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

gain professional experience in different departments (33). In thematic terms, at this point the respondent's description provides two chances for detailed follow-up questions. On the one hand, Kuno's experience in the different departments during his internship could have been consolidated with another narrationgenerating question like, for instance: ‘Could you please tellme more about being in all these different departments?’ In his response, Kuno would then have outlined possibilities for the further exploration of his experiences in the mentioned departments (i.e. ‘office’ and ‘purchase’). Descriptions of experiences usually involve evaluations; thus they can contribute to a more coherent image of the background of his professional decisions. On the other hand, ideally, the interviewer should have noticed that the brother had already taken up the same profession as the father, who is also a role model for Kuno's professional interests. This peculiar relation between professional choice and kinship could have been addressed, for instance, by employing the communication strategy of mirroring. However, within the flow of the interview, the first topic is more suitable for detailed followup, especially at the beginning of an interview when the development of pre-interpretations is facilitated by the availability of rich and extensive descriptions (here: of the internship). The second issue of professional choice could enter the interviewer's notes and is better addressed at a later stage. In methodical terms, the interviewer should have noticed that his strategy of opening up the topic of apprenticeship search at one point (‘… how did you do that?’) indeed produced a short narrative sequence. It is richer than the answers before and results in a set of issues that suggest many detailing follow-up questions. Considering that the questionanswer scheme dominated the interview, this brief narrative sequence also indicates that the respondent managed to liberate himself somewhat from this pattern. For the interviewer, this would have been the chance to remember the principles of the PCI and to get out of this circle of being forced to provide a series of questions closely related to the issues of the interview guide.

… and how it fails Instead, as the next lines of the transcript indicate (34–7), the interviewer does not take the respondent seriously; he ignores the topics he offers, and falls back into the questionanswer scheme. With his very selective question concerning details of the internship, which may be relevant regarding the interview guide, he undermines Kuno's potential to involve himself more strongly in shaping the conversation.

Page 26 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

4.7 Social/Personal Characteristics How do I ensure that I do not miss important comparative background information about respondents? The main part of the interview may not provide enough opportunities to collect important background information about the respondent which is critical for a broader assessment of the interview statements. Some details about social background and personal characteristics may be provided implicitly, and some may be addressed occasionally throughout the interview, for instance, by means of ad-hoc questions. Yet in order to concentrate this kind of data collection and to facilitate its completeness and comparability there should be a moment towards the end of the interview that is dedicated to it. As the collection of this information can itself trigger narratives or invite follow-up questions it is recommended that the recording of the conversation continues. Selected passages can then be considered for additional transcription. While the original research interest determines many of the specific details under consideration here, there is basic information about the respondent that should not be missed. Typically, we should at least register the age of the respondent (e.g. the year and month of birth) in addition to gender and family status. Social scientists with some interest in issues of social stratification or social mobility will further investigate education (optionally also for other family members), professional history, and their housing situation. Migration history, ethnicity, denomination and details concerning health are among important additional social indicators. Details and possible misunderstandings can be clarified here. For an example from STUDY C see Box 4.6. In case the PCIs are part of mixed-methods projects, specific information may be required in order to match the respondent's profile with the results of a preceding survey or a classification resulting from it. Other standards may apply to PCIs carried out as expert interviews with a different research focus. In such cases, this step may be omitted altogether. When combined with the brief questionnaire that sometimes follows this step (see below), the systematic collection of background data is also an important indicator for the respondent that the interview is entering its final stage. The style of communication changes at this point and takes the shape of a question-answer pattern. Afterwards, it is gradually dissolved and replaced by more casual conversation in the next step.

Box 4.6: Example of Social Background Data for Comparison (Study C) Social background data (Keep the recorder running!) Pseudonym:

Page 27 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Age - year of birth:

Education Brief reconstruction of the school career: How many years concluded (when dropped out, when repeated, etc.):

Activities with regard to employment and work Work experience (jobs, occasional jobs): Job search experience (labour exchange, friends, acquaintances, parents, etc.):

Ethnicity/citizenship Family - parents, brothers and sisters, grandparents Father - education, job: Mother - education, job: Brothers and sisters - education, job:

Housing/living circumstances Type of property (house/flat)? Where? Number of rooms? Own room?

Preferred choice of career If you could choose, what would you like to become in terms of profession?

Page 28 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

4.8 Short Questionnaire Is there an easier and faster way to collect some comparative information? At the end of the core interview, a short (!) questionnaire can support the collection of information on social characteristics (e.g. age, parents’ occupation, education, etc.). It can also be used to confront respondents with standardised items from survey questionnaires. This may be particularly useful when the PCIs are part of a mixed-methods project and profiles of the respondents of the qualitative interview are matched with those from a preceding survey. The questionnaire can also be a first step in identifying cases for in-depth analysis (see Box 5.3). Alternatively, and in case the respondents of the PCI are recruited out of the quantitative sample, some of the items from the survey may be replicated at the end of the PCI in order to explore their reliability and validity. In the overall design of the interview, the questionnaire can serve as a marker that concludes the main part of the interview. In some cases, especially when the interviewer needs to have some specific information about the respondent in advance, a short questionnaire can be used before the main interview starts with the opening question. Getting to know certain details at the beginning of the interview may facilitate starting the conversation and can help to focus it in case time limits are very strict. Advance knowledge also reduces the necessity to interrupt the interview for ad-hoc questions clarifying necessary details later on. In case the questionnaire is used at the beginning, the interviewer needs to make sure that it does not corrupt or undermine the PCI's purpose of dissolving the inquiry-response cycle that usually characterises interviews. The following set of items (Box 4.7) was used in STUDY C exploring the meaning of unemployment in postSoviet Lithuania among young people in transition to working life. After the main interview was finished, respondents were asked to fill in a short, one-page questionnaire. Had it been used at the beginning of the interview it probably would have influenced their accounts in many ways by introducing and establishing external relevancies. On the basis of the answers, anchor cases for in-depth analysis were selected from the sample of 30 interviews by means of cluster analysis (see also Box 5.3).

Box 4.7: Example of a Short Questionnaire (Study C) Please indicate to what extent you agree or disagree with the following statements on a scale from 1 (strongly disagree) to 5 (strongly agree): • • • •

It is very important for me to have a job. If I won lots of money I would still want to work. Work is one of the most important things in life. If one works hard enough, one is likely to make a good life for oneself.

Page 29 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

• • • • • •

It is the duty of a good citizen to have a job and to work. Having a job means having achieved something in life. Life is very difficult; I do not know what to do. Life is becoming more and more difficult. It is very difficult for young people today because nobody knows how life will be. I agree with some people who believe that life was easier when Lithuania was still part of the Soviet Union.

4.9 Exit and Debriefing How should I conclude the conversation and the encounter? The transition of the communication from an (asymmetrical) interview ‘back’ to everyday conversation between interviewer and respondent at the end of the encounter has several purposes. In general, together with the introductory phase, this exit phase has the function of bracketing and framing a very un usual, at times emotionally quite intensive, social interaction, and at the same time bringing it to a pleasant conclusion. Apart from appreciating the respondent's participation and efforts, the interviewer invites further contributions and final thoughts that may well open up new aspects of the discussion. This moment can also be used for asking the respondents whether they would be available for follow-up research; it includes offering them the chance to contact the interviewer themselves. This should not be offered in cases where issues such as anonymity and confidentiality are particularly important. Towards the end of the interview, respondents should also be given another opportunity to ask questions they may have concerning the research. In the course of the interview the respondents are likely to change their perception of the exercise and become more interested in the background and purpose of the study. Often, they actually only then have questions and perhaps also express their curiosity towards the person with whom they have spent the last one or two hours. The kind of interaction and conversation created during the PCI, and the interpersonal relationship it involves, may constitute a rather unique and intense experience for the respondents. The dissolution of this potential bond of trust and (one-sided) intimacy should be tackled with care. Accordingly, respondents can also be asked to assess their general feelings about the interview now that it is over. The intensive reflection of subjectively relevant issues can leave a deep impression on the respondents, both intellectually and emotionally. At the end of the interview they may have the need to continue their reflection and appraisal of what they said. Or they can be deeply affected by it, even confused. From our experience, the interview can have an empowering effect on respondents: for them, the deep yet professional interest of the interviewer and the sensitive communication style are sometimes literally ‘taken Page 30 of 31

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

personally’ and interpreted as a form of particular appreciation. This may trigger the necessity to talk, after the interview, about other more or less related topics with somebody who is ‘interested’ and competent. The transition from interview to a sort of counselling conversation is common. It is the responsibility of the interviewer to deal with these issues appropriately and with integrity, and to reciprocate the respondent's readiness to share knowledge. After all, it was their joint interest in a certain problem that brought them together. The fact that the interview context may become particularly relevant here should already be considered when selecting the location and institutional or private setting of the meeting. Also, time should be calculated generously in order to accommodate such needs. http://dx.doi.org/10.4135/9781446288030.n4

Page 31 of 31

The Problem-Centred Interview: Principles and Practice

Processing PCIs In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 95-122 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Processing PCIs 5.1 Postscript What do I need to note immediately after the interview? While the previous step mainly serves the debriefing of the respondent, the production of a postscript is an important part of the debriefing of the interviewer, or of ‘self-debriefing’ as Wengraf (2001: 142) calls it. The PCI's instrument of the postscript supports exactly this purpose. The idea of the postscript goes back to earlier qualitative-methodological suggestions to consider the interviewer as part of the research situation as well as the genuinely communicative character of any social science data collection (e.g. Cicourel, 1974a; Laslett and Rapoport, 1975). With their assumptions and doubts, their assessment of the situation and their non-verbal articulations, interviewers contribute aspects to the outcome of the interview that are not sufficiently registered in the recording and transcription. The same is largely true for otherwise non-registered information about the respondents. In order to capture these aspects of the encounter, a postscript – short for post-communication description – is produced immediately after the interview. Postscripts should be written – though sometimes it may be easier to record them – without delay after the interview and without following conventions of order or self-censorship. The importance of this point cannot be emphasised enough. The postscript is a collection of impressions and details concerning the events: (1) before the actual conversation (e.g. establishing contact, the expectations expressed by the respondents, important moments before the interview, etc.); (2) during the conversation (e.g. non-verbal aspects, observations, emotions, thoughts, etc.); and (3) after the conversation (e.g. concluding discussion, questions, follow-up plans, etc.). In his study of Argentine fertility Cicourel (1974a: 84) asked the interviewers to write down ‘post-interview descriptions of the general atmosphere of the session; problems they had experienced with respect to general or particular questions; their feelings about whether the subject had been lying, evading the issue or had not understood the question; and problems in their relationship with the respondent’. Postscripts can include the contents of the whole interview in note form; they can point to particularly relevant passages and reflect the overall quality of the interview. In case of very limited resources, like in the context of BA or MA theses, postscripts can be the basis for choosing single cases for transcription and analysis as well as, altogether, the limitation of the sample. Postscripts complement the tape recording with information about the immediate context, the place and atmosphere of the interview as well as non-verbal and (positive and negative) emotional aspects. They may include ad-hoc interpretations of the interviewer and a brief overall assessment of the interview together with remarks concerning dominant and neglected topics, all of which can become part of the process of analysis. The registration of this information is crucial for many reasons; it becomes especially important when interviews were not undertaken by the main researcher, when they are analysed in a team, or when they are used for qualitative secondary analysis. In all of these Page 2 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

cases, the problem of ‘not having been there’ is most obvious (Heaton, 2004). The postscript then has the advantage of providing some first-hand background information about the interview process and the relationship established between the interviewer and the respondent. The example of a postscript (Box 5.1) produced by one of the interviewers in the frame of a study of young people's perspectives on unemployment in post-communism (Reiter, 2008a: 75–6) is able to illustrate the immediate value of such a protocol. It instantly takes the reader into the situation. Together with the basic social background information it bears witness to the difficult circumstances of both interviewing and giving an interview. In short, Dimitrijus, as the young man had called himself, fell asleep during the interview after some 20 minutes. He was ready to participate in the interview and interrupted his work cleaning the windows of a supermarket in Vilnius in the cold month of February to follow the interviewer to a room in a nearby social work institution. Obviously, he was overwhelmed by the warmth of the room, the soft sofa and his physical inactivity. The interviewer, on the other hand, was desperate and made some unsuccessful attempts to get him back. It was her first interview and although she had been prepared for quite a few exceptional experiences she did not expect this to happen. The interviewer's postscript and the notes concerning Dimitrijus's social background provide a vivid picture of the situation. In the end, this interview could not be fully included in the study since it was incomplete. Yet much of it reads like a summary of the findings. And many details are striking, for instance: the young man's appearance, his poor education, his age when he started working, his parents’ unemployment, his striving for independence, his plan to go abroad, the ‘helplessness’ of the public labour exchange, his ethnicity, the fact that this respondent is male, even the interviewer's reaction.

Box 5.1: Postscript and Social Background Information of a Young Man (Study C) Respondent: Dimitrijus (20); 02/2004; 40 min. Place of interview: It was a room in the basement. There were soft furnishings in the room. It was a very quiet room. The first impression: The first impression was very bad. First of all, we had to go and pick him up from the supermarket not far away. He was working there (cleaning windows). Second, he had a lot of wounds on his face, hands; probably a day or two days ago he had a fight or something like this. It really didn't look nice. Third, I understood that he is Russian and got scared (for reasons of language; the authors). So I started the interview not in a very good mood. The appearance of the respondent: Well, the respondent looked really poor. His clothes were really very dirty (and he wasn't wearing a uniform during his work; it was his regular

Page 3 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

clothes - jeans and sweater). He looked just the same as people you sometimes see living in the streets. I should say he looked like a very unpleasant young man.

Behaviour of the respondent: Well, considering the fact that it was my first interview, his behaviour for me was more than shocking. He was gaping all the time and finally fell asleep. When I woke him up he told me that he worked all night and all day, so he was exhausted. And it was warm and nice in the room so it was a perfect place for him to sleep (it was cold and rainy outside). The flow of the interview: He started to talk about work because he was informed (by the social workers; the authors) that the interview will be about work. He was looking for concrete questions; he was too tired to think. I tried to involve him somehow; I started to skip from one topic to another but finally I had to give up. To see him with closed eyes and gaping all the time was too much for me for the first time. Education: Compulsory education completed with certificate. He studied to become a cook for some time. Activities with regard to employment and work: The respondent worked in a building lot; he washed cars in the streets, sold newspapers. He has been working from the age of nine, because he didn't want to ask for money from his parents; he wanted to earn it by himself. He is searching for a job with the help of acquaintances. He was registered at the labour exchange for half a year but, as he said, ‘There was no help’. Ethnicity: Byelorussian - Citizenship: Lithuanian Family: The respondent's parents are married. They are both unemployed. Housing: The respondent lives with his parents. Preferred choice of career: Future plans - to go to work abroad. He has a dream - to go to the army and to stay there. Mostly he likes to work with cars, repairing cars. Indicative and dense remarks like these about atmosphere, location, emotions and flow of the conversation can help interviewers to recall details that would otherwise easily vanish from memory. Authentic and ‘thick’ postscripts allow the interviewer to mentally return to the original situation and to how it ‘felt’ even after years. In any case, length is not the main criterion here. Postscripts may turn out to include valuable clues for later analysis and interpretation and can contribute to consolidating the overall impression of the research subject. Like in the above example, they may even be used to support and validate the overall coherence of ‘findings’. As social research is not usually accompanied Page 4 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

by professional supervision - unlike social work or psychotherapy where conversations are characterised by a similar degree of intensity - postscripts can have an important psycho-hygienic function. They should be used to jot down emotions and to record impressions that burden the interviewer, or that burdened the interview. Finally, postscripts force interviewers to critically reflect their interviewing technique and help improve interviewing skills.

5.2 Transcription How and what parts of the interview should I transform into text? The transcription of the main parts of the interview and of additional passages from preliminary or concluding talks is indispensable for the analysis and interpretation of interviews. Yet transcribing can be extremely timeconsuming, and as the transcribed interviews are, first of all, a means to an end - i.e. transcriptions make interviews accessible for processing them in the analysis - they should be produced with consideration. If the transcription is done by the researcher, she should take time to make notes and write memos along the way. This is an invaluable source of interpretation and of a first understanding of the meaning of what has been said. The interview you hear when you transcribe is different from the one you hear while interviewing: the focus of attention can now be broadened, thoughts can unfold, and the tension of immediacy inherent in the interaction in terms of asking follow-up questions or active listening and non-verbal reactions is gone. You will have first key insights, develop hunches and identify statements for analysis. Almost always this process of careful listening to the interview is accompanied by the frustration of discovering lost opportunities for follow-up questions and information gaps. This is unavoidable. Write it down and perhaps some of it can be clarified during a second interview or other opportunities. There are many different conventions for transcriptions – ranging from extremely detailed and complex (as for conversation analysis) to pragmatic (for example, in studies which involve, for instance, the translation of the transcribed material (Kvale and Brinkmann, 2009: Chapter 10)). In the former, dialect, intonation, turntaking and overlap, breaks, etc. are painstakingly transformed into text and signs following highly codified conventions (e.g. Ten Have, 2007). In the latter, translation costs and the unavoidable ‘foreignisation’ of the originally spoken language through translating it (Temple and Koterba, 2009) requires different standards of both transcription and analysis. Adding punctuation to what is heard (on tape) is already a form of interpretation of spoken language and suggests a certain structure of meaning. And two independent transcribers are likely to produce two different versions of transcribed text, adding issues of ‘intertran-scriber reliability’ to the list of quality criteria in research (Bucholtz, 2007). Translation is tricky without imposing your own order in the flow of words during transcription. As a general rule, the purpose of the interviews within the research design, together with the method of analysis, should determine how detailed the transcription should be. Transcriptions should be produced precisely for this purpose. It may be advisable to produce more detailed transcriptions only for selected key Page 5 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

passages. Interview transcripts prepared for publication will likely be revised once more, in order to provide readable accounts of what has been said. It is advisable to re-check key passages in the original recording in case of doubt or when they support particularly far-reaching conclusions.

5.3 Analysis and Interpretation What should I do with all this information? The interpretive paradigm does not suggest a method of choice for the analysis of interview data. This applies also to the PCI. In principle, PCIs can be analysed using the whole spectrum of available techniques for text analysis. Like the technique of data collection, the choice of the technique(s) for analysis has to follow the principle of object orientation and depends on the research questions and design. They can range from reductive content analysis to extensive biographical reconstructions, the integration of quantitative and qualitative evidence, or forms of sequence analysis including the ‘art’ of objective hermeneutics that is particularly relevant in the German context (e.g. Bryman and Hardy, 2009; Charmaz, 2006; Dey, 1993; Fielding and Fielding, 1986; Gibbs, 2007; Miles and Huberman, 1994; Reichertz, 2004a; Schreier, 2012; Silverman, 2006). Also the possibility of secondary data analysis should be considered before actually collecting new data (Heaton, 2004; Medjedović and Witzel, 2010). The question of qualitative analysis cannot be discussed exhaustively in this book with its focus on PCIbased data collection. Thus, we restrict ourselves here to providing a general outline of how the principle of

problem centring can be considered also in the analysis and interpretation of the interviews. In the sense of an overview it is a somewhat stylised account that touches upon the key issues without being able to specify every aspect. The backbone of the problem-centred approach to analysis and interpretation is, like the interviewing technique, 30 years old. In the meantime, single aspects of analysing qualitative data have been further developed. The use of software also greatly facilitates all steps of the analysis. The early reflections of the use of software for computer-assisted qualitative data analysis (CAQDAS) were closely related to the ‘Bremen school’ of methods development (e.g. Kelle, 1995), which was the primary environment of the original design and application of the PCI. Since then, the use of computer software for qualitative data analysis has become state of the art. However, the ease of text management should not make us forget to organise our search for meaning according to the basic principles of the PCI. The purpose of this chapter is to exemplify how these principles play out in the praxis of analysis and interpretation (see also Kühn and Witzel, 2004; and Witzel, 1996). The starting point is one of the key distinguishing features of the PCI: already the interaction and communication during the interview is dedicated to the production and clarification of (pre-)interpretations. Although these interpretations are preliminary, they need to be considered in the systematic analysis that follows. The design of the PCI as a technique of dialogic

and interpretive data collection recognises the dialogic interview interaction as a form of joint negotiation of Page 6 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

meaning and understanding.

Basic Approach to Analysis and Interpretation – Open, Axial and Selective Analysis The specific interaction and dialogue during the PCI on the basis of communication strategies and preinterpretations produces interpretive knowledge. This knowledge from the interview enters the process of analysis as a form of interpretive prior knowledge, which needs to be explicated and ‘controlled’. In principle, the interpretation of the material continues the dialogue between researcher/ interpreter and respondent but without the actual participation of the respondent; she is now represented through the interview material. The basic approach of the analysis is the following: original interview material, interpretive prior knowledge (i.e. revised pre-interpretations from the conversation, and notes included in the postscript), and interpretations from the ex-post analysis (considering the sensitising conceptual framework, the interview guide and the sampling criteria) all enter into an iterative process of constituting, examining and refining interpretive

hypotheses. As the analysis evolves, these empirically-grounded hypotheses are gradually corroborated and substantiated, and finally presented as generalisable ‘findings’. The practical introduction to Grounded Theory by Strauss and Corbin (1990) can be used as a general guide to the analysis of PCIs. In particular, the suggested forms of ‘coding’ are connectable to the problemcentred processing of interviews. They distinguish between forms of open, axial and selective coding that refer to different steps in the analysis. As they define ‘coding’ very generally as ‘the process of data analysis’ (ibid.: 61), we use, in the following, the distinction between open, axial and selective analysis. This comes closer to the comprehensive procedure that the processing and interpretation of data implies. Coding, in the conventional and narrow sense of indexing data and labelling phenomena, is only one part of it. The basic step of open analysis is used initially for understanding and appropriating the texture of the interviews through a ‘process of breaking down, examining, comparing, conceptualising, and categorising data’ (ibid.: 61). After first ideas about how to interpret and categorise interview passages are available, the next step of axial analysis starts ‘making connections between categories’ (ibid.: 96). At this point, first causal assumptions are investigated about the connections between the respondents’ opinions and actions and their context, conditions and consequences. If the interview material is very extensive, we recommend doing these first two steps on the analysis on the basis of a selection of interviews. Otherwise the efforts involved in the iterative qualitative analysis may get out of hand. The third step of selective analysis examines and differentiates the findings from the first two steps, which have the status of interpretive hypotheses. This step focuses the analysis on certain aspects and deepens related interpretations by organising the analysis along ‘core categories’. It is a ‘process of selecting the core category, systematically relating it to other categories, validating those relationships, and filling in categories that need further refinement and development’ (ibid.: 116). In case the first two steps were done on the basis of a sub-sample of the interviews, the core categories now define the direction of interpretive attention which is also applied to the remaining material. Page 7 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

It is important to understand that the order of these steps is not static but can be variegated. The process is rather cyclical and iterative. Openness and theory-driven interpretation do not form a contrastive pair. In line with the overall design of the PCI, this approach emphasises that the researcher is involved in both data production and interpretation. The process of reconstructing meaning out of the material and then submitting it to scrutiny on the basis of other sections of the data corresponds to the PCI's basic principle of combining inductive and deductive moves which are essential to any kind of (hermeneutic) understanding (see also our thoughts about the ‘Dr Jekyll-Mr Hyde syndrome’ in section 2.3). Apart from the collected and successively transcribed interview material, the ‘ingredients’ of the analysis and interpretation of PCIs mainly result from research questions, the sensitising conceptual framework, the interview guide, the sampling criteria, the (revised) pre-interpretations from the conversation, and notes included in the postscript. The actual analysis starts with (i) the basic coding and indexing of the interviews regarding topics and pre-interpretations. We illustrate it with an example from STUDY A (see example in

Table 5.2). This is followed by two further steps that could be called (ii) vertical analysis and interpretation related to single cases and interviews, and (iii) horizontal analysis and interpretation in the sense of crosscase analyses. The typical steps involved in analysing PCIs are summarised in Table 5.1.

Table 5.1 Typical steps of analysing PCIs (i) Basic coding and reconstruction of pre-interpretations

–

Key categories from the interview guide

–

‘In-vivo-codes’

–

Interpretive codes and first interpretations

(ii) Vertical analysis and interpretation

–

Case descriptions and biographical chronology

–

Dossier of the interpreter

–

Identification of key themes

–

?

Biography-oriented analysis (gestalt)

?

Topic-oriented analysis

Validation

(iii) Horizontal analysis and interpretation

–

Thematic cross-case analysis

–

Synthesis and representation of findings

?

e.g. Empirically-grounded concept development

?

e.g. Empirically-grounded typologies

(i) Basic Coding and Reconstruction of Pre-Interpretations Page 8 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The basic way of coding and indexing a PCI does not follow any specific rules and is best illustrated with an example (see Table 5.2).

Page 9 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Table 5.2 Interview with a metal fitter (STUDY A)

Page 10 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Page 11 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The example in Table 5.2 includes two basic forms of indexing and coding. The column on the left contains (theory-driven, deductive) key categories from the interview guide and the conceptual framework as well as terms referring to what is relevant for the respondent himself (according to the principle of openness). This is a rather descriptive step that identifies topics and stages of the biographical process by means of a list of categories established in advance and extended in the process of coding. This step prepares thematic text retrieval and analysis and is indispensable for the reconstruction of the respondent's line of argument, which can contain mental loops, reconstructions from different angles, repetitions, spontaneous ideas and recurrent specifications of topics due to the gradual mobilisation of memory. The column on the right serves a more analytic purpose. It contains two types of information. On the one hand, ‘in-vivo-codes’, i.e. catchy everyday terms, can point to important topics and stages. On the other hand, and like in the example, it is open to various forms and degrees of interpretation ranging from simple exclamation or question marks to short sentences with the character of ‘memos’ in the sense of Glaser and Strauss (1967). A basic form of coding like this, where text is assigned to categories that serve as virtual containers of meaning, is unavoidable for any kind of text analysis. There are specialised books and papers dedicated to issues of software-supported coding and text retrieval (e.g. Lewins and Silver, 2007; Saldana, 2009). In any case, it should be done with software for computer-assisted qualitative data analysis (CAQDAS). For an extended description of how to use a qualitative database for the analysis of PCIs see Kühn and Witzel (2004).

(ii) Vertical Analysis and Interpretation Case descriptions and biographical chronology. After the basic coding and indexing are done, case descriptions summarise the main features of a single case on a few pages. They have the purpose of familiarising oneself with details of the cases by sorting, for instance, thematic information scattered across the whole interview. The case description also includes the respondent's interpretation of situations and her reasoning behind decisions and actions in certain contexts. In the example in Table 5.2, the unrealised options and related retrospective interpretations were also included in order to represent the whole process of searching for professional training. Especially in projects with more than one researcher, as well as in longitudinal research, short (!) case descriptions are absolutely critical. They help to share the material, to prepare the second or third round of interviewing, and to facilitate re-analysis. If the research is interested in micro-causalities on the level of individuals, the biography of the respondent is also reconstructed and events are ordered chronologically; this explicates the course of action and its temporal dimension. This step can be combined with aspects of the conceptual framework of the study which can then be used in the sense of a heuristic (see example in Box 5.2).

Box 5.2: Biographical Chronology on the Basis of the Conceptual

Page 12 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Framework (Study A) In STUDY A, the reconstruction of biographical chronologies followed the so-called ARBmodel (Witzel, 2001), which is described in more detail in section 6.1. In short, the three dimensions of aspiration, realisation and balancing (ARB) serve as a simple action model and heuristic tool for the organisation and interpretation of the material. Each biographical stage related to the respondent's professional development and was analysed in the context of other options, in relation to concrete steps of realising them, and with regard to associated evaluations. The biographical chronology, which is part of the description of the case of the metal fitter (see Table 5.2), organises the material according to these three categories.

Secondary School Leaving Certificate Bike mechanic (not realised option) Aspiration: He wants to turn his hobby into a profession and has six years of experience in a bicycle shop. Realisation: He submits unsolicited applications and accepts a detour through a job creation programme. Balancing: He gives in to the pressure of his father who does not accept the detour.

Apprenticeship as Electrician Aspiration: He gives in to the pressure of the father and his opinion that an apprenticeship has to correspond to the needs of the labour market. Realisation: His father organises access to an apprenticeship in small enterprise. Balancing: He tries to change the company because of negative working conditions (e.g. non-compliance with security regulations, work instead of training) through unsolicited applications and with the support of the labour exchange. The attempt fails: ‘These craft enterprises are in cahoots with each other. Yes, they also call each other, and the foreman did not want to let me go.’ He develops psychosomatic problems and terminates the training.

Apprenticeship as Machine Fitter Aspiration: He recalls his original wish to become a mechanic and receives the suggestion of an acquaintance that has learned the profession. He wants to work in the maintenance

Page 13 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

of machines and plans to apply his knowledge in his hobby (‘to build a tandem or so’).

Realisation: He insists on an apprenticeship as machine fitter at the labour exchange. Balancing: He completes the apprenticeship successfully. The training is better than being in a crafts enterprise (e.g. good atmosphere). Later goal: his own bicycle shop. ‘I learned this profession more because of the skills that you learn, in order to follow my own interests.’ The aim of this step is not yet that of a fine analysis of the material. The case description, as well as the biographical chronology, like in the above example, should first of all guarantee the application of two basic principles of analysis of cases that are emphasised by Gerhardt (1986): the preservation of both the meaning

context and the character and progression of the case. She writes: ‘At the end of this understanding of the case we have an outline of the course (of action and events) that proceeds for each case in terms of phases, steps, turns and new beginnings until we reach (only analytically relevant) preliminary final points, which can be relativised on the basis of new research later on’ (Ibid.: 68). As these criteria apply to both steps, the case description can even be replaced by an extended biographical chronology - in our example, it can be combined with underlying conceptual heuristics and illustrated with representative quotations. According to the principle of problem centring, analysis and interpretation do not follow a naive-inductivist approach - theoretical assumptions have to be explicated and specified. At the same time, and in line with the interpretive paradigm, the actor's perspective has to be preserved: theoretical constructions must not superimpose relevancies and action orientations of the respondents. The sensitising conceptual framework of organising and explicating prior knowledge (see section 3.2) and heuristics like the ARB-model have exactly this function of combining openness with theory-guidedness. Altogether, by applying a heuristic-analytic framework, the basic approach of the analysis of PCIs follows the procedure described in Strauss and Corbin (1990). After the open analysis, the second step of axial analysis operates on the basis of a ‘coding paradigm’. This paradigm model of analysis essentially represents a causal model which includes context, causal conditions, intervening conditions, strategies of action and interaction, and consequences of action. In methodological terms, the coding paradigm is not a theoretical model that could be tested in the sense of a deductive approach. It rather consists of elements that are abstract and

empirically not contentful (see section 3.2) and cannot be falsified. In the sense of a ‘theoretical axis’ the action model serves only as a frame for formulating an empirically contentful and concrete middle-range theory. This theory is gradually constructed on the basis of the data by filling up the theoretical axis with empirical content – i.e. with conditions, strategies and consequences of action that characterise the field of research (Kelle, 1994). In the course of such a procedure, theoretical statements can only be formulated afterwards, which requires specific strategies of validation; they are described below.

Dossier of the interpreter. In addition to the case description, the interpreter should write a short comment Page 14 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

about the texture of the interview material and the particularities of the case as well as interpretive obscurities. This dossier briefly discusses the detail of the explorations and the narratives as well as methodical errors, unanswered questions and obvious contradictions in the material. This view of the interpreter should be compared to the postscript of the interviewer to get an additional perspective on specific problem constellations and extraordinary courses of events.

Identification of key themes. The case-specific development formulates ideas for the interpretation that are oriented either towards topics or towards the biography of the respondent. A topic-oriented analysis narrows the interpretation to certain issues like, in the case of the example above, unemployment, dropping out, moving out from their parents’ home, or the specific contextual conditions. A biography-oriented analysis focuses the interpretation on the gestalt of the case and reconstructs interpretive patterns and action patterns that are typical for the case. Here, we can only outline the topic-oriented interpretation of PCIs on the vertical level leading to empirically-grounded typologies after the additional horizontal analysis and interpretation (see below). For reasons of space we cannot cover the alternative focus on biographical case reconstructions (see, for example, Rosenthal, 2006). The first step of the topic-oriented analysis consists of summarising recurrent issues in narrative and dialogic sequences of the single interviews into key themes. Each of these key themes is synthesised by means of a condensed summary combining original text passages, paraphrases and analytic statements. The search for key themes corresponds to the step of open analysis following Strauss and Corbin (1990). In order to underline the openness of this step and its closeness to the relevance structure of the respondent, it can also be based on in-vivo-codes. The second step examines the relation of relevance structures and context

conditions in a way that the process character of the case is maintained. In the example of STUDY A, the ARB-model was used to explicate the dynamic of biographical acting by means of a fine analysis of the interaction between, for instance, decisions with calculated and unintended consequences on the one hand, and the social context and the living circumstances on the other (see section 6.1). By means of the ARB-model it was possible to identify continuities and breaks in the biographical development and how discrepancies between individual aspirations and institutional expectations were dealt with. In this way, an action-theoretical frame was integrated into the analysis of key themes and it was possible to understand the subjective logic of the case in the perspective of these themes. This second step corresponds to the idea of

axial analysis integrating open and theory-driven procedures. The attribution of text passages to key themes and their labelling constitutes the first level of concept formation. It goes hand in hand with the formulation of empirically-contentful categories and theoretical statements which need to be examined and validated.

Validation. Theoretical statements resulting from qualitative data analysis as a dialogue between empirical material and prior knowledge may be empirically grounded. Yet in methodological terms their status as ‘findings’ is still preliminary and ‘precarious’: single text passages can be interpreted in more than one way, i.e. usually there are several possible readings. At first, theoretical statements therefore have the character of preliminary interpretive hypotheses that still need to be validated and assured; their value needs to be Page 15 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

controlled and checked. The issue of validity of findings is central to qualitative research; in fact, it is one of the main areas of its legitimisation vis-à-vis quantitative approaches. Thus, the list of criteria of validity in qualitative research is expanding together with the related body of literature (e.g. Flick, 2008; Seale, 1999). We want to restrict ourselves here to suggesting a minimum of two strategies of validation: the first uses the text, or the case, as a source of control; the second uses the opinions of different interpreters as a source of control. In this way, the PCI's dialogic approach to (re)constructing results is continued on a different level. The first strategy of validation has the purpose of examining, substantiating, modifying or rejecting the various interpretive hypotheses developed by the research team on the basis of the interview or the information available for one specific case. In short, an interpretive hypothesis can be maintained when it is

empirically saturated, i.e. when no counter-evidence can be found in the material. In methodological terms, this corresponds to a basic falsifiable attitude during data analysis: interpretive hypotheses can be falsified on the basis of counter-evidence, yet the lack of counter-evidence does not necessarily imply their ‘verification’ (see Popper, 1989: 47 ff.). In practice, interpretations and counter-interpretations should be developed and confronted in the course of the analysis until the most convincing one is left. Sometimes it could be useful to contrast the evidence with that from other cases already at this stage. The interpreters need to have access to earlier interpretations and to the original transcripts and sound or video files at all times. The PCI is particularly suitable to being subjected to this kind of validation because of its dialogic reconstruction of

problems. The interview is designed to provide a gradual specification, already during the conversation, of the issue under investigation by means of an interaction of various forms of prior knowledge as well as preinterpretations developed during the interview on the one hand, with the respondent's point of view on the other. In particular, the communication strategies of specific explorations, which ask the respondent to refer to key issues repeatedly, are able to address resulting contradictions that arise in the course of the interview. At least some of the analysis should be accompanied by a second form of validation in a team of researchers debating the key themes and their interpretation in working groups or workshops dedicated to it. Also, additional empirical material not directly associated with the case but nevertheless relevant can be brought to the interpretation. These groups discuss pre-interpretations of the first interpreter as well as suggestions and

opinions of the different interpreters in order to further explicate theoretical assumptions and to differentiate the range of possible readings and ways of understanding the material. The search for the most probable interpretive hypothesis goes through a process of discussing many. In order to introduce some limits to the imagination of the interpreters they are required to relate their ideas back to the text by searching for associated evidence and counter-evidence. In this way, selected empirically-grounded hypotheses are gradually eliminated in a process combining both strategies of validation.

(iii) Horizontal Analysis and Interpretation Thematic cross-case analysis. Our theoretical knowledge grows with every additional case that is analysed. From the analysis of the second case onwards, we are involved in a process of comparison: every additional case is interpreted against the background of the cases before. Openness towards the new information Page 16 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

once again enters into a dialogue with our knowledge interest that is gradually consolidating. It would be unwise not to use the tension inherent in this confrontation for the development of new ideas. Ideally, this tension is resolved in an abductive flash of inspiration that discovers novel insights into the structure of the data (Reichertz, 2004b; see Box 2.4). Yet the interpretive movement is iterative and every additional step of validation (forwards) potentially requires the re-analysis of ‘older’ cases (backwards). It is not difficult to imagine that this iterative interpretation bears the risk of exponentially increasing the amount of work involved in the qualitative analysis. One way of escaping this risk is applying a process called systematic contrasting

of cases: usually the comparison involves a sub-sample of cases that are least or most similar with regard to key themes as well as background categories like gender, residence or profession. This form of horizontal analysis differentiates the key themes with regard to all cases in the sample. With the aim of developing key categories, cross connections are established and reflected in memos and short theoretical discussions. The development of empirically-grounded typologies and concepts is one way of consolidating and synthesising

findings from such a thematic cross-case analysis. It is a procedure that is closely associated with the ‘Bremen school’ of social research (see Kelle and Kluge, 2010; and Kluge, 1999, 2000). We want to illustrate some of the related general steps here on the basis of STUDY A, which was also the background of the example above.

An

Example:

Empirically-Grounded

Development

of

Modes

of

Biographical Agency In STUDY A, the analysis of cases like the above example of the metal fitter was organised towards the development of a typology of modes of biographical agency regarding professional careers (see Heinz, 2002; Witzel and Kühn, 2000). It was based on the idea that the analysis of the interviews could be summarised along one central theme – here: biographical self-realisation in view of constrained opportunity structures – which in turn served as the basis for a final round of re-analysis. Thus, the chronology of biographical options was analysed again in more detail, and the process of balancing and evaluating the realisation of biographical options and aspirations was scrutinised in the perspective of the respondents (i.e. ARB-model) (see section 6.1). More concretely, the analysis focused on the following aspects, trying to synthesise them into profiles: the respondent's way of dealing with biographically relevant institutions; their coping with experiences related to their status passage to the world of work; and the various interests, results and evaluations related to their efforts. On the basis of the respondents’ reflection of experiences, two key principles structuring biographical agency were identified: the relationship between self-assertion and self-development on the one side, and the limits and restrictions to the formation of one's biography through agency on the other. The different modes of biographical agency were the result of the respondents’ practical involvement in and processing of options. They represent coagulated knowledge resulting from a chain of aspirations, realisations and balancing; they are ‘styles’ of dealing with structural requirements on the basis of their own claims. The aim of the contrasting case analysis was the development of a generalisable set of modes or types of biographical agency. This was done in four conceptual and empirical steps. First, the main aspect of each Page 17 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

type was defined and examined for its compatibility with the respondents’ balancing and appraisals. Second, the typical patterns of behaviour associated with each mode were briefly described and specified. Third, the brief description of the previous step was complemented by a more extensive empirical description based on key quotations and longer paraphrases especially of appraisals related to realised options. In this way, the contours of the different types were gradually unfolded and sharpened. Biographical turning points and reorientations in single interviews were taken as indicators for a transition from one dominant biographical mode to another. The in-depth analysis of subjective reasons for such a transition contributed to the conceptual

strength and discriminatory power of the different modes of biographical agency. Finally, the types were consolidated by integrating additional aspects (e.g. the communication in families or partnerships), which supported and extended their profile.

Selective coding and analysis, and differentiation of dimensions. In practical terms, the development and intensive validation of these modes of biographical agency were grounded on a sub-sample of 20 cases. A consideration of more or all cases for this iterative step would have greatly slowed down the whole process of interpretation. The resulting preliminary typology was then used in the sense of an interpretive hypothesis and applied to the remaining empirical material. In a quasi-deductive and theory-driven step of selective analysis (Strauss, 1987: 14 ff.) the dimensions of the typology were further differentiated. Their commonalities and differences crystallised in the process of a targeted search for evidence and counter-evidence. Gradually, the preliminary typology was modified and consolidated in a process of empirical saturation. This step was also accompanied by validation as well as the partial re-analysis of interview passages. The intensive interaction

of deductive and inductive steps in the process of selective analysis turned out to be the most time-consuming part of the analysis, because at this stage, each step has to pass through the whole body of interviews. Especially here, the use of data analysis software is indispensable and greatly facilitates the search process. This is why the careful preparation of the typology in its preliminary stage on the basis of selected interviews, and before it is applied to all interviews, is most important.

The typology as a way of representing the findings. The final typology of modes of biographical agency consists of six types that are distinct regarding the orientation towards the five dimensions of work, qualification, career, income and company (see Heinz, 2002; Witzel and Kühn, 2000). The category of status arrangement represents two modes of biographical agency that consider the professional biography as more or less concluded. According to the mode of company orientation, young adults anticipate the end of their professional development and strongly associate it with the company as a kind of professional home that provides support, security and interpersonal recognition in exchange for trust and loyalty. The mode called

wage-worker habitus describes young people that also search for professional continuity but would change job if given a better offer. They consider employment as a necessity and try to establish a pragmatic balance between investment and financial rewards. The category of advancement ambition includes two modes with a strong emphasis on professional progress. The mode of career orientation refers to young people oriented towards promotions within companies and according to institutional mobility pathways, while the mode of optimising chances is characterised by striving for a multitude of options by accumulating additional competencies and qualifications. The category of gaining autonomy represents two modes of biographical Page 18 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

agency that are characterised by a certain distance to dependent employment. In the mode of personal

growth the occupation is chosen according to possibilities for self-realisation, and the professional career is kept open. Finally, the mode called self-employment habitus focuses on independent entrepreneurial activities and business success. A typology like this is one way of representing findings from PCIs. Box 5.3 illustrates a second example of a typology resulting from a horizontal cross-case analysis in STUDY C that followed the interpretive hypotheses established on the basis of a vertical in-depth analysis of a single case.

Box 5.3: From Single Case Interpretation to Typology (Study C) In STUDY C the analysis of the interview material followed the two levels of the sensitising framework - the biographical level of uncertainty and the normative level of recognition (see Box 3.2). The first strand of vertical analysis (I. in Figure 5.1) concentrated on individual cases and captured - by considering and reconstructing single cases in their entirety - the biographical context of work, and unemployment-related accounts in individual transition experiences. The second strand of horizontal analysis (II. in Figure 5.1) cut across cases and provided - through a cross-case analysis of categories and topics - a differentiated picture of the normative validity and meaning of work and unemployment in the young people's accounts. By systematically contrasting and comparing cases and categories, this mode of analysis followed the strategy of developing empirically-grounded typologies (III. in Figure 5.1). Following the PCI's principle of process orientation, prior knowledge in the form of sensitising concepts (SC) was relevant at various stages of the analysis. Findings were developed gradually and empirically grounded in a problem-centred dialogue between these concepts and research interests on the one side and the empirical material on the other.

Page 19 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 5.1 Flowchart of data analysis and interpretation (STUDY C)

The process of analysing data started with a numerical classification of the 30 cases through cluster analysis on the basis of the ten work- and unemployment-related items included in the brief questionnaire at the end of the interview (see Box 4.7) (O. in Figure 5.1). This step prepared the further analysis in two directions as indicated above. On the basis of this classification, eight so-called ‘anchor cases’ were identified and subjected to an in-depth interpretive case analysis (I.). Pragmatic reasons behind this reduction of material were the necessity to get the analysis started, and to keep the time needed within a feasible limit. In addition, an extensive analysis of smaller parts of the data at an early stage and its later confrontation with the remaining data corpus are advisable in the frame of what Silverman (2000: 179) calls a ‘constant comparative method’. It is Page 20 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

part of a strategy for increasing validity. The numerical classification should guarantee the consideration of the possible breadth of the data and reduce the arbitrariness of case selection. The comparison of whole cases revealed biographical patterns for individual transitions as a background of work- and unemployment-related accounts. The intensive work with very few cases on certain thematic issues also determined how to arrange the findings from the analysis into a readable form. This in-depth analysis of cases generated a set of common topics resulting in a list of general and specific codes, which were afterwards organised into thematic code families. These general and specific codes were then applied to all cases (including those used to produce the codes); the list was complemented by a few extra codes as the analysis moved on. The inclusion of all interview material at this stage (II.), as well as in later steps of the analysis, should meet the ‘totality demand’ expressed by Silverman (2000: 180) when calling for an inspection and analysis of ‘all parts’ of the data ‘at some point’. The application of the same coding system to all interviews should ensure comparability. The interpretation produced empirically-grounded typologies regarding three topics (III.): (i) patterns of youth transitions and related time perspectives of biographical uncertainty (see Reiter, 2009a, 2010a, 2012); (ii) young women's imagined gender-work relations (see Reiter, 2008b, 2010b); and (iii) a heuristic typology of relations between individuals, the state and the unemployed in society. In the following we outline the third classification and how the cross-case analysis on which it is based can be read as a ‘response’ to the findings from a single anchor chase (with the pseudonym Saulius) where the most relevant issues are addressed. In order to illustrate the first step of orienting the cross-case analysis along criteria from single cases let us look at one crucial segment from the interview with Saulius. It is the first time during the interview that he is asked to talk about the issue of unemployment. His answer helped to establish the three main dimensions of the cross-case analysis that followed.

I: OK, now look, regarding this work theme, but something a bit different: when there's no work. The fact is that there are many unemployed in Lithuania; unemployment is high. What do you think could be the reason for that? S: Hmm. [These] people are lazybones. [He is laughing] … Those who want [one] can surely find a job. It is not that … there is no work … [as though one were] in some village: ‘I came here and cannot find a job.’ You have to look for it: those who look for work, find it. If you lie on a couch with a glass of brandy [laughs], life will surely not get better because of that. The state, of course, cannot create work for everybody, but for many. If you cannot find a job, so what? You can go abroad.

Page 21 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

I don't think that it is absolutely impossible to get a job. Of course, it is possible. But sometimes people are very demanding and lazy, exactly those who do not have [work]. They think that it is much easier to live out of the state's pocket by having lots of children, that that is much easier than to work like normal people (emphases added; authors). The case of Saulius has many aspects that have to be neglected here (see Reiter, 2007a, including a full display of the relevant interview passage). In any case, his opinion provided a first and preliminary interpretation of young people's views on the problem of unemployment. In the sense of a specific conclusion (concerning one case and based on a vertical analysis) it suggested a universe of possible general hypotheses qua directions of further cross-case analysis (of all cases) as well as relevant key dimensions of such an analysis. In this passage, and throughout the whole interview, Saulius …

a b

c

… constructs an image of the unemployed (including that of ‘lazy’ alcoholics); … reflects citizenship relationships established on the basis of the new role of the state (‘The state … cannot create work for everybody, but for many’); … discusses possible responses to the new problem of mass unemployment (including going ‘abroad’ for work or ‘having lots children’ for benefits).

In the next horizontal step of the analysis, these three dimensions were further explored, substantiated and diversified on the basis of all cases. This systematic analysis of all 30 cases resulted in a differentiated picture of the problem based on the construction of a three-dimensional space of the typology of relations between individuals, the state and the unemployed in society (see Reiter, 2007b, 2008a: Chapter 8, 2009b). In relation to (a): The first dimension of the image of the unemployed represents private notions of recognition or misrecognition (in the sense of Honneth, 1996) towards this new societal group. It ranges from sympathy to negative stereotype. The young people's perception of the unemployed is not one-dimensional or entirely negative; sometimes it is even full of contradictions. Three elements can be distinguished.

i

ii

Page 22 of 32

Stereotypes refer to laziness or reluctance to take jobs, to the association of unemployment and alcohol, as well as to the abuse of benefits. Articulations of sympathy refer to expressions of pity, to the consideration The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

iii

of economic explanations like the lack of jobs, and to the declaration of the necessity of benefits for the mere survival of people. The reference to exceptions within the group of unemployed people (e.g. relatives, neighbours, etc.) points to individuals as victims of, for instance, the impact of the socio-economic transformation on their local working environment.

In relation to (b): The second dimension of citizenship relations refers to perceived public notions of recognition or misrecognition towards the unemployed. It ranges from contentment to frustration. State-citizen relations are established between the state and the (unemployed) individual, and both parties have their share of responsibilities in this relationship. The typology considers these two levels.

i

ii

Respondents discuss ‘duties’ of citizens to participate in employment as well as ‘duties’ of the state to facilitate or provide employment and welfare benefits. Notions of dissatisfaction with state performance are common. They are either directly associated with ‘politicians’ or ‘parliamentarians’, or with state agencies like the labour exchange. Or they are, sometimes implicitly, expressed in terms of an emphasis on advantages of the situation during Soviet Union times.

In relation to (c): The third dimension of response refers to three ways of answering to the unfavourable situation of mass unemployment by means of action. Hirschman's (1970, 1993) distinction of exit, voice and loyalty as responses to discontent was instructive here. This dimension ranges from loyalty to exit.

i

Going abroad as the legitimate exit is articulated in terms of plans or the chance to leave and in knowing people abroad. Depending on the perspective, emigration can be a solution (for some) as well as a suggestion (to others); it can be an individual option or an observed behaviour.

ii

The option of non-work refers to the refusal to work for low salaries and can be a form of protest or ‘voice’. A third way of challenging the triangle consists in forms of cheating the state for benefits associated with the stereotypical image of the unemployed. In the sense of a positive commitment it refers to a kind of

iii

Page 23 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

loyalty to the past citizenship relations of state socialism, where the entitlement of citizens to being supported did not depend on insurance from employment. These categories were further specified and opened up a space in which each of the respondents was positioned according to the interview. Five types of relations between individuals, the state and the unemployed were identified. Table 5.3 includes the detailed matrix of interview issues for five prototypical cases that were used to illustrate the types.

Page 24 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Table 5.3 Matrix of interview issues (STUDY C)

Page 25 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Page 26 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The ‘x’ indicates the ‘presence’ of a topic, meaning that related issues were approved or observed (in the interview, questionnaire, background information and postscript). Obviously, the ‘x’ does not indicate the specific quality of the observation. The subcategories emerged out of the interviews and were not already part of the interview design. Thus, the ‘x’ is missing where the issue was not observed, or where respondents indicated their disagreement. These five types, which are presented here in tabular form only (Table 5.4), provide a differentiated overview of the problem of unemployment in post-communism from the perspective of young people. The internal homogeneity of each type results from the specific combination of, and emphasis on, the single aspects of these dimensions. It is a heuristic typology and, as all the dimensions were not ‘designed’ a priori, they are not entirely separable but partly overlap.

Page 27 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Table 5.4 Heuristic typology of relations between individuals, the state and the unemployed (STUDY C)

Page 28 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

5.4 a Note on Resource-Sensitive Planning of PCIs Giving due consideration to time and other resources required to do an investigation is a notoriously weak spot of qualitative research design. Its greater flexibility and openness to discovering novel aspects of social phenomena appear to provide researchers with good reasons to neglect the concrete planning of their study. Furthermore, persuasive principles of achieving generalisable results or ‘conceptual representativeness’ (Strübing, 2004: 33) may encourage researchers to be reluctant to determine pragmatic boundaries of research. For instance, the rule of ‘theoretical saturation’ in sampling, according to Grounded Theory, can easily be misunderstood in this direction. Somewhat cryptically, theoretical saturation asks for continuing with the sampling until ‘no new or relevant data’ emerge, ‘category development is dense’, and ‘the relationships between categories well established and validated’. And, ‘[u]nless you strive for this saturation, your theory will be conceptually inadequate’ (Strauss and Corbin, 1990: 188). Rules and principles like these seem to suggest that there is no obvious and predictable end to qualitative research that could be anticipated and considered in the research design. Indeed qualitative research does not define standard cutoff points in terms of ‘enough’ observations. In practice, however, academic ideals like these are contrasted with increasingly restrictive resource requirements for research. With budgetary aspects gaining in importance, even studies for professional qualifications (i.e. BA, MA or PhD research) have to be done outside the academic ivory towers. PCI research is not different in this respect, and as soon as it targets external funding it needs to indicate in its research design both the time frame and resources involved. A few aspects can be considered in advance in order to arrive at a more realistic design of PCIs as part of the overall research programme. In principle, in this respect, it is first of all the research question that dictates, metaphorically speaking, where the journey goes and how hard it will be. Thus, frank answers to the following threefold question should be found as early as possible: in order to answer my research questions, what do I have to do and how, and what are my resources? An early and realistic assessment of required activities, time and human resources helps to move the project from mere interest and commitment, to feasibility. Box

5.4 includes a more extended list of questions for planning PCIs.

Box 5.4: Questions for Planning PCIs • How much time do I have for the study, and what are my resources? • Approximately how many interviews will I need in order to answer my research questions? How many can I realistically make, and approximately how long will they be? • How easy or difficult will it be to find respondents? Does it involve travelling? • How long will it take to transcribe the interviews, and what conventions of

Page 29 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

• • • •

transcription will I apply? Can I hire somebody to do this? How much material will I have to process? How long does it take to read through it, and how often will I be able to read it? What kind of analysis will I apply? Do I have access to software for qualitative data analysis? How much time should I set aside to interpret the material and to synthesise the interpretation into a final report?

Qualitative interviewing involves a series of steps, some of which need to be repeated for each interview. Thus, establishing a list of these steps, specifying what they include and making them visible in relation to the overall time frame, which is usually known in advance, is a good start. Apart from the optional training of interviewers, the main steps related to interviewing alone are: (1) identifying and locating interview partners; (2) arranging appointments; (3) travelling there (and back); (4) doing the interviews; and (5) transcribing, analysing and interpreting them. Depending on the sampling plan, the first three of these steps can sometimes be done for several interviews at the same time. It is very difficult to give an average estimation of time required for these steps, but we think that, following Flick's (2007: 52) conservative estimation, doubling the time of the interview itself will probably not be sufficient. Finally, getting the appropriate respondents in front of the microphone can be extremely unpredictable; postponements, lastminute cancellations and no-shows are common problems consuming both time and motivation. Anticipating these kinds of delays is strongly advised. For these reasons the consideration of secondary data analyses is also advisable (Witzel et al., 2008). The recorded part of a single PCI typically lasts between 30 and 120 minutes, with a standard length of about 60 to 90 minutes. The whole interviewing encounter may easily take about one to two hours or more: preparatory talks and getting familiar with each other before the actual interview starts, as well as certain conventions of debriefing and bringing the encounter to an end, need to be considered (see sections 3.5, 4.1 and 4.9). One may anticipate interviews with businessmen, politicians or experts to be brief and efficient. Yet the opposite may be the case. For instance, in the frame of her study of former communist elites, Obelene (2009b) once accompanied a businessman in his car for several hours from one business meeting to another in order to get the interview she needed. For the ‘average’ person, an interview may have the character of a special social event. Being invited to somebody's home for the interview usually triggers social conventions associated with such visits – a tour through the house or the property, or simply a cup of tea before and after the interview, may extend the whole encounter. Generally, this extra time is well invested: off-the-record exchange is invaluable for establishing trust and gaining additional information. However, it does make it difficult to estimate the duration of an interview in advance. As a rule of thumb, one PCI should be scheduled for one day; if it can be arranged easily, a maximum of two interviews can be done in a row. The time required for transcription depends on the writing skills of the transcribers, the technical equipment Page 30 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

– a footswitch to control the playback of the recording is essential – and, most importantly, the transcription conventions applied. As discussed in section 5.2, transcriptions may be extremely time-consuming and should first of all be adjusted to the needs of the analysis. Following pragmatic conventions, ten minutes of interview takes about one hour to transcribe, and therefore one hour of interview takes about six hours to transcribe. In this example, some additional two hours should be calculated for correcting the interview transcript. In other words, one full working day is easily absorbed by finalising a pragmatic transcription of one rather short interview. The transcription of all interviews can produce hundreds of pages of text. Check the time that it takes to read one single page and multiply it with the amount of pages! The resulting time is by no means also sufficient to do the analysis including the various steps of coding and interpretation. And how many times will you be able to read through the whole pile of transcripts? Furthermore, the quality of the interpretation improves if time is dedicated to single cases or interviews at the beginning of the analysis. The use of qualitative data analysis software (QDAS) is highly recommended (Lewins and Silver, 2007); getting familiar with it may take an additional week. The auto-coding options offered by these software programs are helpful and can greatly reduce the time it takes; yet, usually it can only be applied if the transcription is done in plain language and includes punctuation marks. The interpretation itself – together with synthesising the findings into case summaries, research reports or papers – can hardly be calculated in detail in advance. Forms of reductive or summarising analyses like qualitative content analysis (Mayring, 2004; Schreier, 2012) require less time than forms of extensive and explicative data analysis that quickly multiply the amount of textual material that needs to be processed (Reichertz, 2004a). In reality, the depth of analysis and interpretation too often simply depends on the time that is left. Thus, ideally, at least one-third of the overall duration of the project should be reserved for this last step. Add more time if you are an absolute beginner! Again, the realistic anticipation and planning of all the activities that these steps involve is the key to quality. Altogether, calculating one week of full-time work per interview for conducting, transcribing and coding it is not too much. Now, let us also take a look at the resource side of things. The monthly average working time of one person working full time (and not more) amounts to some 170 hours. Considering that hardly anybody, even students, is able to devote 100 percent of their time to research, it is more realistic to assume that some 100 hours a month, or some five hours per working day, are dedicated to research. That means one interview requires a processing time of about one-and-a-half weeks; two can be done in three weeks; and 20 interviews in 30 weeks. Taking away holidays and a few days of sickness leave, conducting and processing 20 to 30 interviews comes close to a full working year of one person. Additional time should be reserved in case the interview topics or the respondents are difficult, and several interviews are likely to fail. At least in non-profit settings, the assumption of compensating additional work with overtime should not be the basis of a professional calculation of resources. Funding for some of the more mechanical tasks, like transcribing or editing, to be outsourced can certainly make things easier. We advise students and research novices to be rather generous in planning their time as it always turns out to be more work than envisaged. Besides, as Marshall and Rossman (1995: 136) point out, ‘not all critical events Page 31 of 32

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

can be anticipated’ and ‘planning for more time than initially appears necessary is prudent’. For instance, despite transnational policy attempts in standardising curricula into the usual BA and MA structure, student life is not all that predictable: the ‘once in a lifetime’ chance of travelling to an exotic country, the excitement around starting or ending a relationship, depression, or bereavement are just a few examples of how time schedules can be mixed up. And usually there is nobody else to take over. Furthermore, the practice of qualitative research itself can hold time-consuming surprises and it is advisable to consider in advance ‘shortcut strategies in the methodological procedures’, as Flick (2007: 52) calls them. For instance, the use of (cultural) peers as interviewers could be such a methodologically sound strategy (see sections 3.4 and 6.2). Box 5.3 informs about how the problem of time was solved in STUDY C by means of selecting anchor cases for detailed analysis. After all, at least in the course of planning the practical aspects of research, one can of course still modify – i.e. specify and focus – the overall research question! http://dx.doi.org/10.4135/9781446288030.n5

Page 32 of 32

The Problem-Centred Interview: Principles and Practice

Examples of Working with PCIs In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 123-173 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Examples of Working with PCIs The previous chapters provided an introduction to the basic principles and steps of problem-centred interviewing, illustrating them with a few practical issues. This chapter is now dedicated to providing more extensive examples: it describes and reflects the use of the PCI on the basis of STUDIES A and B (see Table

6.1 and Box 1.2 for an overview). We first describe key features (e.g. research interest and questions, sampling, etc.), discuss the rationale of choosing the PCI, and finally focus on selected aspects of interviewing. The presentation of the first and most complex study, STUDY A, explicates the application of the PCI in the context of a funded perennial mixedmethods project combining quantitative and qualitative analyses. We illustrate how various layers of prior knowledge merge into a sensitising framework, and how it is represented in the interview guide. We reflect aspects of interviewer training and the organisation of access to the respondents. The detailed analysis of the typical opening question is followed by a discussion of two variations of the opening account and their implications for follow-up questions. In order to avoid redundancies, the second example of STUDY B focuses on complementary features related to the use of PCIs as expert interviews in the context of a commissioned peer research project. We introduce background, research questions, sampling, and interviewer training, and discuss the special situation of judges (as co-experts) interviewing judges (as experts), as well as the important aspect of their particular (prior) knowledge. Finally, we discuss the advantages of applying a consistent grid of questions. Typical pitfalls and interviewing errors are integrated into the representation and are discussed only now because an understanding of them requires some familiarity with the basic concepts discussed in the previous chapters.

6.1 Study A – ‘Transitions to Employment’ Research Interest and Approach The interdisciplinary project ‘Transitions to Employment’ (cf. Heinz et al., 1998; Kühn, 2004; Kühn and Witzel, 2000; Schaeper et al., 2000, 2001; Zinn, 2001) was one of the sub-studies of the Collaborative Research Centre 186 ‘Status Passages and Social Risks in the Life Course’ funded by the German Science Foundation from 1989 to 2001 (see: http://www.sfb186.uni-bremen.de/frames/main.htm). The study focused on the transitions of young people from the German dual system of vocational education and training (VET) to the occupation-centred labour market, and early employment careers of young professionals. The focus on the transitions between different public and private statuses (e.g. forms of employment, unemployment, family, education and training, etc.) with a prospective longitudinal approach (Elliott et al., Page 2 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

2008; Witzel, 2010) had its background in the social science discourse of the time. The guiding thesis of individualisation assumed the erosion of standard forms of employment and of the normal life course organised around gainful employment. In order to investigate this dynamic phenomenon, the professional biographies of members of one cohort of VET graduates were studied on the basis of several quantitative and qualitative waves of data collection. The research addressed: occupational risks and opportunities in two regional labour markets; patterns of transitions into vocational training and employment; and the initial years of employment. Additionally, the project analysed behaviour within structures on the basis of the action strategies of young adults realising their transitions into employment and their career plans as well as outcomes. The mixed-method approach combining quantitative and qualitative data allowed analysis and quantification of the structural impact of, for example, gender, occupation, educational attainment and region on transition patterns and pathways, while also considering individual aspirations, intentions and perceptions of structural features. Following Giddens's (1984) suggestion to overcome the opposition of structure and agency, the study assumed purposive actors: their knowledge is bounded and their action constrained, but not determined, by socio-cultural factors. Subjectivity was considered an important mediator between social structures, social institutions and individual action. At the same time, life-course patterns and their underlying active and purposeful biographical construction were additionally observed by means of quantitative methods. The choice of the PCI as the method for collecting biographical data was due to the focus of the qualitative analysis on the respondents’ own versions of their transition histories and the ways they managed to negotiate their early employment careers. The analysis of life-course dynamics and biographies with PCIs differs from other approaches which base their interest in the ‘narrative biography’ (e.g. Smith, 1998) of their respondents on collections of personal letters and other documents. Instead, PCI-based biographical analysis is a more direct, systematic and context-sensitive exploration of a person's life history. The interview conversation is an immediate form of confronting the research challenge of collecting the whole story which, in this case, was about the causes and consequences of the pathways of youths going from secondary education straight into the employment system.

Sample The mixed-methods approach was able to combine the substantive interest of the study (‘What’) informed by quantitative findings with their specification and explanation (‘How’ and ‘Why’) in the perspective of the respondents from the qualitative analysis. In order to bring these two perspectives together, the qualitative sample was drawn from the quantitative sample after preliminary data analysis (see Figure 6.1).

Page 3 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 6.1 Sampling plan (STUDY A)

The four waves of quantitative sampling (between 1989 and 1997) followed two groups: the complete cohort of young adults that received apprenticeship certificates in one of six major occupations in the city state of Bremen in 1989; and one-third of apprenticeship graduates in the same occupations and in the same year in Munich, the capital of Bavaria. Bremen represented a depressed labour market whereas Munich's labour market was booming. The six occupations were among the most popular: bank clerk, office clerk, engine fitter, car mechanic, hairdresser and sales clerk. These occupations represented a range of employment opportunities, with bank clerks being at the top and hairdressers at the bottom; and, reflecting the gendered structure of the apprenticeship system, they were male-dominated (engine fitter, car mechanic), femaledominated (office clerk, hairdresser) and mixed (bank and sales clerk). Orientations and decisions in the background of experiences and events collected with the questionnaire were specified with data from the qualitative sample. In three waves between 1990 and 1995 it was based on a theoretically-informed selection (cf. Strauss, 1987) of members of the first wave of the quantitative sample. In order to include all relevant groups, the qualitative sample considered the following criteria: the structure of opportunities (labour-market regions with good and bad prospects in these occupations); pathways and transitions from training to job (direct/immediate versus indirect/delayed entry to apprenticeship); and the prospects of staying in the training company. The additional consideration of ‘special cases’ with unusual biographies – for example, a young woman with post-secondary education training as a car mechanic – provided insights into the complexity of occupational choices. Conceptually, the purposive selection of interview respondents from the survey participants corresponds to the utilisation of individualised prior

knowledge in the PCIs. It enriched the chain of dialectic knowledge production about individual respondents, allowed confronting knowledge from the two sources of data, and altogether increased the validity of findings.

Prior

Knowledge

Page 4 of 50

and

Sensitising

Framework

Forms

of

Prior

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Knowledge Most of the interviewers were at an early stage of their career and thus biographically close to the problem of transition to employment. The diversity of their experiences constituted a specific everyday knowledge of the issue. It was the basis for the appropriation of other aspects of prior knowledge and needed to be reflected in the interviewer training (see below). With first interviews available, the analysis of interviewing errors included discussions of possible bias through everyday knowledge. The second form of prior knowledge regarded the context and institutional framework of the VET system and the institutional design and opportunities of transitions to employment. Context knowledge informed about how individual action was embedded within ‘objective’ opportunity structures; it included statistical data and consisted of information about, for instance: • the general situation in the six professions (content of training; possibilities for qualification, further education and advancement); • the organisation of the training (access, conditions, dropout rates); • various pathways to employment and forms of status changes (e.g. the option of being hired by training institutions; further education; change of profession; interruption of employment due to family formation; compulsory military or alternative service; unemployment); • the state of the labour market according to the economic situation and opportunity structures (job placement, internal labour markets, income situation and employment structure). Expert interviews with representatives of labour exchange offices complemented the prior context knowledge and provided insights into the informal practice of job search and job placement. Furthermore, in the frame of a pilot study, researchers and interviewers acquired insider knowledge beyond living conditions. For instance, in order to familiarise themselves with the language and culture of young adults they participated in extracurricular educational events.

Research knowledge based on reviews of specialist literature was produced on several occasions: in the context of applying for funding, preparing each round of data collection, and at the beginning of each funding period of the longitudinal study. Following the interdisciplinary approach, literature was reviewed from sociology, from social psychology, and from youth, vocational, biographical and life-course research. However, most of this literature turned out to be of little use. The reported findings were valid only for some stages of the research or for earlier historical periods; and they could often be interpreted only within a specific theoretical frame of reference. Furthermore, owing to the lack of longitudinal research, it was impossible to develop concepts for the identification of patterns of existing employment biographies during the preparation of the study. In the end, an approach that followed the principles of Grounded Theory seemed appropriate for this exploration of largely unknown territory. The longitudinal research process itself was an additional source of research knowledge. Interviewers exchanged experiences and interpretations of interview passages as well as practical solutions for specific Page 5 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

interviewing challenges. The longitudinal character of research meant that interpretations and findings from previous research stages were useful. As the study moved on, this cumulative aspect of deepening knowledge resulted in a gradual improvement of the thematic sensitivity in the interviews. Researchers needed to find ways of dealing with these three forms of prior knowledge. They made use of their different theoretical preferences, their substantive qualifications, and the prior knowledge generated both during the interviews (in the sense of pre-interpretations) and during preliminary data analysis. Yet, following the principle of openness, they also had to make sure that this knowledge did not narrow down the empirical work. They tried to establish a balance through: developing a self-critical attitude towards their preconceptions; interviewer training; and a culture of permanent exchange in the project and among associated researchers. The main research question concerning the professional biographies of young adults was addressed by the longitudinal study on three levels – it analysed sequences of events, related the conditions of the social context to individual action, and reconstructed individual life-course patterns across the manifold transitions based on retrospective, prospective and present-related accounts. In order to conceptualise these aspects, two sets of sensitising concepts were introduced: ‘pathways and transition patterns’, and ‘agency

within structure’. These concepts specified the direction of research by asking for empirical evidence on: how respondents plan, constitute and assimilate their biographies; how they interpret structural framework conditions either as restrictions or as accessible options; and how they draw on available resources in their action.

Figure 6.2 illustrates the dimensions of prior knowledge available in STUDY A. Starting with their own experiences it includes (moving clockwise): everyday, contextual and research knowledge. This establishes the sensitising prior knowledge of pathways and transition patterns and the overall approach of studying agency within structure.

Page 6 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 6.2 Prior knowledge (STUDY A)

Sensitising Framework – the ARB-Model The complexity of the research question required a conceptual framework that could instruct both data collection and interpretation. It should suggest a meaningful grid to organise the respondents’ statements that is at the same time open for further interpretation. The solution consisted in the introduction of a sensitising

framework regarding professional options based on the aspects of aspirations, realisations and balancing – the so-called ARB-model (Witzel, 2001). These categories were introduced to provide an alternative perspective on the reconstruction of biographical orientations and actions; following the principle of openness, the burden of established concepts from action theory was avoided. The distinction of these three dimensions served as a kind of sorting mechanism for orientations and actions at different stages of the biography

Page 7 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

articulated in the interview. Due to their interrelated character, these dimensions could be combined into and used as a sensitising framework for the analysis of professional options. In practice, these three dimensions were applied to various situations relevant in professional careers. Let us consider the expected first job after professional training: the respondent refers to this job as a professional option – i.e. it remains open what the next step will actually be, and whether the respondent will (be able to) realise this option. The motivation behind mentioning this option refers to the category of aspirations. Aspirations represent individual justifications for goals of action in the sense of interests, expectations and hopes for the future. They can be derived from everyday experience as well as theory and research. Yet the category of aspiration leaves open whether these expectations represent actual claims, deliberate professional plans, or rational decisions. It is the task of the analysis to find this out and to specify them empirically. In other words, the notion of aspirations is a general reference to an important aspect of the ‘problem’ of youth transitions to employment that is specified only as the study moves on. In conceptual terms, aspirations – as well as realisations and balancing – are sensitising concepts which we described as

empirically less contentful and open to being filled with substantive meaning by the respondents’ perspectives in the course of the study (see section 3.2). The second category of realisations addresses concrete career steps. They include, for instance, experiences in the labour market, expectations and selection strategies on the side of employees, the relevance of resources (education, social networks, etc.), and the assertion or revision of goals and ambitions. The category of balancing addresses consequences and results of action and how the individual evaluates them. The term refers to the phenomena of biog-raphisation and individualisation underlining the increasing importance of biographical reflexivity: modern biographies are normatively open to being constructed by individuals; biographical reflexivity is part of modern life styles (e.g. Beck, 1992; Giddens, 1991). The process of balancing searches for and attributes new meanings to completed actions. These meanings are the result of a process of examining the consequences of professional decisions and choices for their biographical sustainability. On this basis, the respondents revise some of their standards of evaluation and develop consequences of action for the future. Life-course stages establish a temporal relationship, i.e. they are interconnected by mechanisms of socialisation and selection throughout the biography. In terms of time perspective aspirations, for instance, represent a double horizon. First they are backwards oriented and connect the present situation with previous stages; they are essentially based on the reflection of experiences and the available history of a given situation. Yet they also refer to future life-course stages and biographical development. The collection of aspirations in interviews allows for the reconstruction of prospective and retrospective aspects, and how the individual makes sense of them. The heuristic ARB-model provides a sensitising framework that also reflects the dynamic longitudinal approach with its emphasis on the dimension of time. In the context of STUDY A, PCIs collected information about aspirations, realisations and balancing at different moments and allowed the reconstruction of how they were related to each other at these moments as well as over time.

Page 8 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Figure 6.3 illustrates the ARB-model on the basis of an exemplary sequence that follows the initial balancing (B) after the conclusion of the first life-course stage of dual apprenticeship training. The second stage of the first job involves all three elements of aspirations (A′), realisations (R′) and balancing (B′). These three aspects are again relevant in the third stage, which relates to the one before.

Figure 6.3 The ARB-model in the life course (STUDY A)

Interview Guide, Interviewer Training and Field Access Interview Guide In line with the sensitising framework, the main function of the interview guide was to prepare the direction

of the questions. Its structure reflects the thematic and temporal aspects of the research interest and the conceptual background. Owing to the longitudinal character of the study involving a variety of life stages, different interview guides had to be prepared for each wave of data collection. In practice, and following the guiding sensitising concept of ‘agency within structure’, contextual knowledge about each life stage was necessary to organise the interview guide regarding thematic and temporal aspects. Detailed knowledge about the peculiarities of the interplay of the VET-system and the labour market were critical here. In this way, similar issues could be explored for a variety of situations.

Box 6.1 includes the structure of the interview guide for the first wave and its emphasis on the educational biography. The extended interview guide can be found in Appendix 1.

Box 6.1: Main Structure of Interview Guide for Wave I (1989/1990) (Study A) 1 2

3

Page 9 of 50

Introduction - Opening Question Pre-Professional Situation 2.1 Occupational options 2.2 Search for Apprenticeship Training Apprenticeship training place 3.1 Transitory situation 3.2 If a job is offered 3.3 If a job is not offered 3.4 If there are graduates from measures/vocational school 3.5 Unemployment 3.6 Armed forces/alternative service The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

4 5

3.7 If a job is offered in another company Workplace Future situation

The already-accomplished life-course stages are addressed in points 2 to 4. Their order is prescribed by the logic of the VET-system. According to the sampling criterion of selecting graduates from vocational training, these points can be expected to become relevant at successive stages of the professional biography during the interview. Owing to the emphasis of this first wave of data collection on the educational biography, the life-course stages concerning professional choice (2) and training (3) were further differentiated. Point 5 invites the respondents to talk about their expectations concerning future life-course stages. It addresses perspectives that are currently available in the sense of stages that appear connectable to whatever situation the respondent is in. They can include: family formation; change of employer or a new position involving retraining; start of university studies or return to school; and preparing for new professional options. An additional way of structuring the interview guide consists in the incorporation of references to detailed

research questions. They are formulated as keywords or as full questions; both should recall specific research topics related to each life-course stage. For instance, with regard to the search for an apprenticeship training place the interviewer should explore, among others, the conditions in the training company, the reasons for successful application, and whether expectations were fulfilled. The respondent's comparison of these experiences to those of friends or acquaintances in similar situations further deepens the exploration. Research questions like these suggest the direction the conversation should take regarding the interview guide sub-topics. They can be reformulated and transformed into interview questions when applied in certain thematic contexts. For instance, this is the case when interviewers are overwhelmed by the wealth of detail and information. They may lose track of the general direction of the questions and the idea of problem centring altogether. Finally, the interview guide includes fully formulated sample questions that can be adapted and modified to everyday language according to the situation. They can be used to get the interview started and to introduce specific topics; or they address the specific fields under investigation. Below, we discuss the use of questions for opening the interview and for following up in more detail. The combination of detailed research questions and pre-formulated sample questions applied to a variety of situations should first of all facilitate orientation, memory, completeness and comparability. The order of life-course stages suggested in the interview guide is applied in a flexible way and according to the respondent (see below for details).

Interviewer Training The study about youth transitions to employment involved a number of interviewers over an extended period of time. Thus, the general suggestion for PCIs that the researcher should also do all the interviews (i.e. Page 10 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

researcher/interviewer) was not feasible, and interviewers had to be trained. The training prepared them for their double role during the conversation: first, as well-informed traveller/ expert on the issue, they had to apply the prior knowledge accumulated before and during the data collection (about the context, and from research); and second, as learners they had to be open to the subjective life-world of the respondents in order to revise this prior knowledge. The interviewer training was typically organised in three steps, as follows. (1) At the beginning, the interviewers learned about the background of the study and the main research

questions, and grew familiar with the prior knowledge represented in the interview guide. This introduced them to the research task and should have enabled them to gain additional insights, extending and revising available knowledge. It also established a common understanding of the particular problem centring of the interviews shared by the research team. One of the challenges resulted from the permanent overload of interviewers and research assistants during labour-intensive periods of data collection. The first wave of the study included some 200 interviews with an average duration of about 1.5 hours in the north and south of Germany. At the end of the 1980s, no cheap airlines or fast trains were connecting major European cities, and the distance of more than 700 kilometres between Bremen and Munich was a real obstacle. Owing to the longitudinal character of the study, the usual work of data management and analysis – i.e. transcribing and interpreting interviews and postscripts; producing case summaries, etc. – was complemented by the necessary preparation of the next round of interviewing. In practice, this meant that, in order to prepare the second round of data collection, comparative

but individualised follow-up questions for these 200 respondents needed to be formulated. Notes about earlier steps in the professional development and related future perspectives provided additional background knowledge. They helped to understand the overall context of the professional biography and helped focus the questions for the following interviews. This preparation had to be done without knowing who or how many of the respondents would eventually drop out of the process – and about 40 percent did. The related preparatory work was essentially lost. Additional interviewers hired in the process were introduced to the role as researcher/interviewer and equipped with the relevant prior knowledge and interviewing skills. In line with the German university tradition of involving students in research, the project recruited and trained students that would then participate in the study over a longer period. They worked on the basis of internships, contracts or as student assistants; they participated in discussions and team meetings and were fully involved in contributing to the production of results. Some of them wrote their thesis about specialised research questions associated with the project and were able to use the collected interview material for their own work. (2) A second purpose of the interviewer training was the introduction of the participants to the principles of

qualitative-interpretive research. In particular, the positive role of prior knowledge as sensitising knowledge for problem-centred interviewing was emphasised. It was crucial for the interviewers to understand the importance of a sensitising orientation as the precondition for productive and valid interviewing. The interviewers had to develop a self-critical orientation towards all sorts of knowledge that can enter the research process. Page 11 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The participants were sensitised for blind spots they may have developed during their own academic socialisation and which prevented them from assuming the perspective of members of other social strata. In theory, students of social sciences are aware of cultural differences between social groups. The interviewer training contrasted this by discussing interview examples of culturally grounded misunderstandings between researchers/interviewers and respondents. The reflection of elitist attitudes of the well-educated, intellectual interviewers towards the respondents is one example (see Pitfall 6.1).

Pitfall 6.1: Elitist Prejudices and Their Reflection in the Interviewer Training It was common that interviews with hairdressers, which had the lowest level of education, were at first disregarded in the research and considered irrelevant because of their apparently ‘simple’ character. The real value of these interviews for the study of life-course issues was understood only at a later stage of interpretation. The training material included examples from the interviews with hairdressers concerning conflicts with superiors. For instance, one of the hairdressers is confronted with the suspicion of having feigned a sick call, and she defends herself against being implicitly accused of lacking work ethics. The description of the conflict with her boss takes the form of a dialogue reproduced in direct speech. The contributions are consistently introduced by the respondent in the form of a simulated dialogue alternating ‘I said …’ and ‘he said …’. Additionally, she recounts a similar conflict between a colleague and her boss which followed the same pattern. The participants of the interviewer training were amused about this apparently unsophisticated way of talking and constructing a dialogue, and considered it naive. Furthermore, they were unable to understand that this example represented more than just the articulation of discontent, and that it could actually contribute to answering the research question. Some suggested interrupting this apparently irrelevant narration and to move on to another topic instead. Through the discussion of the examples, the participants understood that narrative accounts always include subjectively relevant meanings. Especially with an attitude of intellectual superiority, these meanings are easily overlooked even if the frequency of such episodes alone indicates their significance. Interviewers with a culturally different, perhaps elitist set of expectations towards a narrative may not be able to penetrate the surface of the conversation and understand the underlying meaning. From their intellectual point of view, they may find that the exemplary accounts lack the ‘usual’ ingredients of storytelling such as logical conclusions or final appraisals. Against this background of cultural misunderstandings, the participants were asked to develop an impartial approach to the transcripts. Regarding the example, they were invited to discover the subjective meaning behind the hairdresser's description of the unjust

Page 12 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

treatment by her superior. The moral indignation was obvious in the interview passage, and so was the urgency with which these complaints were articulated. Finally, the fact that similar conflicts of other colleagues with their boss were recounted in the same way made the participants understand the meaning behind the pattern: in this way, the otherwise only implicit negative appraisal about the superior could be doubled and served as proof that the complaint about the boss was actually justified. Once cultural prejudices were overcome, participants were asked to search for the reasons behind this particular pattern of expressing complaints. It became obvious from a second look into the example that these complaints referred to shortcomings of interpersonal relations in general, and not only between a superior and his employees. The interpretation moved from initial doubts about the relevance of the interview passage to uncovering the generally negative work climate at the hairdressers. Yet despite their complaints, the hairdressers did not change job or employer; they accepted their professional ‘dead-end’ situation. In view of the problematic working conditions and the low income prospects, they aspired to a feeling of belonging to the company, to a good work climate and to a caring and attentive attitude from their superiors. Through careful interpretation of this example, the participants became aware of their cultural bias. They finally understood that the complaints conveyed, in a negative way, the significance of the ideal of a cosy workplace for coping with precarious professional biographies. Other hairdressers that were at the end of their professional career were able to consider this ideal as realised and could thus call the workplace their ‘home’ or paraphrase it with words like: ‘We are like a family.’ The problem of social desirability in interviews was another important aspect of the training here related to the interview topic of professional career patterns. Some respondents wanted to conform to supposed expectations concerning career strategies. They obviously felt that in their self-presentation they should emphasise their commitment to professional advancement by, for instance, mentioning their readiness for further training. The project tried to deal with the problem of social desirability in various ways. First, the issue was openly addressed in the context of the first contact to the respondent. The researcher emphasised the focus on individual actions, experiences and opinions and not on the evaluation of individual achievements. The positive atmosphere of the conversation was a second means of reducing social pressure. The accepting attitude of the interviewer should invite the respondent to be authentic. Explicit evaluation may bias the accounts and should be avoided (see Pitfall 6.2). Finally, the longitudinal character of the project allowed discovering differences between intended and realised actions. For instance, in the second or third round of interviewing it was possible to find out whether (or which) further education plans articulated in the first interview were realised.

Page 13 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Pitfall

6.2:

Positive

Judgemental

Comments

and

Their

Negative

Consequences One opportunistic reason for interviewers to articulate agreement consists in the attempt to establish and maintain a good atmosphere and relationship with the respondent (see the role of the accomplice in the expert interview, Bogner and Menz, 2009). The complete loss of distance from the respondent, also known as the phenomenon of ‘going native’ in ethnographic research, is another reason. For instance, interviewers may enjoy articulating their agreement with opinions of the respondents; or, perhaps out of solidarity, they may want to communicate their sympathy regarding problems or achievements and failures of the respondents in a way that goes beyond interpretive reconstructions. Unsuccessful interviews are thus often characterised by exclusively positive valuations. In principle, judgemental comments imply that interviewer and respondent reverse roles: contrary to the actual purpose of research, the interviewer starts to act as the subject of the research in the sense of the carrier of orientations, opinions and values. As people tend to orient their description and performance towards positive responses and avoid negative feedback, appraisals also tempt respondents into providing only socially desirable answers. Reinforcing contributions from the interviewer establish a commitment from both participants to a specific interpretation, and it therefore becomes harder for the respondent to reconsider or modify his opinions. Respondents interpret interviewer assurances as invitations to conclude an issue prematurely, or they continue their apparently ‘successful’ account by underlining them with new and redundant examples. The following is an example from the interview of Emil Bank:

A: … Well, one can say, really, they (i.e. the boss and his employees; authors) made an effort to train us. I: Huh, these are really good conditions. A: Yes, so the conditions, that I found, in the first year of apprenticeship (were) perhaps not ideal, but in the second year of apprenticeship, where we went through the subjects that I was very interested in like loan and construction financing, it was quite unique. And where I was, in the loan department, it was so that we did not only study during working hours but we also came together in the afternoon, well late afternoon and discussed. And I mean, when you are together like that, then you get on to the topic of the bank again. And also then we

Page 14 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

continued, I mean we simply talked about it. And anytime when I have a question now, I ask him. The interviewer's reaction to Emil Bank's few words about professional education expressing his positive opinion about the relationship to the instructors and the employer is strongly affirmative. At first glance, this seems unproblematic: the relationship between interviewer and respondent is positive; and in terms of interviewing, the interviewer's comment does not seem to block or interrupt the account. It seems the approval of the respondent's original appraisal (‘made an effort’) by the interviewer (‘really good conditions’) encourages Emil to continue with an even longer narration that first of all further supports his opinion. Yet Emil's attachment to this positive perspective is not undifferentiated: he reduces his perception of ideal conditions of education to the second year of apprenticeship, and, in so doing, partly contradicts the interviewer's sweeping judgement. This can be explained with the fact that the interview is in an advanced stage, and that in his efforts of reconstruction, Emil was able to emancipate himself somewhat from the interviewer. (This is, by the way, one of the main arguments in favour of the robustness of the technique of the PCI.) However, the risk of socially desirable answers is not averted altogether. And in the perspective of Emil, the expectation also of negative assessments by the interviewer persists. In the course of the interview, the interviewer makes several affirmative valuations; one could even consider it part of his interviewing style. For instance, later on, the interviewer comments on achievements regarding job applications by saying: ‘Yes, well this also went very smoothly in your case …’; or, a desired change to a different department of the bank is commented on with the words: ‘That's great …’ The consequences of these reinforcements are far-reaching and also affect the interpretation of the interview: Emil's optimistic and positive view of the period of transition from education to work appears strikingly consistent. Yet it is difficult to say whether Emil's overall assessment, articulated later on - ‘I mean, it is fun to do the job of a bank clerk as such, be it at the cash register or in the area of bonds or credits, actually pretty much everywhere …’ - reflects his real attitude or is rather the product of social desirability and positive reinforcement throughout the interview. (3) Finally, the participants were trained to use the PCI's combination of active and appropriate orientation towards the problem by conducting the interview in an attitude of relaxed attentiveness. They were systematically introduced to basic interviewing techniques and possible mistakes on the basis of examples, and sensitised to the process of mutual interpretation that constitutes a dialogue. They learned when and how to use communication strategies that focus on generating either narratives or comprehension. And they were Page 15 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

trained to use the sample questions included in the interview guide and to distinguish research from interview questions. Additionally, and depending on their needs, they were trained to deal with problematic situations on the basis of hypothetical assumptions (‘What to do if … ?’). The interviewing workshops, which were not only offered at the beginning of the data collection but also throughout the process, followed the principle of reproductive learning: the participants were asked to assume the role of interviewers and invited to simulate conversation. All team members participated in the interview analysis which included the reflection of interviewer interventions and how they affected the accounts of the respondents. A self-critical attitude, a positive atmosphere and a climate of solidarity in the research team facilitated the evaluation and discussion of individual interviews and interviewer mistakes. The examples used for the interviewer training originated from a pilot study on the problem of youth transitions from school to work. The longitudinal project itself also generated many positive and negative examples that could be used whenever new interviewers had to be trained, and the first attempts at interviewing by new project members carrying out pilot interviews were also evaluated.

Access, Interview Settings and Introductory Explanation The establishment of sustainable research contacts is an often under-estimated part of the process. Usually, in order to achieve rich and valid dialogues about issues that are relevant for both sides, the different perspectives, interests and backgrounds of researcher/interviewer and respondent have to be brought together. For the interviewer, the first contact is at the same time the beginning of a journey into the life-world of the respondent. Careful negotiations are important at this stage because they can help to eliminate concerns and resentments that respondents may have; they are a valuable investment facilitating the prevention of short, evasive and socially desirable answers. Considerate access to the field also promotes participation in the study. It is necessary to avoid bias in the sample, which is anyway declining rather fast in the course of a longitudinal approach involving differentiated sample criteria. In the context of the mixed-methods study on youth transitions to employment, the first direct contact to the respondents was made during the survey that preceded the qualitative interviews. The accompanying letter indicated that additional interviews with selected people on a voluntary basis would start six months later. In case of agreement, the respondents were asked to provide their contact details, which were then stored separately from the questionnaires. Because of the expected job mobility of the young people during the course of the longitudinal project, the addresses of their parents were also collected during the first qualitative interview. Their initial agreement to participate in the survey underlined their readiness to become part of the study, indicating a general interest that could be built upon. Being part of an intensive face-to-face interview, however, is a completely different situation. Therefore, the starting point of all efforts to involve respondents was the assumption that it was unclear how they would perceive the academic world and what they would know about it in their perspective of everyday knowledge. Would young people who probably have no idea about social sciences be trustful or rather sceptical? And how should we deal with this uncertainty?

Page 16 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

The clarification of the problem-centred character of the conversation was important at this stage. It included an explanation of the purpose and content of the interview, as well as the general circumstances of the research. Unlike typical psychological experiments, the purpose of the research was not hidden. Quite the contrary, the explanation and ‘translation’ of the basic research goals in everyday language should enable them to understand the background of the study, their own role as experts, and expectations associated with their participation. In this way, the problem centring was twofold: it became clear that the topic was limited for scientific purposes, and that it concerned specific and subjectively relevant issues of the everyday or working life of the respondents. First of all, the respondents were addressed in plain language and as experts for the various routes to a profession. The interviewers explained that they were interested in the respondents’ expectations, hopes and disappointments, and highlighted their eagerness to learn about the story behind their current professional situation. Already at this point, the interviewers made use of formulations similar to the opening question of the actual PCI. They asked for permission to record the interview and explained that it would then be transcribed and analysed by the research team. The comparison with other interviews would produce a diverse picture of various pathways to working life and related experiences. The respondents were reassured that their contribution was relevant and that they actually had something to say with regard to the issue. Second, during the initial contact, the interviewer had to gain the respondent's trust. He introduced himself, for instance, as a student at the university that conducts the project in which the respondent already participated. He assured confidentiality and underlined that the contents of the interview would not be distributed, either in his private or in his professional sphere. He also mentioned precautions regarding the anonymity of the material and offered to say more about it if required (especially concerning the safe storage of data and references lists for secondary analysis and follow-up interviews in the context of longitudinal research). Names and other sensitive details (like names of companies, schools, friends, locations, etc.) would be unrecognisable. Once established, the relationship of mutual trust between interviewer and respondent was cultivated, and if possible, the same person contacted the respondent for follow-up interviews in the context of the longitudinal project. Third, the interviewer introduced the PCI as methodical alternative to the questionnaire, which is limited in terms of possibilities to collect detailed experiences of the professional biography. Instead, a conversation would provide exactly the opportunity to elaborate on certain experiences. They would be free in their responses and could take time to develop answers. The interviewer also indicated that he would support the reconstruction of experiences with questions. This was necessary because, without previous experience, the invitation for storytelling was often perceived as pressure to perform. The respondent should rather perceive the interview as a chance to present extensively his experiences and opinions as an expert of his own life course. And, on the basis of this introduction, he should understand how he could and should comment on

the topics of the interview. The experience with many respondents in the course of the longitudinal project indicates that there were Page 17 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

several other aspects influencing the willingness to participate. For instance, respondents who had experienced failures in their professional biography were less likely to participate. After they were identified on the basis of the survey, the interviewer told them during the first phone contact that problems in transitions to working life were common and often depended on the labour market situation. The interviewers underlined that the study was exactly about identifying differences that young people encountered while trying to enter the world of work – success or achievement were not important, but the readiness to share one's experiences were. A small amount of money was offered as an additional incentive. It was a form of recognition of the respondent's participation and compensation for possible travel costs. Obviously, monetary incentives have the advantage of immediately representing some sort of exchange value and of being considered ‘useful’, and the young people certainly appreciated it. The location of the interview was another important aspect of negotiating agreement. The question was, whether it was advisable to magnify the intensity of face-to-face interviews by doing them in a private setting. We assumed that young people had different preferences and offered to do the interviews either in the neutral office location at the university, or at the respondent's home (see Box 6.3). Details of agreed appointments were typically registered in a telephone note (see Box 6.2). Time and place were confirmed in writing.

Box 6.2: Telephone Note of First Contact and Appointment Introduction: name, University of Bremen Introduction of project Practicalities (duration of interview, etc.) Agreement on time and place Announcement of confirmation letter for appointment IP-number: ……….. Interviewer: ………. Date: ………..

Interview appointment Date: ……….. Location: ………..

Further information • e.g. new address, tel.-nr., times of absence from/to

Page 18 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

• military/alternative service from/to Reasons for failure of interview The longitudinal character of the study required the fairly long-term commitment of the respondents. Possible conflicts with other life domains like work or leisure were unavoidable but could hardly be anticipated. During the first two waves of the study, only a few decided to discontinue their participation and were not available for further interviewing. This was interpreted as a success of the project - making the interviews also relevant for the respondents: they were accompanied in their reflection of experiences of success or failure and, through talking about it, could arrive at a profound understanding of their biographies in times of transition. However, as the respondents settled into their jobs and developed life routines, the interviewers had to improve their efforts to persuade them to continue with their participation. For instance, they had to underline the fact that the third round was the last. Furthermore, it was part of the maintenance of the panel to keep permanent contact with the respondents at an appropriate level of intensity and frequency. On a regular basis, leaflets and flyers updated the respondents on the progress of the project. In the end, only a few respondents refused to continue and dropped out of the study - most of them were among the hairdressers in Munich, who were in altogether precarious working and living situations - their earnings were low in a city with high living costs, and possibilities for professional advancement were lacking. Many were unsuccessful in trying to find alternative work in jobs that required little qualification. ‘No time’ or ‘I don't feel like it’ were typical explanations for refusing to participate.

Box 6.3: Impressions from Doing the Interviews Few respondents wanted to be interviewed at the university; the ones that did were perhaps simply curious about the unfamiliar university life. Usually, it was more convenient to do the interview at the respondent's home, which was most commonly their parents’ flat. When a respondent had their own flat, their partner was usually around, at least for part of the time. An element of (gendered) control was obvious from this constellation: parents and partners often assumed a sceptical role towards the interviewer. Young women appeared to be under special protection. Some parents, especially mothers, were concerned their daughters could become problematic cases and often complemented the child's story with their own remarks. Partners, on the other hand, considered answering certain questions a joint challenge; they contributed their own work-related perspectives and talked about common agreements with regard to life planning. Typically, interviews with young women were done in the living room, which was shared

Page 19 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

with other family members, and where simple things like the shortage of free sockets could cause technical problems with regard to recording. To describe a common scenario: the interviewer sat on a chair opposite the respondent, a young office clerk maybe, and in front of a well-laid coffee table. Her father would sort stamps in another corner of the living room; her mother would be busy in the kitchen and occasionally shout through the open door that the daughter should not forget about her latest job application. Interviews with young men followed an equally standardised pattern. Either the interviewer was left alone with him in the living room, or undertook the interview in the son's own room in the parents’ flat. Mothers would occasionally knock on the door to offer coffee. In this way, the parents contributed to documenting the son's independence regarding his professional career. Observations like these were an important part of the interview and registered in the postscripts together with other peculiarities of the conversation. Additionally, aspects of warming up before the actual interview were noted; they captured the process of how interviewer and respondent grew familiar with each other, or how issues already discussed during the first telephone contact were specified. At this stage, doubts and concerns, often related to matters of confidentiality and anonymity, could be addressed. In order to prepare respondents once again for the topic, the interviewers repeated the purpose of the study, emphasising that respondents were free to articulate their opinions and comments on the issues.

Interviewing: Opening Question In its short version, the opening question of the first interview (in Wave I) was typically formulated in the following way: ‘You did learn the profession of … (e.g. hairdresser). How did this actually come about?

Please, tell me about it.’ The very first sentence follows the discussions about the research purpose and summarises it. The sentence translates the problem centring of the interviewer – i.e. the interest in biographical work in the sense of acting and orienting in the context of the transition to employment – into the comprehensible everyday language of the respondent. This formulation is an ideal-typical example and a tried version of the opening question that was adapted depending on the concrete situation. Shorter or more colloquial versions could be appropriate and are not necessarily ‘wrong’ (see Table 5.2). For instance, more casual formulations were often used in case the relationship between interviewer and respondent was already quite close and informal (e.g. because of previous contacts or earlier interviews). The first sentence – ‘You did learn the profession of … (e.g. hairdresser)’, is not only the repetition of a

Page 20 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

banal detail that is evident from the sample and questionnaire; it also has the function of reminding the respondent of the topic, which was introduced as relevant and as something he could comment upon. It also addresses the active role of the respondent in the process of professional orientation: she did embrace the option of becoming a hairdresser and went through the training programme. Furthermore this version keeps

the current status open. It would have been wrong to say: ‘You are a hairdresser now.’ This would have carried certain implications. At this point, prior knowledge from the survey and from available statistics about transitions to employment becomes important. It shows that six months after the completion of vocational training, a variety of life-course statuses were possible, including unemployment, work in jobs without skills requirements, additional training in a different profession, etc. The open way of formulating the first sentence leaves room for the respondent to actually specify the current status. The second sentence – ‘How did this actually come about?’ – is a narration-generating question that suggests a narrative frame for the response. It reflects the process orientation of the interview and the knowledge from a pilot study of pre-professional socialisation – i.e. the ways of pursuing different professional options may be complex. For instance, the eventual training as hairdresser usually has a history, which should be recounted in detail by the respondent in a practical perspective. In general, it is an advantage to formulate questions towards the description of concrete actions and events because this usually also involves appraisals and emotions. In everyday communication, experiences are hardly ever presented as mere chains of events; their presentation is usually accompanied by appraisals, expectations and balancing, or explanations of goals. Therefore, the orientation of questions to the level of action and events makes it easier for respondents to ‘get into’ the topic. Through recounting past experiences, step by step, they will also reconstruct their subjective

meaning. Biographical narrations and descriptions of this kind are closely related to existential questions of reproduction, self-realisation and identity formation as well as, more concretely, one's leeway regarding work, family and leisure. The respondent has to deal with these experiences according to circumstances like the labour market situation and according to his own resources such as education or family support. In methodological terms, the assumption of the existence of a subjective meaning of events and actions is part of the prior knowledge of the project. It is a form of general social science research knowledge about how people relate to their experiences. Yet the presentation of this subjective meaning is not demanded explicitly from respondents in the interview. Instead, the open formulation of the question should allow them to provide subjective appraisals in a more authentic way. Think of an alternative formulation like, for instance: ‘Does the

vocational training come up to your expectations?’ Such an explicit suggestion would necessarily provoke social desirability and narrow down possible answers. Respondents are likely to answer in the affirmative because they feel prompted to present the outcomes of their biographical choices as accomplishments, or at least as successful ways of coming to terms with difficult circumstances. They tend to assume that problems concerning their way of life cannot be addressed openly, but that efforts of successful coping are socially desirable. Instead, the open question – ‘How did this actually come about?’ – helps to take this pressure out of the interview from the very beginning. It should be left to the respondents to present themselves as active subjects or as victims of circumstances, or both.

Page 21 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Finally, the invitation at the end of the opening question – ‘Please, tell me about it’ – also has a distinct

methodical function. It complements the narrative frame introduced before with a short reminder of the importance of storytelling: ideally, the communicative form of the narration should be used. The emphasis on the unfolding of a story encourages the respondent to establish his own logic of representation on the basis of memory and priorities. The actual account then depends on the articulateness and readiness to support the efforts of the interviewer to comprehend the description of orientations and actions.

Opening Account The first sequence that follows the opening question is particularly important for the interview: it ratifies (or not) the ‘communication contract’ between interviewer and respondent and largely predetermines the possibilities for what narrative researchers call ‘immanent follow-up questions’ (see section 4.6). Its length immediately indicates the involvement of the respondent and the quality of both the opening question and the preparation for the conversation. In the present study, the length of opening accounts ranged from one to more than 50 lines of text. However, the length does not necessarily indicate how much respondents could actually talk about their experiences, for instance, of more or less complex/difficult transitions, or specific personal problems (e.g. health, detachment from the parental home, etc.). And there are many reasons for short answers at the beginning (see Pitfall 6.3).

Pitfall 6.3: (Too)-Short Opening Accounts The opening account of Heide Office - ‘Heide’ is the given first name, and ‘Office’ her job environment - is very short but clearly refers to a complicated story of professional orientation in her background.

H: I studied electrical engineering before and wanted to do something completely different, because I did not enjoy it anymore. And then I actually wanted to learn industrial clerk, but did not get a training place and changed to office clerk. It is puzzling that this representation hardly suggests any qualitative differences between the occupations mentioned although they entail very different chances and risks in the labour market. There could be other reasons why, for instance, this opening account is so short. Let's look at some of them.

First, the respondent might be uncertain about the consequences of the opening question. The new role as research subject is as unusual as the fact that, instead of concise answers, she is asked to tell as much as she wants and without restrictions in terms of scope and detail. Insufficient mutual trust might be a second reason that can result

Page 22 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

from uncertainty about the exposure of delicate information. For instance, Roswitha Office wants to make sure about this and concludes her opening account by saying: ‘… I can talk about all this, right?’ Opening questions that neutralise the narrative stimulus may be a third reason for restricted talkativeness at the beginning of an interview. For instance:

I: How did this actually come about, when did you start thinking about what you wanted to become later? R: Well, at grade 9, at the application or so. The narrative question (‘how…’) was extended by a non-narrative part (‘when …’). The process orientation of the openly formulated first part is undone by the succeeding detailed question about a specific moment. Unsurprisingly, a single-line answer followed. This is a common pitfall that can be avoided through careful training. A fourth reason may be that the transition was linear and unproblematic and extensive self-reflection not necessary: regarding the ARB-model, neither job aspirations, nor the use of opportunity structures for the realisation of options or the balancing of employment experiences, caused concern. The following example is able to illustrate such a linear story; it is interesting also for other aspects of interviewing. Luke Hairdresser is a case of professional heredity - circumstances, decisions, experiences and appraisals all appear unproblematic and have a taken-for-granted character.

L: It came about because we have our own shop. Yes, I simply don't know differently. We've had it now for 12 years, the shop. In former times I was often there, assisted a bit after school. Yes, and I just liked it, that's why I came up with this. In Luke's view, there is not much to tell because it was clear from the beginning that he would take over the parental barbershop. Once this decision was made, aspects of the transition to employment were largely pre-defined, e.g. his professional education, his transition into the training company, and his future income depending on his parents. Another passage from this interview underlines the apparent simplicity of the events that is reflected also in the predominance of short answers: the small business of his parents put him on track. Later probing questions hardly produce any new insights in terms of alternatives in his professional orientation.

L: Yes OK, eventually, in case we had not had the shop, maybe I'd have done

Page 23 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

something else, OK. I'd have done something with guns, electronics. I: But that … Did you ever follow this up, this electronic stuff? L: No, actually not. Not at all, no. The interviewer does not give in and, with little success, switches to a more dialogic form of conversation and connects his questions to details of the accounts, including single terms (‘electronic’). Then he follows a hunch that the young hairdresser may not have been free to choose his profession but exposed to pressure from his parental home. As the conversation unfolds he comes back to this suspicion, suggests alternatives and seems puzzled about the extent to which the respondent takes his decision to become hairdresser for granted. At some point, as the next passage indicates, probably because he thinks that the relationship with the respondent is stable enough, the interviewer even uses the communication strategy of confrontation to probe the coherence of the answers - ‘I mean, somehow usually one does not do anything without a reason …’. All the accounts that follow support the interpretation of a linear and unproblematic transition to employment; and also the level of detail hardly improved:

I: On your own, hm. And there you … how did you decide this, to say, ‘I don't feel like this, but hairdresser, that's for me’? L: (laughs) Yes, I actually also don't know, I cannot even answer this directly. I: I mean, usually one doesn't do anything without a reason, well somehow, some ideas … L: Ah well, the idea is that I want to take over the shop, that I want to get the master's certificate, and want to have a good future in this way. In the context of this study, two variations of the first account were typical: the first is organised chronologically, the second dramatically. We will take a look at them on the basis of empirical examples to clarify the following details: • What kind of information does the opening account include, and does it provide a first overview of the history of the respondent's transition to employment? • What possible starting points for follow-up does it include? • How does the interviewer deal with additional ideas for follow-up emerging from the opening account?

Page 24 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Variation I: Chronologic Organisation of the Account The opening account of Anton Mechanic is a typical description of the structure of the history of a professional biography that is chronologically organised. The topics he addresses refer to life-course stages (temporal structure) as well as to aspirations and balancing regarding the realisation of different professional options (thematic structure).

A: Well, first it was … first I learned salesman for two years (Stage 1). I saw that this just does not fulfil me. You stand there like a fool in the shop until half past six, and first of all your income is small, you don't do what you expect, giving advice, but you just take the shelf out, put a number on it, cash desk (Balancing 1). And then, already before I used to tinker around a lot with mopeds, and then did the driving licence and immediately had a car. And then I understood, that you, if you don't know anything or simply don't know what you have to know, that you have to take the thing to the garage. That you just have to learn (Aspiration Stage 2). And then I started to do casual jobs at a tyre dealer (Stage 2). He always also repaired cars on the side, and these cars, they were kind of fun for me (Balancing Stage 2). And then I just … and he also said to me, the best thing if you want to continue here: to learn (Aspiration Stage 3). And then I just started the apprenticeship one year later (Stage 3). Well, at a company that is actually lousy but now I have the certificate and that is just enough for me for now (Balancing Stage 3). Now, well now the Armed Forces, for now Armed Forces (Stage 4). But afterwards I will in any case continue (Aspiration Stage 5). Well, the main reason was actually that you can help yourself and simply have fun, fun with cars (Aspiration

Stages 2 and 3). The interviewer tries to understand these first insights into Anton's experiences and reflections by concentrating on the temporal and thematic structure of the professional stages (see Table 6.2).

Table 6.2 Temporal and thematic structure – example 1 Temporal structure

Thematic structure (ARB-model)

(stages) 1Salesman 2

Unskilled worker

Balancing: lacking challenges, low income Aspiration: personal benefits from the car mechanic apprenticeship Balancing: positive experiences with car repairs

3

Car mechanic apprenticeship

Aspiration: reinforced in the professional option as car mechanic by employer Balancing: negative evaluation of the training company but satisfaction with the formal apprenticeship certificate Aspiration (Stages 2 and 3): retrospective summary of the justification of the option of car mechanic

4

Currently: military

Page 25 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

service Future: 5as

work carAspiration: determination in targeting future stage: job as car mechanic

mechanic The temporal structure includes basic information that should be utilised for the interview. Its immanent logic constitutes the starting point for exploring further detail following the ARB-model. The extent of the opening account is never known in advance and there can be a lot of valuable information in one paragraph! We recommend noting at least the various stages that are described. In narratives, these stages are usually intertwined with actions, orientations and appraisals. The ARB-scheme helps to organise notes, and supports structuring pre-interpretations and developing strategies for follow-up questions. In the example of Anton, there does not seem to be much tension between the various stages and related aspirations, realisations and balancing; he recounts these stages systematically and with little emotional involvement. The final step of getting a training place in his preferred profession is not constructed as a dramatic highlight in the narrative; it is not introduced as a particularly important experience. Thus, in terms of follow-up, this interview does not suggest focusing on this event only. In opening accounts like this it is better to organise the follow-up along the stages in which the respondent unfolds the story. The inherent time line facilitates and consolidates the structure and gestalt of the narrative; and it establishes a framework that integrates the single aspects of the story. It is beneficial for the conversation if the respondent's time structure is also adopted by the interviewer. In the present example the interviewer should, in his first followup question, come back to the respondent's first stage as salesman by asking, for instance: ‘How did it come

about that you first learned to be a salesman?’ While listening to the opening account, the interviewer develops first ideas for additional follow-up, which should ideally connect directly to the opening account. These ideas depend largely on prior research

knowledge or previous interviews. The interviewer could, for instance, notice the following: in his explanation of the option of car mechanic, Anton refers twice to the personal benefits of this education, i.e. he can fix his own cars and does not have to ‘take the thing to the garage’. The interviewer could also notice that, in his professional aspiration, Anton deviates from the conventional societal division of labour: he does not talk about simply earning a living as car mechanic, which would then indirectly pay for his private needs (like repairing a car). The interviewer could also take up the respondent's point about income; it is important for professional choices and is even part of the interview guide. The interviewer could, for instance, explore his

pre-interpretation that Anton's emphasis on the personal benefits of the profession may be related to his low income. In terms of contributing to the household, the ability to repair cars would save him money in their maintenance and could earn him something on the side. It is important that the interviewer notes these questions or keeps them in mind, but delays them until they fit into the treatment of a specific stage or topic during the interview. Overall, the first interview stage is dedicated mostly to the collection, rather than in-depth exploration, of interview material. The preferred communication

Page 26 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

strategy is that of generating narration. Yet this should not be understood as a dogma (see Box 4.4)! First of all, the interviewer should actually understand what respondents want to say; they construct the story in a subjective perspective. For instance, the reflection of aspirations, realisations and balancing related to the steps towards employment involve meaningful assumptions, which are not necessarily obvious to the interviewer. They need to be unlocked through comprehension-generating questions (see section 4.6.).

Variation II: Dramatic Organisation of the Account The account of Uta Hairdresser also begins with an overview of the history of the professional biography that is temporally structured. Yet the crucial difference to the previous example consists in the dramatic climax of the narrative.

U: Yes, well, at the very beginning I actually didn't want to become a hairdresser (Aspiration Stage

3). Well, I left school and didn't find an apprenticeship training place (Realisation Stage 1). I wanted, actually wanted either hotel manageress or somehow into the office or so (Aspiration Stage 1). And then I didn't get an apprenticeship training place (Realisation Stage 1), I continued with two more years of commercial school (Stage 2), then I left school (Balancing Stage 2) because I got an apprenticeship training place as hairdresser (Stage 3) through connections (Realisation Stage

3). And well, then I continued. And then, well after the apprenticeship they did … well, during the apprenticeship they told me that, well they would not keep me. Before they used to say: ‘I'll keep her, I'll keep her’ (Stage 4). And then actually I was never really told that I would be dismissed after the apprenticeship. That was always only … my colleagues knew about it, my boss knew it, I was the only one, and I stood there somehow and did not know anything. And then, when it was close to the exams, actually I really got to know it only at the graduation ceremony. That's when they told me, well as if they would do me a favour, when they would say, I was dismissed (Realisation Stage 4). The temporal and thematic structure here is more confused and needs to be reconstructed (see Box 6.5). Apart from rejecting the hairdresser's profession in favour of an education as hotel manageress or an unspecified office job, Uta Hairdresser does not explicitly articulate any other aspirations in her opening account. Also her expectation to be able to continue working in the training company is never expressed as an aspiration. Neither her failed attempts to be trained in her favourite job, nor the temporary bypassing of the preferred transition to work by continuing commercial school, are at the centre of her opening account. Rather she elaborates on the disappointment experienced at the end of her training: the company's decision to dismiss her contradicted earlier positive signals. She is upset to be the last person to learn about it, while her colleagues were already informed about her failure. In a final twist of this episode, Uta recounts the employer's attempts to sugar-coat the negative decision: she was supposed to interpret her dismissal as a beneficial act that considers her otherwise more promising professional perspective. This episode is very significant in methodical terms; the intensity and emotional involvement reflect its high relevance for the respondent. Interviewers should be sensitive to such passages where respondents explicate Page 27 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

subjective relevance structures like this. If the interviewers are responsive to such indicative passages, they provide the opportunity to gain trust and underline the approach of being taken seriously. The respondent will perceive the interviewer's interest as a signal that he shares the moral evaluation of the employer's actions as a deviation from common standards. The follow-up for an opening account that includes an emotional episode like this has to be organised differently from the previous example of Variation I. Strictly following the order of stages by starting with the one that is most distant in terms of time would be counterproductive. Instead, the emotionally charged aspect of the account – here the unfair treatment and dismissal labelled as a ‘favour’ – should be taken as the starting point for further exploration. In the example, the interviewer does exactly this.

Table 6.3 Temporal and thematic structure – example 2, part 1 Temporal structure (stages) 1

Hotel

manageress,

(unspecified) – not realised

2Commercial school 3Hairdresser apprenticeship 4

Thematic structure (ARB-model) officeAspiration and realisation: failure to find a training place as hotel manageress

Balancing: discontinuation of commercial school in favour of an apprenticeship training as hairdresser

Realisation: successful search for a training place as hairdresser through contacts

Work in training company – notRealisation: disappointed expectation to be able to stay in the training realised

company

I: What? Why favour? I do not at all understand this. How come? The interviewer seems surprised by the twist at the end of the opening account (‘What?’). He spontaneously connects his follow-up to it by expressing his astonishment in the form of a specific exploration (‘Why favour?

…’). Uta's answer further elaborates the episode:

U: Well, that one does not remain there always as an apprentice, that one would get to know other companies (Realisation and Balancing Stage 4), and then, well, I was unemployed (Stage 5), and went to the labour exchange. And then I actually said, at the labour exchange I said then, that I do no longer want to get into this profession (Balancing Stage 3). And they told me, ‘No, that does not work, we don't do that, we cannot, cannot re-train you. You are too young, you have to pay yourself, and you have no impairment there’ (Realisation Stage 6). Well, and then they placed me. Yes, on the 3rd of August I was at the labour exchange, I registered unemployed, and on the 3rd of September I got a job from the labour exchange, again at a hairdresser's (Stage 7). And I had to introduce myself there, and they took me immediately. And that's where I am now (Realisation

Stage 7). Uta explains how she understands the reasons for the company's decision not to keep her: the supposed advantage would consist in starting in a different company where she would not be regarded as an apprentice anymore and could broaden her horizon. Page 28 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Uta's answer connects immediately to her professional biography indicating that she did not fully conclude her response to the opening question. In retrospect, the interviewer's intervention, trying to clarify the employer's particular behaviour, may appear as an interruption of her unfolding narrative. However, the interviewer's spontaneous reaction is a rather natural response to the climax she provided. Only afterwards it becomes clear that this was not the conclusion. Yet the interviewer's spontaneous reaction has positive effects on the conversation: Uta feels that she is taken seriously because her interpretation, and the way she provides hints for it, is understood by the interviewer and reacted upon accordingly. She can therefore continue her opening narrative despite the interruption. After the establishment of a moral agreement with the interviewer she can smoothly continue talking about stages and related aspirations, realisations and balancing. Other forms of interruption like the two described at the end of this part (Pitfall 6.4) can have more disturbing effects.

Table 6.4 Temporal and thematic structure – example 2, part 2 Temporal

structure

(stages) 3 4

Hairdresser

Balancing: no aspirations to continue as hairdresser

apprenticeship Work

Thematic structure (ARB-model)

trainingRealisation and balancing: the advantages of her dismissal in the view of the

in

company – not realised employer

5Unemployment 6 7

Retraining

–

notRealisation: rejection of occupational change through retraining by the labour

realised Employment hairdresser

exchange because of lacking medical condition of intolerance as

Realisation: employment as hairdresser through labour exchange placement

The second interview sequence finally concludes the problematic history and invites ideas for follow-up. They can be picked up at a later stage in the context of exploring certain issues. For instance, also in this passage the lack of aspirations is striking, i.e. Uta only expresses what she does not want. This could be a possible starting point for follow-up. Other possible questions resulting from prior research knowledge could refer to the role of commercial schools. Typically, they are bridging measures for young women that did not get an apprenticeship place in their favourite occupation and they try to improve their chances of getting a job. One could, for instance, ask why Uta neglects these alternative educational opportunities by leaving school and returning exactly to the originally aspired occupation of hairdresser. This moment of contradiction is also inherent in other aspects like her efforts of continuing in her training company and the disappointment regarding the dismissal, and in the later-expressed intention to make use of her unemployment by entering a retraining programme. The interviewer understands that the opening account has finally come to an end and starts the follow-up according to the general recommendations: he follows the time line of the narrative and asks the respondent for details about pre-professional stages. His first question takes the respondent back to school.

I: Alright. That was a very quick sweeping blow. If we could go through it once again in more detail, Page 29 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

so all the way back to school once again. Were there also considerations of a different kind, back to this period, where you wondered: ‘What can I actually do now, or what must be done next?’ The straight-forward formulation of this question contributes to the relaxed atmosphere of the conversation and conveys to the respondent the genuine involvement of the interviewer in the story. The use of everyday language with all its redundancies, expletives, interruptions and other flaws is familiar to the respondent. It reduces the pressure to formulate the response in a sophisticated way. Summarising and mirroring the whole account in a nutshell as a ‘quick sweeping blow’ is a smart formulation and is skilfully continued by asking for an elaboration of the thematic overview (‘If we could …’). Combined with the initial recognition of the contribution (‘Alright’), it gives the question a vivid twist underlining the interviewer's participation in the description without running the risk of providing false appraisals. The formulation continues to contribute to the good interview atmosphere: the interviewer does not imitate the artificial role of a completely neutral agent moving from one sober and pre-formulated question to the next. With a few sentences he leads the respondent to a moment in her biography that she should take as a starting point for further descriptions. The interviewer tries to recall details of the complex narrative and moves carefully towards the sequence of stages. At first, he remembers commercial school as a realised option (‘so all the way back to school’). By exploring the kind of professional ideas that stood at the very beginning of the history, he obviously recalls Uta's uncertainty about alternatives. The formulation of the question invites answers in the direction of wishful thinking (‘what could I actually do now’) and introduces the idea of a job as duty following the imminent completion of school (‘what has to be done next now actually’). By articulating his thoughts in the sense of ‘thinking-aloud’ (Ericsson and Simon, 1980), the interviewer formulates his exploration of possible options in an understandable way. For several reasons, this is a good strategy. The respondent can appreciate the sincere efforts of the interviewer to increase his knowledge by asking questions that bring him closer to her own perspective. In addition, the interviewer creates trust and motivates her to continue. Both aspects are preconditions of an extensive and valid account.

Pitfall 6.4: Interruptions The reasons for interruptions may be manifold and, as we have just seen, there may be good reasons for them. Yet usually interruptions involve a loss of distance to the respondent, a break with the role of listener, and certain impatience on the side of the interviewer. Typically, interruptions either express the need for immediate clarification of certain aspects of the account that were difficult to understand (acoustically) - then they could be called clarifying interruptions - or they result from the fact that the interviewer is being entangled in pre-interpretations and gets distracted from listening to the respondent. The disturbed balance between the interviewer's attention to the ongoing account of the respondent and the interpretive reflection of what has been said is a common problem

Page 30 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

in interviewing. The result of this brooding about pre-interpretations, e.g. a spontaneous idea, is taken up and expressed impatiently in the form of an immediate intervention - let's call this type an erratic interruption. In general, interruptions contradict the priority in qualitative interviewing of focusing on the respondent, and may throw the interview back into an unproductive question-answer scheme. There are situations when interruptions for the sake of clarification are legitimate; yet they should be the exception. For instance, clarifying interruptions are indicated when sudden noise or unclear formulations obscure important details. Such interruptions are legitimate and should be made in a friendly way and corresponding to everyday situations, for instance, by saying: ‘I beg your pardon; I just could not understand what you were saying.’ Apart from these instances, there are no substantive reasons for interruptions because spontaneous action (and the retrospective description thereof) evolves without the necessity for clarifications. Erratic interruptions (which often occur when interviewers are afraid of losing track of details or forgetting spontaneous ideas from pre-interpretations) should also be avoided. Instead, interviewers should take notes of key issues and address them later in the followup. Often, open points are clarified by the respondent in the course of the account. Interviewers should refrain from interruptions even if they may be unproblematic owing to the established relationship of trust. The problem is that respondents may develop a sense of tolerance for interruptions and interpret these interviewing mistakes as part of the impatient curiosity and involvement of the interviewer in their experiences.

Example of an erratic interruption I: So the … you were trained as a hairdresser, and I would for now like to talk to you about how it happened that you learned this profession, and uhm …, yes, when did you actually start to think about your future profession, and how did it happen that you took this apprenticeship training? C: Well, I thought rather late about it in concrete terms, at the end of … well during Abitur (i.e. graduating with A levels; authors), that is, in the 13th (class) or so, that's when I thought about what I could do concretely. I: … So, one moment, I have to check this now. You have Abitur (i.e. A levels; authors)? C: Hm, is this bad?

Page 31 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

I: No, uhm, that's not bad, but it does not correspond to the information I have here … C: … But perhaps not the … I: Here, it says quali (i.e. qualified general school; authors) … C: … No … I: … Well, we have to check this again then. Now, OK …, interesting … C: … Well, I got it all here, no. I got it all here, yes. Uhm, yes …, well. As I said, very late. Uhm, the plan to become hairdresser, it is, well … it consists of very different things (…) (extract from a narration that continues for 75 lines altogether). The interviewer starts with a spontaneous opening question close to everyday communication and complements it with a question about the moment when the respondent started thinking about her interests in the profession of hairdresser. Carola Hairdresser is interrupted shortly after starting her opening account. The contextualisation of her first ideas about a possible profession prompts the interviewer to mentally compare this information with his social background data about the respondent. As soon as he realises the inconsistency, he is irritated and reacts spontaneously with an erratic interruption. He cuts off the respondent's opening account: first by demanding a clarification, and then by mirroring her contradictory answer. What follows is a sequence of rather unproductive questions and answers that could have spoiled the interview. The interruption is hardly justified: there is no urgent need for pinning down the actual qualifications at the beginning of the interview. At this point the subject matter should be unfolded and the basic biographical structure outlined for a reconstructive understanding of the history of entering a professional track.

How is the erratic interruption overcome? It seems the two parties manage to clarify the interruption: the interviewer stops insisting and Carola continues in an apparently undeterred way (for 75 lines of text altogether). However, such a remarkably extended account after an interruption should not be taken for granted, or attributed to the respondent's eagerness to talk about her life. The fact that Carola maintains her cooperative attitude, returns to her story, and continues her account despite the interruption, may have other reasons.

Page 32 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

First, Carola's reaction to the interruption and to the interviewer's surprise about her A level education does not take the form of an irritation. The formulation of her response - ‘Hm, is this bad?’ - could well be an expression of irony and resistance to the interviewer. She seems to react to somebody who is irritated by encountering a hairdresser with A level education. Second, if we stick to this reading of Carola's response as a subtle critique, it does not burden the interview progress. The way in which the irritation is overcome can be reconstructed by means of a documentary method of interpretation (see Box 2.3 and section 4.6). If at first Carola interpreted the astonishment of the interviewer as irritation, she obviously becomes aware of her misinterpretation as the interview continues. The interviewer's further comments - i.e. ‘does not correspond to the information’, ‘it says quali’ - do not support this interpretation. A new interpretive pattern referring to ‘faulty data management in the research project’ seems more appropriate. This shift in the framing of the interpretation of the interviewer's position unburdens the situation: she no longer has reason to doubt his positive attitude, which is crucial for successful communication. Moreover, it is remarkable that after this brief erratic interruption has been resolved with the interviewer's evaluation (i.e. ‘check again’, ‘interesting’), Carola does not await another question before continuing the interview. She interprets it as a now concluded intermezzo and continues her narration without hesitation. Her initial difficulties to find the right words indicate her efforts to continue with her narration right where it was interrupted. This interview sequence shows that, in this case, it is not the robustness of the PCI as a method but the respondent's competence to reflect and articulate herself that is crucial for overcoming the interviewer mistake of interrupting. Interviewers cannot rely on individual qualities of respondents, and erratic interruptions can easily spoil the conversation.

6.2 Study B – ‘Costs of Custodianship Law’ This second project description exemplifies methodical peculiarities of using PCIs as expert interviews. We restrict ourselves to aspects complementing the description in the previous chapter - the discussion of issues of sampling, prior knowledge interviewer training, etc. is shorter here. In order to also discuss the chances and risks of peer research, our examples focus on interviews that were conducted by judges (as peer interviewers) with judges (as respondents). The overall approach and purpose of this study was very different: PCIs were used in a commissioned and applied investigation. The research question was practice-oriented; the study had to be done within a short time frame, and was concluded with a report to the commissioning party. Further research and additional publications from the material required the agreement of the ordering party.

Page 33 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Research Interest and Approach The study with the title ‘Cost reduction in the practice of custodianship’ combined secondary data analysis with expert interviews and was conducted from April to December 2002 for the German federal state of Lower Saxony (Ackermann et al., 2004a, b; Haase et al., 2003). The new custodianship legislation, which came into force in 1992 and was celebrated as the ‘reform of the century’, completely redefined the legal understanding of care for adults in need. For instance, groups of mentally ill people, addicts and dementia sufferers were established, and required intensive and expensive services. The new approach focused on the well-being of the persons concerned and on their needs, instead of incapacitating and exposing them to the largely anonymous administration of authorities. The new law strengthened their rights, suggested a policy of oneto-one care, and emphasised the idea of rehabilitation. Some adaptations contributed to the rising costs of the increasingly recognised custodianship law but could not explain why and how the costs of administering the new legislation had exploded by 80 times since its introduction in 1992. It was the task of the project to investigate this cost explosion of care both in total and on average per case. Already a preliminary exploration suggested the complexity of the issue. The process involves many actors and institutions that interact on the basis of a strict division of tasks: voluntary custodians (e.g. family members) interact with professionals, representatives of courts and public agencies, as well as with institutions like retirement homes and hospitals. They all have to interpret the relevant legislation and co-construct the legal practice interactively; they use their influence, translate experiences into changing practices of care, and react to changing contextual factors and job requirements. In the end, the complexity of the resulting web of responsibilities and interpretations was the suspected cause of the alarming cost increase. The study should reconstruct the development of costs over the years and describe custodianship practice from the perspective of different actors. The interaction of single cost-determining aspects of the administration of the legislation was analysed on the basis of secondary analyses of data available on the national and regional level. In addition, expert interviews were conducted with judges, judicial officers, members of administrative agencies, etc. in order to disentangle the complexity of changes in the actual practice of channelling clients into the programme. The project team consisted of: five district judges and a judicial officer as interviewers (all of them had participated in an earlier working group in 2001 and had collected extensive background material concerning the development of costs in the practice of custodianship), one academic researcher and assistants from the University of Bremen. Because of the complexity of the issue, qualitative expert interviews on the basis of PCIs were conducted. They were used to clarify the different situations in the various districts and were able to identify certain trends (e.g. about the composition of the clientele in the districts) that complemented the limited statistical data. In the sense of ‘systematising expert interviews’, they focused on ‘knowledge of action and experience, which has been derived from practice, is reflexively accessible, and can be spontaneously communicated’ (Bogner and Menz, 2009: 46–7). Their quality as expert conversations was twofold: Page 34 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

respondents as experts in different operational positions were interviewed by experts of custodianship law and practice. In this way, the shortage of prior (research) knowledge due to lacking studies was compensated for by the expertise of the peer interviewers.

Sample In order to achieve a reasonable coverage of the 80 districts of Lower Saxony, the criteria for selecting 16 of them were: number of judges, size of the judicial district by inhabitants; costs of supervision per inhabitant, regional structure, and increase in costs from 1997 to 2001. Three districts were considered special cases and were included owing the particularly high share of volunteers or health care proxies (Vorsorgevollmacht) as well as one district that used a novel way of accounting (i.e. based on a lump sum) for freelance professionals or those organised in associations. In total, 59 people were interviewed: 13 judges, 16 judicial officers, 13 representatives of public care agencies, 11 professionals, and 6 members of associations.

Prior Knowledge For this short applied research project the main source of prior knowledge consisted in a variety of statistical data about the development of the amount of people under custodianship. The Ministry provided additional reports about the costs including raw data that were analysed in advance. The five judges and the judicial officer participating as interviewers were a most valuable source of information providing everyday, contextual and specialised prior knowledge. Their expertise compensated for the lack of relevant studies; in particular their specialised legal knowledge benefited the academic researchers who would have otherwise needed a lot of time to acquire this knowledge. The researchers, on the other hand, prepared the members of the working group for the implementation of the study and trained them as peer interviewers (see below).

Interview Guide and Interviewer Training: Interview Guide The interview guide had to be adapted to the fact that experts interviewed experts. The first part of the interview guide (see Appendix 2 for the interview guide for judges) addressed the practical involvement of actors in the custodianship procedure. Their practical knowledge was primarily based on their specific role related to the overall procedure. Furthermore, it was possible to find out how experts associated the question of costs to typical or untypical cases, obstructive or conductive conditions, more or less useful regulations, and to the perception of the changing societal need for care. The second part of the interview guide directly addressed the increase of cases, the overall costs, and the costs per case. The third part explored the respondents’ view of the core of custodianship law together with aspects of the procedure that should be changed or are worth being maintained. In this way the participating experts should be invited to reflect their view on the problem and to provide ideas for possible alternative procedures. The aim of this last part in particular was to make sure that the respondents had a chance to display their expert knowledge vis-à-vis the expert interviewers. Page 35 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Interviewer Training The training of the project team included the discussion of the interview guide and was done at the University of Bremen. Owing to their practical experiences with social law, it was easy for the judges to understand the

methodical principles of the PCI and to develop the right attitude towards their role as interviewers. From their counselling work at court, they were already familiar with basic aspects of non-directive conversation. As peer interviewers, the five judges and the judicial officer had a high level of expert knowledge. Their position allowed them to assume the role of ‘co-experts’ and to formulate original follow-up questions during the interview. The training had to make sure that their distinct prior knowledge and their special status would not bias the interviews. It should help to reduce their susceptibility to overburdening questions, exaggerating their involvement as experts, inappropriate follow-up (e.g. impatiently pushing the interview ahead), and insinuations of meaning. Mistakes like these restrict the richness of the collected material and the validity of the accounts; and they can happen not only in this special form of expert-to-expert PCIs. Pitfalls 6.5 to 6.8, which are taken from this study, indicate that these mistakes can happen despite interviewer training. They underline the importance of addressing such mistakes in the preparation.

Pitfall 6.5: Overburdening of Questions Especially when interviewers share the respondents’ high level of expertise, they are at risk of packing too much knowledge and interests into single questions by starting with one main question and complementing it with several sub-questions. Even if the respondents could easily answer all sub-questions, they still have to choose which one to address first. The more detailed their answers are, the more likely they may forget about others or run out of time. The following example starts with the kind of yes/no question that should generally be avoided. The respondent is a competent communicator and not affected by this pattern; he understands that he is expected to provide a more extensive answer. Yet he manages to answer only the first part of the question. Probably, after hearing the first part, he already ‘disconnects’, stops listening to the interviewer, and prepares his answer mentally. He only picks up on the term ‘professional’, used in the question, and integrates it into his answer.

I: We talked about the selection of custodians. Do you have cases where it is difficult to find suitable volunteers, and then being in trouble tending to use professional custodians of members of an association, although custodianship is

Page 36 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

not suitable for professional guidance? R: Yes, definitely. Especially here we have problems with deaf, with mentally ill people. It doesn't work because we do not have anybody who can communicate with deaf people. We have professional custodians, although they only go there once in four months. The costs of the assistance, so the costs of the home, everything is set, health care, there are no serious decisions to be made, but it is difficult to find voluntary custodians. There has been a newspaper article about it recently. As a result the home came forward and said that it was not that bad, we can still, we can be of help. Sometimes you need to have the right contact person. I don't know if it works now, but for certain cases it is definitely difficult to find volunteers. Of course it is important for them that they can communicate with the people and although there certainly are cases where I … where there is not much to do in professional terms and a volunteer could do that, but he gets little feedback and says: ‘I don't do that.’ That's a big problem, yes (emphasis added; authors).

Pitfall 6.6: Exaggerated Involvement as Expert Expert interviewers tend to introduce their own opinion into the conversation. This should be avoided as it undermines the PCI's idea of prioritising the respondent's perspective. In the following example about gross costs of support personnel, the interviewer clearly acts as co-expert and does not even formulate his contribution as a question. The response may still include valuable information; yet the articulated appraisal can no longer be attributed to the respondent alone.

R: And I think one would need to set up new models, because, for instance, also concerning the craft profession … I: I know that. During the last interview on Friday as it were, I just said, it also happened then … So, craftsmen have an appropriate possibility, finally this is something we really need to do something about, because it is so unsatisfactory.

Pitfall 6.7: Inappropriate Follow-Up, Impatiently Pushing the Interview Ahead

Page 37 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Overloading the interview with too many issues is a general problem. PCIs are not conventional guided interviews; they just use interview guides to organise the relevant aspects of research. Their main interest lies with exploring the respondent's point of view regarding specific issues. Time pressure introduces to the interview a moment of competition between research interests and topics addressed by the respondent; and we can never really be sure how long an interview will actually last. We already discussed one way of coping with this dilemma under the label of ‘interview guide bureaucracy’ (Hopf, 1978, 2004: 207; see section 3.3), i.e. interviewers neglect follow-up questions because they are more concerned with the list of topics than with in-depth exploration. Sometimes following-up is difficult because it is hard to understand what respondents actually mean. In some cases, however, interviewers simply ask inappropriate follow-up questions. The reasons can be manifold. In the following example it could be anything ranging from disinterest, lack of concentration and attention, to time pressure and interview guide bureaucracy.

R: A little less capable person in custody, there are possibilities for development but you cannot assess it precisely in advance, because you never know in the beginning or even after a year or two you don't know, how far the person can go. I have several younger clients who I'm proud are now in employment and then you realise that they cannot cope with the burden and that's the question then to be able to show. It is simply impossible. There are always new additional aspects. I: Hm, these clients with their problems are, don't behave as the legislator once wished they would, but the problems become always new, always new … R: New problems then. I: Yes, yes. R: Hm. I: OK, good. Then my second topic, thematic area, my second thematic area, question of costs with support personnel …. (point 3 in the interview guide; authors)

Pitfall 6.8: (Mutual) Insinuation of Meaning

Page 38 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

When two experts, or peers more generally, talk about a familiar issue, they usually do it in a way that is difficult to understand for laypeople or non-members of their environment: either their conversation is permeated by specialised terms and language incomprehensible to the outsider, or it lacks the level of detail that nonexperts would need to be able to follow. The latter case becomes problematic when two interview partners that recognise each other as colleagues or peers rely on each other's insinuations of meaning. They assume to be sharing specialised (expert) knowledge, evaluations and experiences. This can prevent certain issues from becoming explicit; both validity and the breadth of the material suffer from an imagined identity of expert and practical knowledge. An example from STUDY B illustrates the danger of joint efforts of reconstructing specific knowledge being unproductive - on both sides, taken-for-granted assumptions remain undisclosed when questions and answers are too vague. In the following, the question is discussed as to whether relatives are suitable as custodians of persons in need of legal care. Ideally, the interviewer would introduce a comparison of volunteering relatives with other non-related volunteers and professional custodians. Interviewer and respondent should clarify, for instance: whether being a relative can undermine the capability of assessing the needs of a person in care; or whether one should rather be in favour of more expensive professional custodians who comply better with the requirements of legal advocacy.

I: Do you have good experiences with relatives as custodians? R: As long as it is not about money. I: At this point it gets difficult? R: At this point it gets difficult. The short question introducing the issue of positive experiences is inappropriate in methodical as well as thematic terms. It applies the conventional question-answer pattern (see Pitfall 4.1), could be answered with ‘yes’ or ‘no’, and altogether targets short answers. Taken seriously, this question about ‘good experiences’ can only encourage summative evaluations; it misses the purpose of treating the issue in a comprehensive and differentiated way. The first response is an evaluating answer implying general agreement: good experiences do exist, but not when money is involved. At this point the interviewer could have taken up this distinction with the respondent. The belated introduction of a narrative stimulus could

Page 39 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

have compensated for the inappropriate question and stimulated an account for exploring this evaluation. For instance, the interviewer could have asked: ‘What are your particular experiences (regarding questions of money)?’ Instead, the interviewer implies that the respondent has experiences of some kind. He continues with paraphrasing and mirroring the respondent's answer on the level of the evaluation, which is then simply repeated by the respondent using the exact wording. For both interview partners the topic is completed at this point just as if a final questionnaire question was answered in a satisfactory way. The whole background of the respondent's experience remains hidden behind a very brief judgement. After all, the two experts participating in the conversation seem to agree on this matter. One can even imagine this brief exchange to be accompanied by their smiling knowingly at each other. In the course of the interviewer training, the expert interviewers were taught to be particularly sensitive towards possible conflicts of loyalty between respondents and the organisation in which they worked and were interviewed. Such conflicts are relevant when individual behaviour or ideas are not in line with the official image of organisations. Usually, such concerns were explicitly refuted by the interviewed judges. They even underlined that anonymity was unnecessary and that they would publicly vouch for their opinion. This reflected their professional status involving self-confidence, independence and immunity and meant that conversational precautions against socially desirable answers were not necessary. Yet the respondents had other concerns. They were worried that some of their frank answers might facilitate cutting measures although they were actually against it. Such a form of bias could, of course, undermine their openness and result in self-restricted answer patterns. Therefore it was crucial to anticipate, in the course of getting in touch with the respondents or at the beginning of the interview, that there may be reasons for cost increase, and that this could be perfectly reasonable in terms of good practice (see introductory explanation of the interview guide from STUDY B in Appendix 2). Furthermore, the design of the interview moved critical questions to the end; the actual possibilities of cutting costs should be explored only after discussing the practice of custodianship. Descriptions on the level of action are usually connected to evaluations and balancing. Thus, in case respondents spontaneously commented on measures influencing costs, it was noted but addressed in more detail only later (see remarks at the end of point 1 of the interview guide in Appendix 2).

Doing Expert Interviews with Judges: Opening Question A typical first sentence of the introduction of the interview was formulated like this:

I think the best way for us to understand your view is to learn about your concrete work as a custodianship judge, which means on the basis of the process of your work involved in a typical custodianship procedure. Page 40 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

First of all, this sentence is not formulated as a question. It is rather the suggestion of a thematic narrative

frame that is connected to the introductory explications during the first contact. The interviewer can make use of the agreement negotiated in the course of the preliminary talks to organise the interview along the respondent's descriptions if he manages to specify the frame of these descriptions, i.e. he has to orient them towards the practical suitability of the regulations. This introduction supports the organisation of the interview in two ways: first, it reduces the breadth of the possible account by highlighting the concrete work involved in a typical case; and second, it refers to the process character of the issue and the possibility to describe it step by step. The emphasis on the view of the respondent underlines that – apart from actions, their preconditions and consequences – the interviewer is also interested in appraisals as part of the respondent's practical involvement. The opening question continues with:

How does a case land on your table, or: how does a case actually start? And how does it continue then? These questions specify the thematic narrative frame by mentioning some practical steps of the work. Unlike in STUDY A about transitions to employment, an explicit narrative prompt was not included in the interview guide; it was considered unnecessary because of the process character of the topic, which already implies a narrative structure. Nevertheless, this entailed a few problematic consequences for the way the interview guide was used, as will be discussed below.

The Importance of the Beginning of the Interview Some of the interviewers did not fully understand the narrative purpose of the last part of the opening question; it was obviously not clear enough: they omitted the question that implicitly targets a narrative account (i.e. ‘And how does it continue then?’) although it was crucial in terms of communication strategies. They did not consider its important narration-generating function. In return, some respondents were irritated, especially when the formulation was even further reduced. For instance, one of the interviewers, a judge, asked: ‘Therefore my first question: how does a case actually land on your table?’ Such a question clearly contradicted the advantages of qualitative interviewing and the replies were formulated as if they were part of the conventional question–answer pattern: the brief question was followed by a brief response. These interviewers had to correct their communication style during the follow-up stage. Their efforts were supported by the experience of the legal experts and their routine in communicating in public: they had something to say and were resilient to interviewing mistakes; and they were not easily interrupted, as the following example in Pitfall 6.9 shows.

Page 41 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Pitfall 6.9: Resistance to the Interviewer's Change of Subject It was quite common in these expert-to-expert interviews that respondents (i.e. judges) resisted being interrupted in their accounts. As the three examples from STUDY B indicate, they were still in the middle of unfolding a specific topic when they refused to follow the lead of the interviewer. Once they had accepted and assumed their roles as the actual subjects of the interview they insisted on concluding their thoughts. In the first two examples, this self-understanding is so strong that they even interrupt the question of the interviewer.

R: During the meetings with the custodianship association it was certainly discussed that it is difficult to win people for this somewhat more problematic group and that it is of course easier to assume a voluntary position in some association that develops activities that are not particularly problematic. I: Let us now come to … R: I just thought of another issue related to these volunteering stories, where we also have sports clubs … I: When you come back to these custodianship files after a while … R: May I add one more here? I: Do you actually have preferences concerning people with specific disease or disability patterns? R: I started and said I attend only to women and no alcoholics. In the meantime I attend to everything. I just thought that as a woman it is easier for me to care for women and that for women it is easier to have a female custodian. Not for all but I think for most. Yes and then after some years there was a request for an old gentleman, and I could easily imagine that and then it became two, three more, but the majority is still more women. And I also had an alcoholic and it went very positively, I could apply custodianship; by the way, it is the only one, where, after three years, I could suggest rescinding custodianship. I: Could you review a custodianship that you are doing. What would be your main field of work? R: I did not finish answering the other question.

Page 42 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

I: I see! R: I attend to young, mentally disabled. I attend to people that still live in their flat, old as well as middle-aged. I attend to … Some of the interventions suggest that the respondents interpret the sudden thematic changes of the interviewer as exaggerated articulations of involvement and curiosity. They react by sticking to their story and consistently further developing their common theme. In principle, the PCI allows for such interventions of the respondents for the sake of the validity of the statements. Experts rather than other respondents may have the kind of self-confidence to resist the lead of the interviewer. Thus, changing subjects before the respondents have finished and impatiently pushing the interview forward should be avoided.

The PCI as Expert Interview: The Role of the Interviewer as Co-Expert (a) The interviewer's self-presentation as competent discussion partner One distinctive requirement of the expert interview is the interviewers’ communicative and interpretive capability to bridge the various knowledge areas of the respondents. As co-experts they need to communicate in terms of substantive issues at eye level with the expert respondent. Their ideas for follow-up questions depend on this capability and the recognition of the interviewer as equal conversation partner. Often, the expert respondent needs to perceive the interviewer as a congenial partner in order to be motivated to provide detailed answers. Specific advantages as well as risks for the reconstruction of the respondents’ knowledge have to be considered in case experts interview their colleagues and peers, like in the following example where both interviewer and respondent are judges. For instance, their overlapping knowledge provides the chance to improve the validity of the collected material. Yet there is also the risk that interviewers neglect to explore aspects that are taken-for-granted and everyday knowledge for both of them. Or they simply, and wrongly, imply them on the basis of their own experience and background knowledge (see Pitfall 6.8). The following example features a judge as interviewer and co-expert. His questions and the way he is able to confront the interviewed judge on the basis of his experience reflect his competence. Their supposed common understanding becomes evident and indicates the fine line between explicating co-expert knowledge and influencing the direction of the answer.

I: When you now have all the things together for the initial appointment. The decision is ready. I don't know, I often use this sentence at some point during the hearing: ‘If everything goes well, then you won't see me for five years.’ R: Yes, exactly (she laughs). Page 43 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

I: It often happens that not everything goes well. How about you? In which way does the file usually come back to the table? On what occasions? Routine controls or rather other things in between? R: Well, I anyway have deadlines for resubmission. Where I think that a lot can happen in the near future or I am simply also interested in how it develops, or where I think perhaps I or the judicial officer who supervises the guardian should follow up and find out whether something really happens on the part of the guardian - in these cases I have the file resubmitted after six months, in other cases after a year, mostly after two years. This is somehow the routine deadline for me (…). In the first sentence the interviewer refers to the background of the arrangement of care, which results in a legal decision (‘initial appointment’). This is followed by an original way of introducing the next step in a typical care procedure: in order to prepare the exploration of issues concerning changes at a later stage (see Appendix 2; point 1.5), the interviewer, a judge himself, confronts the interviewed judge with experiences of his own legal practice. He says that, after the decision making in the hearing, he usually expresses the expectation that the care process evolves within the determined period of five years in a well-ordered way and without problems. He can assume that the interviewed judge shares this experience and also tries to get the files off the table before the legally established deadline. When the interviewer insinuates that this expectation is an ideal that often contradicts reality, the respondent laughs in agreement and confirms the relationship that connects both interview partners. They both know what they are talking about; the overlap of their professional knowledge is confirmed. The interviewer can be confident regarding this professional agreement over typical expectations, and his anticipation of this agreement in his remarks has a positive effect on the situation: the interviewed judge recognises him as a co-expert, and as a competent and understanding colleague and conversation partner. The remark that follows further strengthens their relationship as fellow sufferers from the reality of custodianship cases: contrary to any kind of wishful thinking, care arrangements usually do in fact cause problems. This remark serves as a good transition to the question from the interview guide targeted at the procedure and at concrete steps taken by the judge. Ideally, the interviewer could have ended his intervention by pointing to ways and occasions at which a file is returned to the table. Additionally mentioning common routine controls hardly focuses the conversation, and also the somewhat vague allusion to ‘other things’ is not necessary. Finally, as Pitfall 6.10 illustrates, another advantage of employing peer experts as interviewers consists in the fact that misunderstandings can be clarified more easily owing to the common basic understanding of the issues.

Pitfall 6.10: Misunderstandings and Their Clarification in Peer Expert Interviews

Page 44 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

At first glance the following passage is only about clarifying a misunderstanding in the conversation. Yet there is something more to learn regarding the respondent's definition of the problem and how its aspects are interlinked.

I: How do you deal with a medical opinion in cases where the need for assistance/ custodianship has been approved by a medical doctor while you do not identify this need during the hearing? R: How do I deal with the medical opinion? I: With custodianship proceedings. R: Ah, with custodianship proceedings. It does not amount to an order for custodianship. So, if you like, the opinion is one part of custodianship proceedings, but just one part. So I do not feel bound to the medical opinion. For the judge (respondent) it is clear that the medical opinion is only one part of the proceedings. The detail of ‘dealing with the medical opinion’ brought up by the interviewer does not connect to his perspective because he does not really ‘deal’ with it (in terms of rejecting it in writing, or sending it back to the expert). In other words, for the judge this detail is not part of the context, it does not make sense - at least not in the same way as it does for the interviewer. The self-correction of the interviewer indicates that ‘dealing with the medical opinion’ is indeed crucial for his interpretation of the process; it is a meaningful detail in this very context. In methodological terms this form of negotiated interpretation is an example for the documentary method of interpretation (see sections 2.2 and 4.6). (b) Conditional introduction of knowledge by the interviewer Another issue that is crucial for this kind of expert interview refers to the amount and kind of knowledge introduced by the interviewer: (to what extent) does the introduction of knowledge improve the validity and richness of the material, or does it rather contribute to the risk of biasing the statements of the respondent? Concretely the provision of specific information - in the following example the interviewing judge mentions that there are ‘considerations to

amend the law’ - can be interpreted in two ways: either negatively, because relevant aspects of research are anticipated; or positively, because this specific factual knowledge is altogether irrelevant for the research goal. Let us look at the interview passage, which involves highly indexical specialised terminology, typical of expert interviews.

I: In this context there are considerations to amend the law, namely in such emergency cases like PEG-tubes (i.e. percutaneous endoscopic gastrostomy tubes - short-term measures of artificial prolongation of life through feeding tubes; authors), that from the beginning a relative is able to agree, thus the Texan model. Page 45 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

R: I would think that this is reasonable. Well, precisely in the case of PEG-tubes. In the end there are no other possibilities. I do not really have decision-making ability. I am, so to say, only nodding it through for the physicians. If the physician tells me: ‘The PEG-tube is necessary otherwise the patient dies.’ Should I then say: ‘The PEG-tube is not used’? I don't have another possibility. Also the relatives do not have another possibility. Regarding the artificial prolongation of life as such, to connect to machines and so on, that's certainly something else. But the PEG-tube, this short-term measure, that is such a preliminary thing. I think it would make sense if more things were outsourced. The interviewer's implicit assumption, that the judge is aware of this knowledge, is not necessarily justified. Being up to date with the latest development of legislation requires at least some efforts. Thus, providing this information (‘there are considerations to amend the law’), that in the end constitutes an implicit question (see below), involves a confrontational aspect which needs to be used carefully by the interviewer. Before resorting to any form of confrontation, one has to be sure about having established a positive

relationship with the respondent, and that, in the example, an interviewed judge is certain to be recognised in his role as expert. Only if this is the case, a judge would be able to admit to knowledge gaps, for instance, by answering something like: ‘I've already heard about this but I have yet to get into the details.’ Then he could also accept and even expect additional information by the interviewing colleague in order to establish a factual basis for discussing the topic. Providing such information in this case can be appropriate also for another reason: the discussion of a possible amendment of the law is not the actual topic of the study; it is only an implicit aspect of the investigation of practices and costs. The interviewer wants to arrive at a comparison of the current framework conditions with changes connected to the possible amendment. Such a comparison should conceptually sharpen the perspective on the current situation and create ideas for improvement. In the example, such a comparison suggests itself at the end of a discussion about the details of the use of PEG-tubes and their status within certain legal proceedings (i.e. ‘the Texan model’). The formulation of the interview question first of all targets the respondent's subjective evaluation of certain options and their consequences for the actual practice; the availability of a certain factual knowledge is not under investigation here. Strictly speaking, in the above example, the intervention of the interviewer is not a question but a statement directed at the respondent. Yet obviously none of them is aware of this: the respondent could have answered the question with a simple ‘yes’; or he could have pointed out that he is not familiar with the latest discussions. So, why does the communication nevertheless work? The respondent answers without problems because he agrees with the interviewer about the meaning implied in the intervention. The confrontation of the respondent with possible new legislation is similar to asking: ‘What would be the practical consequences of this?’ The respondent immediately understands this subtext in the context of the preceding dialogue about PEG-tubes. For him the information about possible legal amendments is a signal to comment on the interviewer's suggested alternatives on the basis of his Page 46 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

own experiences. The first part of the answer – ‘I would think that this is reasonable’ – underlines this interpretation. The fact that the respondent goes beyond this spontaneous appraisal refers to a crucial aspect of the interactive approach of the PCI: in the course of the interview the respondent ‘learned’ to participate. He understood that his viewpoints are taken seriously by the interviewer, and that his colleague's communicative interventions have first of all the purpose of stimulating and supporting his explications. His expertise and authority are not questioned (see Pitfall 6.9).

The Advantages of a Consistent Grid of Questions The use of a grid of questions in the interview guide has the following methodical advantage: the method of comparing and contrasting cases, which is usually used in the context of data analysis (see section 5.3), is already applied during the interview. This idea is associated with the often neglected PCI concept of

pre-interpretation: it serves as a knowledge generator during the interview and accompanies the interaction between researcher/interviewer and respondent (see sections 2.3 and 4.6). (The ARB-scheme described in section 6.1 served as a similar means of systematising pre-interpretations in STUDY A). In the case of an expert interview, the use of a consistent grid of questions has a similar function. The interviewer controls the object orientation of the account by keeping certain questions in mind; and he further systematises the interview by formulating ad-hoc questions on the basis of certain adjectives (or synonyms), or by inviting the respondent to reflect differences on the basis of contrastive pairs and thematic comparisons. A few final examples should illustrate this strategy.

Typical versus untypical The use of the grid of questions is first illustrated in the opening statement – ‘I think the best way for us to understand your view is to learn about your concrete work …’ (see above). It establishes the narrative frame and the object orientation of the response. In order to understand the concrete work of a custodianship judge, the opening question suggests describing a typical procedure. Implicitly, it also asks for the description of untypical cases in the later follow-up. The contrast between the two accounts should produce a differentiated overall picture of the work of judges and the possibility of distinguishing between general and specific aspects. The introduction of these terms to the interview conversation may well require some specification in order for the respondent to understand the question, like in the following case.

I: Are there also atypical cases? R: What is an atypical case now? I: Maybe one that can be administered in a particularly short time, or one that is very long, (and) that needs to be reviewed? R: There certainly are atypical cases. In particular, there was a case about a troublemaker who is now, essentially, able to manage his everyday things, but it is now about the question of legal matters Page 47 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

concerning the authority. He does not come along; he has to be summoned to all measures. It takes longer then. And then there are complaints, and in the meantime there are appeals. Of course such things can drag on for a while. Facilitating versus hindering This pair of opposites was used in order to analyse positive or negative aspects of the circumstances of each of the steps.

I: Let us now approach the issue of time burden. Because due to the deadlines you get more files (R: hm) on the table (R: hm, hm). Is the circulation of files in this way rather facilitating or hindering with regard to this early change? R: Well, it is of course facilitating with regard to the early change. It is hindering of course for my workload, that's clear, but of course it facilitates the goal that I pursue. Yes, that would not be a reason for me here to say: ‘I have so much to do, that's why I extend the deadline.’ That's not my goal, I have a different goal. More or less useful The introduction of this distinction was another way of getting the experts to talk about regulations and context features of the procedure worth preserving or not. It helped to collect concrete suggestions for reform or improvement. This distinction was, for instance, addressed in terms of a direct question: ‘Can you imagine useful alternatives that are at the same time cost-effective?’ Or the distinction was made by the respondent himself.

I: If one would exaggerate this now, how could one say it: who actually needs such custodianship, and when? R: Custodianship is certainly necessary for mentally ill people, who no longer manage to do certain things due to their disease, and who have deficits when there is no guardian that arranges some things for them. This is where I think custodianship is of imminent importance and also effective. In terms of usefulness, I have a big problem when it comes to old people. There I often feel that, in the end, custodianship was arranged only to serve the law, while it does not improve the situation of the people … Before versus after The before–after comparison refers to the process character of legislation and framework conditions. For instance, it became apparent that the changing composition of the clientele was one crucial reason for disproportionate increase in costs. There are more and more particularly difficult cases that need intensive care: especially mentally ill people, people with multiple addictions, and, among them, young people. Furthermore, the number of people with dementia is also growing.

This is a development of the last, say, ten years. Not before. But there, one can say successively. It goes according to the need. We have the need. People, I say, get older than before. Such questions that prompt respondents to think about the issue in terms of contrastive pairs and thematic comparisons are very effective means of systematising knowledge. Even in expert interviews, details related Page 48 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

to an issue are recalled only bit by bit and still need to be organised and brought into a temporal order. Expert respondents may also develop an issue for the first time and in this way – i.e. opposite and with the support of a person with a genuine interest in their subjective view on practical aspects of the problem. Highly trained judges were no exception: they had no ready-made frame of reference according to which they could introduce the interviewer in a structured way like in a lecture. (By the way, this is the reason why experts often ask for the interview guide in advance in addition to the usual information about purpose and content of the interview. They want to be prepared for the actual interview and ‘performance’ to come.) Spontaneous reconstructions start from preliminary thoughts and are gradually refined. According to findings from Gestalt psychology, preliminary thoughts tend to be completed in cognitive processes. In particular, narrative interviewing techniques make use of this mechanism by relying on constraints to narrate (Zugzwänge des Erzählens) – like the compulsion for completion (Gestaltschließungszwang) and the compulsion for detail (Detaillierungszwang) (cf. Hopf, 2004: 207; Schütze, 1977). The fact that these dynamics can produce accounts that are beneficial for the respondents’ self-understanding is used in conversational therapy. With regard to social research, providing respondents with the opportunity to reveal their thoughts and taking them seriously can also motivate them to continue with the cooperation. (This is particularly important for longitudinal studies where the process of self-understanding can extend over a long period of time.) The following example illustrates the construction of meaning in the making and is initiated by the introduction of a contrastive pair. The interviewed judge recognises the novel perspective of the interviewer's suggestion and articulates his opinion accordingly.

I: If you compared then and now, do medical doctors assess the situation differently now that they tend more towards the requirement of care? R: Well, the latter, yes. I: Can you think of reasons for this? R: Yes, well, I never really thought about it. Now that you bring up this problem, well, I could imagine as a reason that the expert usually assumes, in case he is asked to prepare an opinion, that this is a case of ‘care’, yes, well, that there is a kind of compulsion. And perhaps that there is a certain expectation projected on the other side. Well, in principle, the court actually expects that I arrive at my opinion with the conclusion that care has to be prescribed. I discussed this already with the two older psychiatrists that I do not at all expect that. At first glance, one could misunderstand the respondent's reply as a reaction to the introduction of an additional topic: he appears to react to the topic like in a conventional guided interview and without unfolding his own ideas. The whole sequence is able to refute this first impression. In his daily routine, the following assumption is relevant for the judge: medical doctors associate the court's invitation to write a medical opinion with the implicit expectation that they provide a positive opinion endorsing supervision. Yet he was not aware Page 49 of 50

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

of the problematic implications of this assumption. Only in the course of the conversation about medical opinions within the systematic logic of the interview was he able to draw conclusions from the reflection of his experiences. http://dx.doi.org/10.4135/9781446288030.n6

Page 50 of 50

The Problem-Centred Interview: Principles and Practice

A Final Note on PCI Errors and Pitfalls In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 174-185 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

A Final Note on PCI Errors and Pitfalls Our final chapter is an attempt to summarise some of the main issues involved in problem-centred interviewing in an alternative way. It complements the general and practical chapters with a discussion of forms of deviating from the basic attitude of problem-centred interviewers towards respondents. The systematic discussion of interviewing errors should help in understanding the basic principles of the PCI from

the ‘negative’ angle of discussing common mistakes. The change in perspective is a didactic one: we cannot expect to be in control of every detail of immediate communication; yet we can and should learn from making mistakes, especially from those related to PCI principles. Interviewing mistakes are often discussed by referring to the problem of interviewer biases and influences. The standardisation of questions and the rather reserved behaviour of the interviewer are among the efforts in avoiding such biases. They follow a certain neutrality doctrine that should help to increase the ‘objectivity’ of the data. Such a standardisation of interviewing behaviour has been criticised by qualitative researchers. For instance, following the classical argument by Watzlawick et al. (1967), one can argue that efforts of restraining oneself (Zurückhaltung) in communication have little effect because it is impossible not to communicate. Or, as Kahn and Cannell (1957: 195) emphasise, the special social relationship between interviewer and respondent always implies ‘influences’ – ‘to say (…) that we want an interview without interviewer influence is a contradiction in terms’. These influences are particularly relevant when the decidedly neutral behaviour of the interviewer (mis-)leads the respondent into orienting her statements along socially desirable values and norms, which she projects onto the (neutral) interviewer (Berger, 1974: 78). Influences of the interviewer are simply unavoidable and the suggestions of the PCI should first of all help to avert mistakes. These suggestions are essentially based on the attitude of the interviewer as a well-informed traveller, as well as the dialogic principles of doing PCIs. They constitute the basis for the management of interviewer influences that is committed to the following: the advancement of knowledge, the overcoming of prejudices of the researcher/interviewer and the facilitation of the reconstruction of actions, experiences and interpretations of the respondent. The principle of problem centring and its orientation of the interview along the thematic relevancies (and their disclosure) of the respondent should help to overcome the basic interviewing problem discussed in the literature – i.e. the contradiction between openness and the articulation of the very specific interests of the researcher/interviewer during the follow-up phase. For instance, unlike the narrative interview, in order to deal with possible strategies of resistance or of evading answers on the part of the respondent, the PCI does not need to refer to the power of ‘constraints to narrate’ (Zugzwänge des Erzählens) (Hopf, 2004: 207). On the contrary, the PCI relies on the (inter-)activities of the interviewer to facilitate the interest of the respondent in explicating everyday problems and related nuances, opinions, apprehensions, successes and failures, etc. to an interested and understanding conversation partner within a protected context and with the chance of the benefit of enhanced self-understanding. Therefore, verbal and non-verbal signals of the interviewer are not considered as forms of negative influencing but express understanding (or not) and empathy regarding Page 2 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

the respondents’ relevancies which are articulated in their accounts. Consequently, these signals are part of the phase of following-up and mirroring. (Depending on the level of detail of the transcription, they may also become noticeable and relevant in the interpretation.)

So, what is left of the neutrality doctrine? Initiative and authentic creativity of the respondent is most desirable during the interview; that's why restraining oneself verbally (Zurückhaltung) as an interviewer is certainly important. The interviewer first needs to ‘sit back’ and listen until she has a sufficient basis to further explore and deepen the respondent's account. Yet cautious listening (zurückhaltend) is not a basic requirement of the PCI. As a default means of securing openness it would cause stress for respondents if the burden of producing a successful account was theirs alone. The communicative and verbal support of the interviewer is necessary in case they lose the narrative thread or are uncertain about how extensive the account should be. The involvement of the interviewer is not a contradiction to the principle of openness: the respondent expects direct feedback which indicates understanding and comprehension. A more interactive form of conversation is generally closer to everyday conversation even if the interviewer still acts professionally. Trained interviewers interact systematically; they try to establish the ‘right’ understanding that they develop self-critically through

pre-interpretations and their revision. What can we finally learn from the debate about the neutrality of the interviewer? Interviews constitute social relations. Errors and misunderstandings cannot be avoided owing to the many spontaneous decisions that have to be made during interactions. For pragmatic reasons it is thus important to take a look at the process of interviewing from the angle of making mistakes. It can help to improve the technique of interviewing in general, and optimise ‘good’ techniques that the interviewer already has.

What Kind of Mistakes Are We Discussing Here? Many interviewing errors are ‘unspecific’. They happen in all sorts of interviewing and can be related to individual weaknesses, lack of concentration, or simply a bad day. These are not the kinds of mistake that we want to discuss here. Instead, we want to systematise specific PCI-related errors that were illustrated on the basis of examples. These are ‘typical’ errors reflecting misunderstandings and deviations from the principles of problem-centred interviewing, and they were collected over many years of working with PCIs. The following

classification of mistakes should allow the reader to understand the consequences of breaking methodical rules. It should enable the interviewer to deduce, understand and avoid errors that could undermine the main goal of reconstructing relevant aspects of the respondent's life-worlds in an unprejudiced way. It should help to refine the interviewer's attitude in a way that respondents are taken seriously in their ‘natural’ meaning context, i.e. they should get the chance and support to exhaustively present the subjective meaning of relevant actions and appraisals. Ideally, respondents that benefit from such an attitude are empowered to develop a certain resistance against interviewer mistakes, which can contribute to the ‘immunity’ and overall quality of the interview (see Pitfalls 6.4 and 6.9). Before we discuss these mistakes, let us recall the principles of problem-centred interviewing.

Page 3 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Basic Principles of Problem-Centred Interviewing a) Bridging the Spheres of Scientific and Practical Knowledge

The starting point of problem-centred interviewing is the joint communicative and interpretive effort of interviewer and respondent to bridge the differences between their interpretations of a certain ‘problem’ in a process of interaction (see Chapter 2). Prior knowledge and practical knowledge enter into a dialogue (see Figure 2.1). The success of this bridging exercise – i.e. the discursive-dialogic reconstruction of a

problem – depends first of all on aspects described under the label of the principle of problem centring. Interviewers use prior knowledge, which includes shares of scientific as well as everyday knowledge, as a sensitising framework for the formulation of appropriate questions. At the same time they need to consider the dangers inherent in mutual insinuations of meaning resulting from the interconnection of the two worlds of meaning, which, in the course of the conversation, may become less and less distinguishable. The fact that the PCI establishes the respondent as the subject of a meaning universe has two important consequences: it constitutes a critical distance of interviewers to their own meanings; and it invites respondents to control and, if necessary, to correct the interviewers’ process of understanding and comprehension (keyword: counterevidence). b) The Interviewer's Double Role of Expert and Learner

The PCI's main task of bridging two worlds of knowledge defines the interviewers’ role in a twofold way: the interviewers’ status as experts is rooted in their prior knowledge; their status as learners is rooted in their ambition to reflect this knowledge self-critically and revise it by confronting the life-world of respondents. In the course of the interview they develop pre-interpretations, by using sensitizing concepts, and challenge them in a knowledge iteration of inductive and deductive steps following the principle of process orientation. c) Practising the Interviewer's Double Role of Active Contributor and Passive Listener

In their roles as experts and learners, interviewers are active by shaping the dialogue in a suitable way (principle of object orientation); they are passive by focusing their attention on the respondents’ relevancies, i.e. they literally ‘sit back’ (‘zurücklehnen’) and listen. d) Facilitating Narratives and Comprehension

Interviewers facilitate and support narrative sequences by means of narration-generating and detailing questions; this goes hand in hand with a more passive approach in an attitude of listening. This explorative part is complemented by comprehension-generating communication strategies with a more active approach in an attitude of understanding. Follow-up questions of comprehension are based on prior knowledge as well as on ad-hoc ideas and pre-interpretations concerning the respondents’ accounts. These questions result in more or less extensive phases of dialogic conversation; they depend on the need for clarification, which results from the complexity of the issue and the competence of the respondent to articulate and structure the Page 4 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

issue. In extreme cases, when respondents are not talkative or have limited communicative capabilities, the dialogic form of interviewing is hardly even accompanied by a narrative flow (flexibility of the PCI) (see Box

4.4). e) The Respondent's Double Role as Subject and Object

Following the basic principles of interviewing, described above in a) to d), interviewers try to establish the respondents as research subjects and abolish their role as objects of inquiry that they would have in other conventional approaches. There, their status is that of ‘functional’ informants providing brief answers to equally brief questions. Respondents would be exposed to an external sequence of questions resulting from the authority of the interview guide (Pitfall 4.1); and they would have to involve themselves in a one-sided attempt to interpret the formulation of the questions and translate them into their world of meaning. Such approaches fail to bridge social scientific and everyday knowledge; they tie respondents down to the role of informants who provide knowledge, which is then subjected to the interpretational sovereignty of science. Finally, the implications of these interpretations and their transfer into the world of practice are beyond the respondents’ control. Needless to say, this aspect concerns the ethics of science as much as the basic principles of social co-existence and politics.

A Classification of PCI Pitfalls Together these principles constitute the basic attitude of the interviewer who has to decide which roles have priority in which phase of the interview and how they should be connected. The pitfalls of problem-centred

interviewing result from disregarding some aspects of these basic principles and reflect the domination of one of the contrastive pairs. In other words, the dialectic tensions between these aspects – scientific/ practical, expert/learner, active/passive, generating narratives/comprehension – enter into conflict. In the end, interviewers leave the path of the PCI's well-informed traveller and exaggerate the role of either miner (I. – see below) or traveller (II. – see below) (Kvale and Brinkmann, 2009: 48) (see Chapter 1). I. Interviewers Remain Trapped within Their World of Meaning and Do Not Use Their Knowledge in a Sensitising Way

In this first category of interviewing errors, interviewers put a one-sided emphasis on their role as miner and ‘undermine’ the status of respondents as subjects. They dominate the conversation, focus mostly on their research knowledge, and cannot overcome their preconceptions. This problematic approach is widespread when training in methodology and methods of qualitative research is inadequate. (a) The most extreme form of exaggerating the expert role consists in the direct use of research questions in the interview (see Pitfalls 3.1 and 6.10). The interviewer under-estimates the difference between prior knowledge and the practical knowledge of the respondent and presupposes that both understand the questions (which directly reflect the underlying research interest) in the same way. The respondent's lack of understanding results in silence or in responses that are thematically incoherent. The interviewer essentially Page 5 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

neglects the everyday life-world and language of the respondent and fails to appreciate the role as learner. The inaccurate translation of research questions into interview questions can often be found in projects of students which (still) tend to stick too closely to their research questions. A few pilot interviews can help to sort out inappropriate questions that are too abstract because they use scientific terminology. (b) The strict execution of the interview guide constitutes another form of exaggerating the role of the miner. The domination of the question–answer pattern indicates that the interviewer's communication is restricted to conventional forms of collecting information. Short questions and the expected short answers atomise the topic and detach it from the context. While the interviewer's constant interventions drive the dialogue forward, the responses consist mostly of isolated flashlights on the issue. The more subtle aspects of the respondent's perspective stay hidden behind the ritualistic interaction. Such an interviewing style either forces respondents to speculate about the meaning and background of the questions, or they don't care about it and simply translate the terminology into their own everyday context without additional thoughts. There is little chance that respondents will produce coherent accounts in a reflexive perspective, gradually remembering past events. Each new question (from the interview guide) changes the topic and discontinues previous thoughts. The interviewer/researcher will collect a sequence of concise answers, at worst consisting only of ‘yes’ and ‘no’. The actual meanings cannot come to the fore because the communication obstructs original contributions of the respondent. Owing to the poor quality of the answers, the interviewer is unable to understand that the questions are actually incomprehensible. As her interpretational sovereignty is uncontested, she sees no reason to change them. The interviewer essentially immunises her perspective: she feeds her prior knowledge and pre-interpretations into a communicative circle that can only confirm them following the logic of subsumption. Three main reasons for the domination of the question–answer pattern can be distinguished.

Lack of trust in the competence of the respondent to develop an original narrative. The interviewer has the good intention to invite the respondent to unfold her perspective and to tell a story. Yet this intention is contradicted by her actual communication. The interviewer may have doubts about the competence of the respondents to accommodate an actual narrative stimulus. She may, for instance, believe that the respondent needs to be supported in activating her memory or in identifying an appropriate starting point for her account. Therefore, she complements her narrative stimulus too early with additional and detailing questions. The contradictory combination of a narrative stimulus and the concerned communication style confuses the respondent regarding her role in the interview conversation and reinforces the pattern (see Pitfall 4.1).

Stressful situation of the interview. Another reason for the exaggeration of the active role is when interviewers feel that they can only cope with the pressure of the situation by dominating the conversation. This is common among inexperienced and poorly prepared interviewers who have difficulties establishing a temporary relationship of cooperation with an unfamiliar person. They finally assume the sole responsibility for the situation and deal with the related stress by activating proxies of communicative security: they use keywords and formulations from the interview guide or rush from one topic to another. If this style of communication is repeated, the insecurity of the interviewer is reinforced and empathy for the respondent gradually becomes Page 6 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

impossible. The interviewer keeps pushing the conversation forward, neglects non-verbal signals, and pauses or interrupts the flow of thought. This style fails to recognise the importance of the interviewer's accepting and empathic attitude for the gradual development of a productive working relationship and an atmosphere of trust.

Exaggerated commitment and interest of the interviewer. The self-confidence of interviewers as experts of the research issue and a certain impatience to find answers to their questions can result in typical mistakes related to rushing through the interview guide. They include the overloading of questions in terms of length, issues and complexity, or the rapid shift between topics and the exaggeration of the expert role (see Pitfalls

6.5, 6.6 and 6.7). (c) The supposed identity or overlap of understandings between interviewer and respondent is another reason for the domination of the question–answer pattern. In this case, it is not established by the interviewer but it is the result of mutual insinuations of meaning that apparently bridge prior and everyday knowledge. Interviewer and respondent believe that they already understand each other through certain suggestions, and mutually confirm this apparent agreement non-verbally or by means of short questions and remarks. Yet the understanding is superficial and they do not explicate each other's meanings, which are thus also not registered for analysis and interpretation (see Pitfall 6.8). (d) Some interviewers cannot overcome their strong preconceptions based on their prior knowledge, which can also include everyday knowledge and private opinions (see Pitfall 6.4). They remain entangled in their universe of meanings and consider the respondent's opinions as a deviation from their own definition of

normality. These interviewers exaggerate the expert role by constructing a sort of objective reality, for instance on the basis of statistical and empirical evidence serving as a benchmark for their assessments. The respondent's ‘deviating’ opinions appear strange to them and might cause feelings of antipathy. Interviewers may lose interest in exploring the underlying subjective meaning altogether and forget that the unfamiliarity of the respondent's perspective was the actual reason for exploring it in the first place. Overcoming the passive rejection of the respondents’ views by confronting them does not solve the basic problem. Interviewers that criticise the respondents’ position abandon the accepting attitude that is so important for qualitative interviewing. The interpretational sovereignty discussed above turns here into a form of normative sovereignty that blocks any kind of understanding. Expert interviews are particularly susceptible to this mistake when peer interviewers are used. They are at risk of transcending their legitimate role as coexperts when they pretend to know better and start competing with the expert respondent. (e) Knowledge can have the negative effect of indifference and ignorance. The apparent paradox that an increase in knowledge, also due to the number of conducted interviews, can go hand in hand with a decreasing number of follow-up questions is called interviewer saturation (Heinz et al., 1986: 191). Prior knowledge and pre-interpretations turn into preconceptions, and certain interpretations are taken for granted. The interviewer stops exploring the respondent's actual position and can no longer validate her interpretations on the basis of authentic material. Page 7 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd II. Interviewers Lose Distance to the Respondents’ Meanings and Disregard Their Prior Knowledge

The second category of interviewing errors results from misunderstanding the principle of openness as an invitation to be absorbed by the perspective of respondents as subjects of the interview. Interviewers believe they understand the meanings of the respondents only by completely disregarding their prior knowledge and assuming the role of a traveller. (a) Interviewers lose the necessary distance to the role of the respondents as subjects of the interview and develop an attitude of partisanship (see Pitfall 6.2). Furthermore, this interviewing error results from the relationship between research and everyday knowledge and from the question as to whether and how the position of the respondent corresponds to whatever is defined as normality by the interviewer. Yet the interviewer assumes a position that represents the extreme opposite to that described before: she idealises the respondent's perspective and thus loses any critical distance. This positive recognition can range from explicit sympathy to solidarity with the problems of the social or perhaps marginal status of the respondent. The ethnographic literature uses the notion of going native to describe phenomena related to over-rapport (Ballinger, 2008). Social or political commitment is one of the motives for assuming such a problematic attitude of solidarity towards the respondent; another is the attempt to facilitate the interview by being recognised by the respondent. The interviewer performs a problematic transition from the desirable understanding of the perspective of the respondents to a compassionate understanding of the respondents’ background and their possible problems. This transition undermines the distance to the respondents’ life-world that is necessary for a problem-centred approach. Even if respondents welcome the interviewer's positive partisanship and reward it with trust and a good interviewing atmosphere, it may yield problematic consequences. • Total agreement can bring an early end to extensive reflection and threaten the original purpose of

differentiated exploration of the respondent's meaning universe. • Such a form of solidarity can lead respondents into the trap of social desirability, in this case, towards interviewers. Respondents perceive sympathetic interviewers essentially as persons that judge. In order to appear in a positive light they try to anticipate requirements for positive judgement and avoid negative appraisal, for instance by omitting certain issues or by adapting their accounts to the expected criteria of the interviewers. Respondents may suppress inconsistencies and ambivalences in decisions, orientations and action. In the end, such a bias towards partisanship is to the disadvantage of the openness of the interview and the validity of the material (see Pitfall 6.2). (b) Interviewers are concerned that interventions could impact negatively on the accounts and overemphasise their role as travellers. They believe that restraining themselves methodically – i.e. disregarding

one's prior knowledge and giving priority to narration-generating questions – could bring them closer to the respondents’ original perspectives. This attitude reflects a regression into the normative-deductive approach: the role of the learner and the principle of openness are mistaken for the ideal of the neutral observer that supposedly characterises qualitative research. Although the complete neglect of prior knowledge is Page 8 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

epistemologically impossible, the interviewer tries to assume the role of the data collector that, as a kind of tabula rasa, registers accounts without influencing them. The selection of narration-generating questions remains arbitrary or is done in an erratic way. The lack of comprehension-generating questions entails the following problems, which thwart the advantage and purpose of qualitative interviews to produce differentiated insights into aspects of the subjective world of respondents: • It leaves knowledge gaps in the overall picture, and the flow of the conversation may not be understandable. • Respondents are not confronted with ad-hoc ideas based on forms of prior knowledge or preinterpretations developed during the interview, and the understanding of their meaning world cannot be expanded. • Without substantive feedback, respondents cannot judge whether the interviewer is really following their account. Non-verbal signs of attention and conventional superficial feedback indicators like ‘mhm’ or ‘aha’ cannot replace direct follow-up questions. Respondents may interpret such superficial signals as indications of agreement and understanding of what has been said, and be confused when interviewers suddenly start asking follow-up questions in a separate, second phase of the interview (like in narrative interviews). The artificial separation of the interactive unity of account and feedback characterising everyday communication could disturb the respondents’ trust and they may wonder whether they are being taken seriously. (c) Respondents usually reflect upon the issue for the first time and are not able to structure their opinion in a systematic, logical and detailed way; and they certainly do not follow didactic principles to introduce their world of meanings to the interviewer. Interviewers are overwhelmed by the complexity of the account, rely on the flow of the narrative, and miss out on asking follow-up questions. Especially beginners have difficulties dealing with the multi-layered nature of the issues expressed in respondents’ situational actions, decisions and orientations. The requirement of being, at the same time, an empathic listener who ‘sits back’, follows and silently pre-interprets the unfolding account, and an active participant in the conversation who supports the respondent's reconstructions by applying PCI communication strategies, may simply be too much. In view of this challenge interviewers may forget to make use of the following tools: • The communication strategy of mirroring can particularly help to structure and understand issues that have already been discussed. It slows the communication down and supports the control of the interpretation. • Interviewers should take notes to support their memory and to register the multi-layered nature of articulations, turns, redundancies and corrections as well as particularities of language and expression. At the beginning of the interview it is not clear how long the accounts will be – taking notes facilitates concentration and prepares follow-up. • Heuristic models or visualising tools like a calendar in the context of life-course research (Bird et al.,

Page 9 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

2000) or the ARB-scheme (see section 6.1) can be used to structure and review the completeness of the exploration of episodes, situations or biographical sequences.

Concluding Remarks At the end of the book the reader can now – similar to during the interview – sit back and account for what she has actually learned, deliberate over what to think about it all, and wonder how to apply it in her own research. The metaphor of well-informed travelling introduced at the beginning of the book may seem plausible in the context of a journey of discovery through a mediaeval town. Yet it may be regarded with scepticism when it comes to translating this approach into research by means of PCIs. Well-informed interviewing may seem too big an effort in terms of time and economic resources. In particular, the discussion of possible mistakes in this chapter may have caused doubts regarding the ability to conform to the standards of the PCI. In order to dispel such doubts we want to conclude with a few encouraging remarks regarding a) economical aspects, and b) the skill of problem-centred interviewing. (a) People always project and realise actions, and they make them meaningful through their purposes and achievements. These processes of meaning constitution refer to different conditions and situations because action is always contextual and embedded within certain living circumstances. Usually, these contexts and interpretations, as well as the assessment of the results of action, change in the course of life. That's why qualitative research is an ambitious enterprise as it tries to grasp this changeable process of meaning production in the perspective of research subjects. The challenge is further increased when the researchers/ interviewers realise why interviewing requires a self-reflexive methodical procedure when they are ‘learning from strangers’ (Weiss, 1995): their own interpretations are constructions that can influence how they reconstruct the meaning/interpretations of the respondents in the course of the interview. Our systematic explication of the basic principles of the PCI and their practical reflection should be understood as a contribution to the development and improvement of standards of qualitative interviewing. This is all the more important as there are only a few elaborate qualitative interviewing methods and because, at least in Germany, a culture of ‘patchwork interviews’ has spread. In an eclectic way, ever new techniques are designed and labelled, which hardly refer to the standards of well-established techniques (see Mey, 2005). It is true that the principle of object orientation suggests the adaption of the method to specific research questions, but this does not require a completely different technique. Although the practical application of the PCI always has to be adapted to the research subjects, these adjustments often do not concern the interview directly but rather its preparatory phase. For instance, researchers in STUDY A prepared themselves through participant observation in order to facilitate appropriate communication with young people. Also, group discussions can prepare PCIs, for example, when they deal with peculiarities of language among respondents. Small research projects or student projects often do without methodical training. This is understandable, as is the fact that the possibilities for exploring an issue in the course of a single interview are not always Page 10 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

exploited because the expected pile of data could not be processed. Yet it is not understandable that the choice of such a simplified and apparently ‘pragmatic’ or ‘realistic’ approach is then attributed to the high quality requirements of a technique like the PCI and not to the conditions of research. PCIs can indeed produce a considerable amount of data, yet calling the method uneconomical is a non-scientific argument: it inappropriately relativises all efforts of searching for a solution for the manifold practical problems related to collecting data in a methodologically appropriate way. The quality of analysis, interpretation and findings depend on the quality of the interviews. If lacking time or human resources one should not simplify the procedure but reconsider the sample size. We recommend following the Grounded Theory approach of theoretical sampling (see section 3.5) where sampling depends on the progress of theory development – new evidence is collected only until the moment of theoretical saturation, i.e. until generalisable findings are produced. This principle can also be applied to data already available in order to reduce the amount of material that is actually analysed. Another alternative that can reduce the amount of work invested in data collection and transcription is rather new but innovative: secondary analyses of data from studies that are already concluded and available in qualitative data archives (Heaton, 2004; Medjedović and Witzel, 2007, 2010). Usually this option is not even considered, and it may turn out to be unnecessary to collect new data, or that a small-scale investigation is sufficient to complement available data. (b) The analysis of mistakes is an important contribution to quality management in interviews. Making mistakes, as such, is no reason for self-doubts regarding one's competences or the requirements of the PCI; and from our practical experiences we can tell that interviewing errors are unavoidable. The important thing is that skill builds upon learning from mistakes. The discussion of mistakes and the related examples should first of all stimulate the self-reflexivity of the researcher/interviewer (see Mruck and Breuer, 2003). In the end, the goal of research is the development of novel insights as well as the revision of prior knowledge and the overcoming of preconceptions. Following Reichertz (2000: paragraph 69) referring to the principle of knowledge sociology – qualitative research is about a ‘systematic and organised production of doubt (in each phase of the research process) and the thereby achieved eradication of mistakes’. Our classification of mistakes should not be understood as a ‘not-to-do list’ which has to be learned by heart in order to be capable of establishing a fruitful interviewing dialogue along PCI principles. It is not about wagging a finger but about explaining how typical PCI mistakes happen. We wanted to show that deviations from key principles and their ‘outcomes’ can be brought into a systematic relationship. This necessarily negative angle of understanding mistakes should strengthen the confidence of the researcher/interviewer rather than undermine it. The seemingly contradictory procedure of using one's prior knowledge, but at the same time keeping it open, is technically demanding and can thus result in the two ideal-typical mistakes of interviewers: they use their prior knowledge in a circular way and remain trapped in their world of meaning; or they consider it dispensable, suspend it and get lost within the respondent's world of meaning. Mistakes are best understood and avoided by means of teamwork and supervision; they are the best means of developing the skill of dealing with mistakes. Qualitative research is hardly manageable without a research team because of the Page 11 of 12

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

extensive data analysis that is usually involved. The independent and critical contributions of team members can also improve self-reflexion. At least in the German context, however, this idea is contrasted by a research culture that turns junior researchers into lone fighters. The great demand for both research workshops and self-organised interpretive communities refers to shortcomings in methods training and the need for counter-balancing isolated research work with forms of exchange and intersubjective methodical control and reflection. The discussion about the possibilities of secondary qualitative data analyses indicates that in terms of cooperation there is still a long way ahead – making one's original data available for colleagues continues to be a problematic issue. The collective ‘production of doubt’ and of alternative readings for the sake of enhancing the quality of interpretations is in its infancy. The qualitative research culture is still characterised by a certain ‘fear of exposure’ regarding the disclosure of possible weaknesses of one's data that hampers progress of insight (Corti, 2000; Medjedović, 2008). The fact that the PCI is rather robust regarding mistakes – at least if interviewers succeed in assuming the role of active contributors and listeners – should reassure sceptics. As soon as respondents perceive the accepting and curious attitude of the interviewer as authentic, they will usually insist on the chance to unfold their perspective. And once their perspective is established as the focus of the interviewer, they will be less irritated by interviewing mistakes. http://dx.doi.org/10.4135/9781446288030.n7

Page 12 of 12

The Problem-Centred Interview: Principles and Practice

Appendices In: The Problem-Centred Interview: Principles and Practice

By: Andreas Witzel & Herwig Reiter Pub. Date: 2013 Access Date: December 25, 2018 Publishing Company: SAGE Publications Ltd City: London Print ISBN: 9781849201001 Online ISBN: 9781446288030 DOI: http://dx.doi.org/10.4135/9781446288030 Print pages: 186-191 © 2012 SAGE Publications Ltd All Rights Reserved. This PDF has been generated from SAGE Research Methods. Please note that the pagination of the online version will vary from the pagination of the print book.

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Appendices Appendix 1 - Interview Guide of Wave I of Study A (1989/1990) 1. Introduction • You did learn the profession of … How did this actually come about? Please tell me about it.

2. Pre-Professional Situation 2.1 Occupational Options

• When did you think about it, when decide? • Any plans before or alongside it (at school, in terms of profession, preferred occupation)? What was important for the decision: future chances regarding career and profession, possibilities for further education, safe job, expectable income, work content, a professional education at all? • Did you talk about it with parents, friends and teachers? • Internship. • Career counselling (employment agency). 2.2 Search for Apprenticeship Training

• How many applications did you submit, where, why there (occupation, company, number of job offers)? How did you search for a job (employment agency, newspapers, telephone book)? Support by parents, acquaintances? • Apprenticeship offers, graduation, gender, un/favourable for job search? • Why did you accept the apprenticeship? Why did the company accept you? • Have the expectations been fulfilled (longer search, more rejections than expected, handling rejections, test results)? What experiences did friends, classmates have?

3. Apprenticeship Training Place • Now I would like to talk about your education … Company, number of trainees, superiors. • Good and bad sides of professional training: a) company, b) vocational school. • In case it was bad, did you think about discontinuation? • Explanation of drawbacks. • Potentials for change. • Own aspirations to work (contents, circumstances, money). Page 2 of 8

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

• Aspirations for organisation: appraisal and dealing with it (colleagues). • Meaning of the final examination (achievement, private life). • Meaning of the reorganisation of the job. • Would you like to have this education again (profession, company)? 3.1 Transitory Situation

• Half a year ago, when you filled in our questionnaire, your training company (did not) ask(ed) you to stay. How did this come about? Was it clear from the beginning who would be able to stay, or did they keep it open until the end of the apprenticeship? • When did you hear about it? Did you expect it? • Was it the same with other trainees? • Criteria for staying (explanation). • Did you want to stay? Why? • What have you done to be taken on? • Other plans/how pursued in practice? • Relationship to trainees in the process? 3.2 in Case a Job Is Offered

• Conditions for the job offer (limitation, participation in the choice of the workplace). 3.3 in Case a Job Is Not Offered

• Consequences for further occupational plans. • What did you do to find the workplace? 3.4 in Case of Graduates from Training Measures

• It was clear for you from the beginning that you would need to find a job after graduating from your apprenticeship training. Did you have difficulties with this perspective? • How did you deal with this? • Evaluation of the vocational education: i.e. what you learnt, as precondition for job search. 3.5 Unemployment

• Did you expect to become unemployed? • Why did you become unemployed (school education, region, profession, gender)? What about colleagues, friends, family? • Burdens (material, psychological). • Further professional plans (reduction of aspirations(?), time perspectives, material security versus work contents). Page 3 of 8

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

• What did you do in order to get a job? • What do you intend to do in case you are still unemployed in half a year? • Meaning for private plans/leisure time interests. 3.6 Armed Forces/Alternative Service

• When do you expect to do your military/alternative service? • Meaning for professional plans (coercion, bridging time, strategy) • Disadvantages in the company (social comparison)? 3.7 in Case a Job Is Offered in Another Company

• How did you get your job? • How many applications did you write for which jobs (semi-skilled, non-skilled activities)? • Where did you apply (regional, national, which companies)? Why there? • How did you search for a job (labour exchange, relations, newspapers)? Job offer, education un/ favourable? Why? • Why did you decide in favour of this position? • Why did the company take you? • Have your expectations been fulfilled (explanation of success/failure)? What were the experiences of classmates, colleagues?

4. Workplace • Usefulness of the professional education (eventually further education necessary). • Good/bad sides of the job. • In case of bad sides: Can you change it or do you need to get used to it? • What does the company expect? • Does the work correspond to your expectations (work content, working conditions, money, opportunities for advancement, accomplishment)? • Are private/leisure interests compatible with occupation (working hours, money, exhaustion, conflicts with partner)? What about colleagues? • What do you miss in your current workplace?

5. Future Situation • Where do you go from here in the next few years (detachment from parental home, family formation, financial aspirations, career, change of job and company)? What do you want to achieve, and until when? • What have you done/do you plan to do in order to reach your goals? Page 4 of 8

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

• On what does the realisation of these goals depend (personal features/qualities/activities, labour market)? Meaning of work/leisure? • Graduation, occupation, company ……… Pay attention to contradictions!

Appendix 2 - Interview Guide for Judges from Study B Establishing Contact As you know, for several years there has been a nationwide debate about the costs of custodianship law that was introduced in 1992. Against the background of this debate and on behalf of the Ministry of Justice of Lower Saxony our project is dedicated to the investigation of the difficulties and possibilities of this law. Our approach is different from previous research that studied the increasing costs of custodianship on the basis of comprehensive questionnaires. Our emphasis is on the experiences of all institutions involved with custodianship law. On the basis of the different practical experiences of judges, judicial officers, custodians and custodianship officers, we first want to investigate the practicability of the regulation. Only then, in a second step, do we want to discuss the costs that it involves. In other words, our aim is not only to identify the reasons for the increasing costs that need to be reduced, but also to find out which costs are actually necessary and reasonable. In this context I want to consult you as an expert of your area. This is not an interview that follows the normal question–answer pattern. Instead, I want to give you the opportunity to talk extensively about your practical experiences and to develop your views and ideas concerning this matter. I will ask follow-up questions only in case I do not understand something or when I want to know more about it. In order to better follow your account I would like to record our conversation. Everything you say will be anonymous.

1. Stages of the Practical Work and Related Topics Opening Question

I think the best way for us to understand your view is to learn about your concrete work as a custodianship judge, which means on the basis of the process of your work involved in a typical custodianship procedure. How does a case land on your table, or: how does a case actually start? And how does it then continue? 1.1 Needs for Custodianship

• How does the custodianship procedure start? • Who normally files the application? • What reasons do the applicants have? • What is the understanding of custodianship among the applicants? Page 5 of 8

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd 1.2 Necessity/Proportionality

• What is your opinion about the reasons of the applicants? • How do you deal with them? • When, or for who, is custodianship necessary in your opinion? • What are the requirements for this necessity (proportionality)? When are they fulfilled? • What do you think about guardianship (Vorsorgevollmacht) as an alternative? 1.3 Field of Activity

• How do you define the fields of activity? • Who is commissioned to provide medical opinion? • How do you deal with the report, what do you learn from it? • How do you assess the report in terms of informing the assignments of the fields of activity? 1.4 Selection of a Custodian

• How is the custodian selected? • Requests of the person concerned. • Requests of family members. • Are the requests of the funding parties relevant: state or person concerned? • Requirements from the medical report. • Social reports of the custodianship office. 1.5 Review/Alterations/Implementation Regarding Changes

• Once custodianship is established, how are you re-involved in the case? • Perception of the judges’ own role: do they consider themselves ‘initial decision makers’ or ‘committed companions’ of the person concerned? • Medical provision/preventive measures/other special circumstances. • Change of fields of activity: extension/reduction. • Change between professional and voluntary custodian, and related criteria? • When and why is custodianship revoked? How does it happen? General pattern of asking: • Typical versus untypical cases. • Obstructive versus conductive conditions. • Useful/reasonable versus less useful/reasonable regulations. • Previous versus current situation. Note: Page 6 of 8

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

Do not explore cost-increasing or cost-reducing measures at this point in case respondents start talking about them spontaneously in the frame of talking about their experiences! Take a note and follow up in the context of the next topic. If necessary, indicate that you will come back to this issue later. The same applies to the topic of less useful/reasonable rules (that we expect to be cost-relevant); take a note and come back to it in the frame of a more extensive exploration in the context of the discussion of alternatives.

2. Costs and Alternatives Opening Question

• You already talked about more and less useful rules on the basis of your practical experience; and to some extent you already addressed the question of costs. In your opinion, what are the main reasons for the costs? 2.1 Disproportionate Increase in Costs of Custodianship Per Case

• Where do the costs mainly occur? • Where did they increase? • Will they further increase? • Where can they be reduced and where should they not be reduced? Note: All questions under 2.1 should be explored systematically along the stages of the practical work (see point 1). 2.2 Increase in Cases in Custody

• Is your own area of responsibility affected by this increase? • Are there moments or occasions for a particularly rapid increase? • Where do you see possible reasons? • Can you think of alternatives that are both reasonable and cost-saving?

3. Core of Custodianship Law/Further Alternatives Opening Question

• We talked about the practical aspects of the custodianship procedure, about what you consider reasonable and less reasonable rules, about obstructive and conductive conditions, and about issues of cost. I would now be interested in your conclusions concerning the core of custodianship law. • Is there a necessary core and ‘essence’ of custodianship law that is worth preserving? • Does this need to be handled in a different way? • Is the current legal situation appropriate for it? Page 7 of 8

The Problem-Centred Interview: Principles and Practice

SAGE

SAGE Research Methods

2012 SAGE Publications, Ltd

• Should we consider the system of custodianship law itself as the core of the problem? • Do you think that the system of custodianship law is altogether worth preserving? • Are there aspects of custodianship law that you would definitely abolish? Why? How would custodianship law need to be changed then? • Do you think that custodianship law should be abolished altogether? Why? What should be established instead? http://dx.doi.org/10.4135/9781446288030.n8

Page 8 of 8

The Problem-Centred Interview: Principles and Practice