1 Introduction

Case study research is an important part of qualitative research in the field of management (Cassell and Symon 2004; Flick 2014; Langley and Abdallah 2011; Patton and Appelbaum 2003) and published articles on this topic provide some groundbreaking insights in management research (e.g. Bansal 2005; Burgelman 1983a; Chandler 1962; Danneels 2010; Graebner 2004; Hoffmann 2007; Moschieri 2011; Pettigrew 1973). As a result of its rising relevance, there are various methodical contributions available that emphasize important aspects of conducting case study research. Some scholars focus on methodological foundations (Bryman 2008a, b; Denzin and Lincoln 2000; Guba and Lincoln 1998), others work on methods for data analysis (e.g. Bryant and Charmaz 2007; Charmaz 2014; Corbin and Strauss 2008; Eisenhardt 1989; Eisenhardt and Graebner 2007) or on the topic of evaluation (Flick 2007; Kirk and Miller 1986; Lincoln and Guba 1985; Silverman 2004). Despite what methodical contributions have achieved during the last decades, qualitative methods are a subject to constant criticism. Apart from other issues, the subjectivity of interpretations and the exclusion of theory are the main points of criticism (Holweg 2012). While there is more or less no doubt today that a naïve empiricism research approach starting with an “empty mind” is not viable, an explication of how theory should be used in qualitative case study research is still underemphasized in literature. The lack of methodical contributions that focus on integration of theory or other forms of pre-existing knowledge has important implications for knowledge creation in management research. First, the general uncertainty about how extensively theory could or should be integrated leads to significant differences between research designs even between these qualitative case study designs that basically follow the same research method (e.g. Martin and Eisenhardt 2010 vs. Shaffer and Hillman 2000). This is unsatisfactory from the perspective of methodical rigor and bears the risk that the trenches between highly standardized quantitative and qualitative research will become even deeper. Second, qualitative case study research cannot take advantage of its full potential as long as pre-existing knowledge is refused under the guise of ‘openness’. The purpose of this article is to explicate how theory could be used during the entire qualitative case study research process without compromising openness. The paper proceeds as follows. First, we discuss types of qualitative case study research and underlying paradigms (Sect. 2). We then present the role of theory during key stages of case study research (Sect. 3). In our concluding outlook (Sect. 4), we discuss how our article contributes to qualitative management research.

2 Qualitative case study research and theory

The term case study research refers to a research design that focuses on a precise description or reconstruction of a phenomenon within a real-life context (Yin 2014) emphasizing the view of the subject. Qualitative case study research can be designed as a single case, in which only one case, for example an individual, a social community or an organization, is thoroughly examined (Flick 2014). Contrary to single cases, multiple case studies base on designs that draw on a logic of replication (Yin 2014) and analyze a series of cases to elaborate on theoretical relationships between them (Eisenhardt 1989). Similar to other forms of empirical research, qualitative case studies are oriented towards three fundamental designs: descriptive (precise description of a phenomenon), explorative (drawing closely to phenomenon) or explanative designs (explaining causal relationship between variables) (Mayring 2010). Basically, qualitative case study research can be classified according to underlying epistemological assumptions (‘paradigms’). Guba and Lincoln (1998) distinguish between four paradigms: positivism, post-positivism, critical theory and constructivism. For the purpose of this paper, we are going to describe briefly the contrasting perspectives of (post-) positivism and constructivism.

  1. 1.

    Positivism as an epistemological program argues that science should concentrate on studying observable facts accessible by the senses (Bryman 2008b). Positivism supports realistic ontology and assumes that there is an ‘external reality’ separate from its description (Guba and Lincoln 1998). It is based on the ideal of objectivity and neutrality. Research should test theories and collect facts that provide the basis for laws, whereby replicable findings are probably ‘true’. Post-positivism represents thinking after positivism and follows a critical realist ontology: the external reality is assumed to be only imperfectly apprehend able and knowledge is conjectural. Both approaches regard replicated findings as probably true or as closer to the truth (truth likeness).

  2. 2.

    Although there are a variety of constructivistic approaches, the main statement of constructivism is that every form of knowledge about reality is constructed by individuals that act toward things on the basis of meanings that the things have for them. These meanings are derived from social interactions and are handled in as well as modified through an interpretative process (Blumer 1969). Social reality therefore is multiple, processual, and constructed. As a consequence, the focal point of research is dedicated to the reconstruction of different ways in which individuals ascribe meaning in a certain context (Myers 2013). Sometimes constructivistic approaches are linked to Weber’s ‘Verstehen’ approach. Understanding in social science is inherently different from positivistic explanation and therefore not nomothetic but idiographic methods are applied. According to constructivistic paradigm, findings are created during the research process (Guba and Lincoln 1998). Furthermore, some constructivistic positions reject generalization as a goal of research (Denzin 1983).

Epistemological assumptions are to some extent contradicting because they follow distinct claims regarding the existence of a subject-independent reality. Although qualitative research was developed to a large extent in contrast to positivistic thinking (Flick 2014), even scholars who belief in a realistic ontology can apply qualitative case study research (Morgan 2007; Haase 2010). Actually, post-positivistic thinking has informed qualitative research, too. Thus, according to underlying epistemological views, the following types of qualitative case study research can be distinguished (see Myers 2013 for further types):

  1. 1.

    Interpretative (constructivistic) case study research is based on the assumption that (social) reality is socially constructed and attempts to describe, reconstruct, and understand phenomena by referring to the meanings that informants attribute to them. Interpretative qualitative research is not restricted to descriptive designs but can also include explanative, experimental designs (e.g. Kleining 1986). The Grounded Theory method by Charmaz or the ‘Gioia Method’ are examples of interpretative case study designs (Charmaz 2014; Langley and Abdallah 2011).

  2. 2.

    In contrast, positivistic qualitative case studies are based on the (post-) positivistic paradigm and aim to test or refine propositions in order to develop a theory, which can be generalized across different settings and is subject of some kind of standardization and formalization (e.g. frequently following a precise research plan). (Post-) positivistic qualitative designs base on a cumulative knowledge ideal and strive for falsification and analytical generalization. Glaser’s version of Grounded Theory is an example of being rooted in naïve empiricism/positivism (Glaser 2005) and the ‘Eisenhardt Method’ can be seen as a post-positivistic case study design (Eisenhardt 1989).

Despite a wide range of methodical contributions in the field of qualitative management research, different perspectives on the role of theory exist. On the one hand, particularly in interpretative case study research, theory plays only a minor role. The perspective of research here is one of an open inductive process, which has to be structured in a way that new knowledge can emerge without being constrained by existing knowledge (Denzin and Lincoln 2000). Qualitative research is frequently assigned to the context of discovery (Reichenbach 1983), in which theoretical knowledge is expendable or even obstructive (Burgelman 2011). Even in some versions of positivistic case study research, theory application is delayed until the analysis is nearly completed. Glaser insists that grounded theory should stay ‘uncontaminated’ by preconceptions (Glaser 2012). However, not only Glaser holds this view but also some other grounded theorists echoed this notion currently (Holton 2007; Nathaniel 2006). Eisenhardt and Graebner (2007) as proponents of a post positivist design claim that for phenomenon-driven research questions only broad hints to previous knowledge are required.

On the other hand, the idea of collecting and analyzing data without use of theory has been heavily criticized. There is an important debate about viability and limitations of openness. Openness in this context is defined as ‘responsiveness’ towards the respondent’s meaning. Scholars refer to the fact that openness and theory-less-ness are different claims (Siggelkow 2007) and that researchers can never approach their project tabula rasa (Hanson 1965). Even interpretative epistemological stances are not contradictory to prior theoretical knowledge (Blumer 1940; Kelle 2005; Kelle and Erzberger 1999). Charmaz (2014: 13) points out that grounded theory can be used “without endorsing mid-century assumptions of an objective external reality”. From this perspective, qualitative case study research is infiltrated by theory just like all empirical research. Interestingly, the relevance of a theoretical grounding was basically echoed in recently formulated requirements of high-ranked general management journals (Bergh et al. 2006; Judge et al. 2007; Piekkari et al. 2009; Suddaby 2006).

Although it is acknowledged that theory plays a pivotal and uncontroversial role in quantitative research, hardly any methodological paper asks pro-actively the question, at what stage and how extant theoretical knowledge could (or even should) be integrated during the case study research process. Table 1 summarizes rather well-known examples of outlines for carrying out qualitative research projects that refer to the problem of theory integration. It shows that different authors assign theoretical knowledge to differing stages of the qualitative research process. Furthermore, the integration of theory remains rather dispersed, is not always disclosed as well as not specified in encompassing canons or methodological frameworks.

Table 1 “Boilerplates” for conducting qualitative research
Full size table

The aim of this paper is to present a methodological framework that integrates and extends the different assumptions of how to use theory during the entire qualitative research process. In general, by aligning different (sets of) variables and illustrating possible relationships frameworks help to structure heterogeneous or new research fields and can generate a heuristic potential for theoretical explanations (Kirsch et al. 2007: 22 ff.). The framework developed below is based on requirements how qualitative research activities could be organized and thus describes a specific methodological framework. It includes several methodological rules that Albert (1987: 74) would call ‘technological rules’. Technological rules do not claim categorical imperatives for all kinds of qualitative research—they recommend methodical means that are more effective to achieve certain research objectives. We are going to validate this recommendation with several arguments throughout the following sections. However, as there are diverse research objectives and different forms of qualitative research, the framework depicted below does not prescribe that all qualitative case study research has to be conducted similarly. Instead, it is seen as helpful if researchers wish to reduce their own subjectivity by increasing the number of possible interpretations (the ‘repertoire to interpret’). Certainly, the extent to which theory is applied in qualitative case study research will vary according to the underlying epistemological view. Our general demand of theory integration explained in the methodological rules below is based mostly on constructivism including certain points of postpositivism and neopragmatism (Rorty 1998). Leitmotiv of pragmatic thinking is that reality is enacted ‘in situ’. Consequently, there is no absolute solid truth and no deterministic explanation. For a research design to be pluralistic and ‘work’ (Creswell 2009) in such settings it should integrate prior knowledge in order to be responsive to possible meanings of socially constructed reality and to be doubtful about the obvious. Therefore, responsiveness and fallibilism as important points of pragmatism are essential for our proposed framework. Prior knowledge is not only applied as a source of inspiration based on the pragmatist idea of abduction but also uses ideas of postpositivism and conceives theory as a means to ensure comprehensible argumentation during the entire research process (King et al. 1994). We are now going to specify the concept of the interpretative repertoire and illustrate how applying previous knowledge during the entire qualitative case study research process could increase it.

3 How theory enhances the repertoire to interpret during the process of case study research

A central argument for an omission of theory in qualitative case study research is that previous theoretical knowledge acts like a filter and hinders new knowledge to emerge (Denzin and Lincoln 2000). For example, Glaser and Strauss (1967) based their grounded theory approach on the demand to ignore the theoretical embeddedness of the studied phenomenon and call for a tabula rasa approach toward reality, whereby researchers are encouraged to drop all their lenses and to free their mind from any theoretical preconceptions on the initial stages of working with data. According to a constructivistic perspective, it becomes evident that the ‘prima facie’ plausibility of this objectivist argument contains at least two flaws:

  1. 1.

    It basically ignores the philosophical position that all perception is based on previous knowledge. In philosophy of science, it is argued whether observations can be considered as a neutral source of information or if they are necessarily infiltrated by theoretical assumptions (Godfrey-Smith 2003). Theory and facts are seen as interdependent (theory-laden approach): interpretation is never a perception without assumptions. This claim is rooted in hermeneutic phenomenology. Heidegger emphasizes that all interpretation is influenced by some form of pre-judgement: “Interpretation is grounded in something (…) we have in advance (…) a fore-conception” (Heidegger 1962: 191), which involves an individual’s background or history including previous knowledge. Theory-ladenness of facts means that those facts can be viewed only through a theoretical window (Guba and Lincoln 1998) whereby all observational judgments are affected by theoretical knowledge of the observer. Therefore, according to fundamental ontological assumptions of the ‘constructivist paradigm’, pre-understanding is not something a person can put aside (Kelle 1995). This creates problems for those approaches, which assume that our knowledge about social reality is based on “pure facts”.

  2. 2.

    It assumes only a piecemeal understanding of the influence of theory and claims that theory hinders ‘open’ perception, which is not essential. Instead, theory can enhance the ‘repertoire to interpret’. The term ‘repertoire to interpret’ is based on research in discursive psychology where it is used to express the background knowledge from which versions of actions, self and social structures are manufactured through talk (Gilbert and Mulkay 1984; Potter and Wetherell 1995). Interpretative repertoires in this perspective are “systems of terms used for characterizing and evaluating actions, events, or other phenomena” (Potter and Wetherell 1987: 149), and there are significant variations in these characterizations as, for example, actors perform different rhetorical devices. An extension of an actor’s interpretative repertoire therefore enables him or her to support and sustain a broader set of actions. Even though our approach is not rooted in the tradition of discourse analysis, the idea of different knowledge systems and their relevance for interpreting social data seems to be fruitful to us. Principally, data interpretation is determined by the self-boundaries of a researcher’s background or history (‘fore-conception’). Since the structure of a certain ‘fore-conception’ may be rather simple or complex, comprehensiveness of interpretations therefore differs as well. A frequently discussed way to reduce these boundaries is to engage different researchers, each of them coding data separately (Denzin 1989). The drawback of this ‘strategy’ is obvious: it is focused only on the stage of coding and cannot be applied to all kinds of studies. The basic idea of our approach is to build on the underlying principle of ‘creating’ multiple avenues to make sense of qualitative data. But instead of involving multiple researchers to ensure heterogeneous readings of data we argue that the repertoire to interpret data can be enlarged in a similar way by integrating extant theoretical knowledge. The term ‘theoretical knowledge’ is defined broadly here and can include certain object-related theories, such as transaction cost theory, theoretical frameworks that tentatively relate variables, which have been extracted from literature (Teece 2007) as well as meta-analyses, reviews or stylized facts that refer to generalized descriptions of a phenomenon. If the repertoire is regarded as a kind of storage, diverse types of theoretical knowledge can enlarge a researcher’s options and provide more alternative ways to interpret data or to ensure a more reflective research process. The repertoire therefore encourages abductive reasoning whether empirical data correspond with extant theory. Therefore, abduction is inextricably linked with creativity. The extent to which interpretation might be biased by diverse theoretical perspectives is therefore rather a matter of (not) applying creativity techniques than of applying theory. Figure 1 shows our underlying framework of integrating theory in qualitative case study research. The structure of this process is based on the stages described in Table 1 and on the work of Maxwell (1996) who defined qualitative research stages in terms of a network with elements such as purposes, conceptual context, methods or outcome.

    Fig. 1
    figure 1

    Stages of integrating theory in qualitative case study research

    Full size image
  3. 1.

    Research motivation: theory and the study’s relevance

The initial stage of any kind of research activity is the explanation of the researchers’ motivation. It is in the introduction of a research proposal where authors offer an explicit structure of their study and establish what Golden-Biddle and Locke (2007) call a “theorized storyline”. A theorized storyline consists of different rhetorical moves developed by authors in the initial paragraphs, or section for managing the relevance of their studies and therefore providing criteria “…according to which some results can be seen as more important than others” (Knorr-Cetina 1981: 110). A theorized storyline constructs the “anatomy of a topic” and should explain its significance, novelty, curiosity, or scope (Colquitt and George 2011). Glaser and Strauss (1967) and Glaser (2005) argue that the literature review should be delayed in order not to contaminate the research by pre-existing knowledge. Several scholars criticize this naïve empiricism view (Clarke 2005; Bruce 2007; Suddaby 2006; Dunne 2011) and argue that there is no objective view of the world and that it is better to make ideas and preconceptions explicit than to leave them implicit (Dey 1993). The initial justification requires putting the study in relation to ‘something else’, which means that authors need to contextualize their studied phenomena with existing conceptual knowledge and thereby convey how their own endeavor challenges, extends, or even rejects existing findings. Gaps in existing literature provide useful rhetorical links for making own contribution and highlighting the study’s significance (Golden-Biddle and Locke 2007: 38 ff.). Especially contradictory empirical findings in literature may be taken as an anchor to challenge current research, stress the study’s significance and emphasize its scope. In her study on routines Howard-Grenville (2005) outlines her conduct by addressing the seemingly opposing findings that routines can be both sources of change and inertia. By following other studies, which also generated theory from in-depths analyses of routines, the author methodically relates her study with findings from existing theory and thus makes space for her own agenda to fill the research gap. Therefore, theory can be regarded as a comprehensive knowledge dispenser of “…important management issues and concepts that enrich the field” (Gephart 2004: 455).

  1. 2.

    Theory supports indication of case study research

Frequently, researchers only apply a single method or research design separately without reflecting on reasons for or against certain methods. However, more recently in methodical discourses an explicit assessment of the appropriateness of the chosen research design is postulated (Flick 2014), because methodical rigor increases if researchers are able to justify their empirical approach. If researchers disclose the indication of the chosen qualitative case study design, a reference to theoretical knowledge can be helpful to strengthen the arguments. (1) A first important indication for qualitative case study research is seen in the research objective/question (Eisenhardt 1989). For example, a research project that aims at hermeneutic ‘Verstehen’ within a context of discovery indicates a qualitative design. Previous theoretical knowledge is strongly interrelated with this indication criterion, because the research object ‘hermeneutic Verstehen’ often results from the existence of divergent findings in literature. Similarly, the purpose of theory building from cases is linked to previous knowledge, because this purpose is “(…) tightly scoped within the context of an existing theory, and the justification rests heavily on the ability of qualitative data to offer insight into the complex social processes” (Eisenhardt and Graebner 2007: 26), which in turn indicate qualitative case study approaches for exploring multifaceted phenomena. (2) Frequently, the maturity level of a given theory is regarded as a further important indicative criterion for qualitative research in general and theory-building from case study research in particular (Edmondson and McManus 2007). The less is known about a certain phenomenon, the more qualitative research methods might be appropriate for exploring the object under study and for gathering more information about it (Steinke 2004). Therefore, reviewing previous literature helps to evaluate the state of knowledge and thus it is inextricably linked to method indication. A reference to theoretical knowledge may furthermore help to specify precisely what Edmondson and McManus (2007) call ‘nascent field of research’. A frequent fallacy here lies in the conclusion that a domain is nascent if there are only few hints in literature that cover the topic. However, not only the number of articles is important, but also theories can provide explanation bases on a safely ground for topics that could be perceived as ‘new’ otherwise. In their case study about value creation in e-businesses, Amit and Zott (2001) do not refrain from theoretical knowledge for the reason that e-businesses were a completely new domain with only a few descriptive papers published on this topic in those days. On the contrary, they applied several object-related theories, such as resource-based view, strategic network theory, or the Schumpeterian innovation approach, and transformed them according to their specified research question. Actually, many ‘new’ research topics can be approached (partially) with previous theoretical knowledge and its explication will improve the indication of case study research by making these choices more comprehensible.

  1. 3.

    Theoretical assumptions via sensitizing concepts help to identify phenomena

While stage (1) refers to benefits of applying previous knowledge to carve out the significance and relevance of a certain research project, at this stage, theoretical assumptions are used to identify phenomena from different angles. As explained before, at least from a constructivist perspective, the ‘deliberate stupidity-view’ that all extant knowledge is expendable is rejected (Kelle 2005). Instead, qualitative case study research should deal pro-actively with the question of how to utilize extant knowledge in a given research domain (Birkinshaw et al. 2011). Some supporters of Yin’s (2014) approach refer to previous knowledge to construct theoretical propositions and test them. In interpretative case study research, on the other hand, theory is applied to identify domains of potential importance for the studied phenomenon. Knowledge of those domains is important because otherwise the researcher is influenced only by his subjective problem definition or facets that are explicit for participants. No ‘instance’ is around that helps to request for further facets that may lie in deeper structures. In this regard, theoretical knowledge can be used as heuristic tools that focus attention on certain nuances (Kelle 1995). Certainly, that does not mean that fixed and specific benchmarks are used to identify phenomena but they rather provide theoretical guidance for formulating sensitizing concepts. Sensitizing concepts “give the user a general sense of reference and guidance in approaching empirical instances” (Blumer 1954: 7). A researcher can use extant theoretical knowledge to merely suggest directions for finding and increasing reflexivity (Haynes 2012). In his exploratory case study on the involvement of different managerial levels in exit decision-making, Brauer (2009) applies Burgelman’s (1994, 1996) process model of business exits. Although existing knowledge about managerial involvement in strategy processes, especially in exit decision-making, is quite limited, Brauer (2009: 344) situates his study in the existing empirical research and thereupon formulates his own theoretical assumptions about the key contingencies that outline the degree of manager’s involvement in divestitures.

Sensitizing concepts also may be a starting point for formulating theoretically rich frameworks. As mentioned before, theoretical frameworks only tentatively relate variables and still enable researchers to explore unknown facets of studied phenomena (Teece 2007). Therefore, theoretical frameworks generate a specified ‘linguistic game’ that already provides certain rules and structures for exploring multifaceted phenomena and stimulates researchers’ own theoretical preconceptions about phenomena. In his field study on innovation-related firm renewal, Danneels (2002) develops his own theoretical framework for exploring the interplay of firm competences and product innovation. The author draws on seemingly dispersed findings in the field of dynamic capabilities, product performance, path dependence, and organizational learning in order to organize his theoretical assumptions about potential relationships between competences and product innovation. In conclusion, existing theories or theoretical frameworks systematize existing knowledge in a certain research area and provide an ‘umbrella’ that organizes and integrates relevant findings and thus clears the researchers’ view for in-depth analyses. By using theories as heuristic tools to formulate sensitizing concepts it is quite unlikely that researchers are caught in what have been known before. In this sense, not the refusal to use theory is important to remain ‘open’, but rather the way it is applied.

  1. 4.

    Case selection: theory supports selecting meaningful cases

Since statistical sampling is not an appropriate method to select cases in qualitative case study research, picking the ‘right’ cases is a pivotal activity here. Theoretical and purposeful sampling techniques can be applied in qualitative research. Contrary to the supposed meaning of the term, ‘theoretical sampling’ describes a sampling strategy, which is basically based on the temporary results of previous cases (Glaser and Strauss 1967) and only barely the result of previous knowledge. It is the sampling method that has the highest degree of openness. However, sometimes researchers might wish to select single or multiple cases purposefully and the question arises if theoretical knowledge supports this process or rather builds a barrier to openness.

While single case studies perfectly suit the purpose of richly describing certain phenomena, multiple case study research provides a more extensive basis for building theory (Eisenhardt and Graebner 2007; Yin 2014). Strategies for single case selection comprise for example the identification of a case with revelatory potential, the extreme case method or the typical case method (Bryman 2008a). Multiple cases are often selected specifically if researchers have initial ideas about ‘problem dimensions’ that are important to be considered (Yin 2011). These dimensions usually serve as inputs to deliberate sampling plans. In their study of dying organizations, Harris and Sutton (1986) used a purposeful sampling since they had previous ‘problem dimensions’ in mind (e.g. private and public, dependent and independent companies). Regardless of whether only one case or multiple cases are selected, the selection process is a critical task because it predetermines possible outcomes. All techniques have a theoretical infiltration in common: the decision to assign the label ‘typical’, ‘extreme’, etc., or the categorization of a case as ‘relevant’ requires a pre-understanding of the sample or the phenomenon under study. It is important to realize that identifying a typical case or a relevant problem means nothing else than the explication of one’s own theoretical assumptions about meaningful ‘problem dimensions’ that should be investigated. In this regard, the researchers’ explicated theoretical assumptions sensitize them for relevant and particularly not yet covered aspects about phenomena in a given field. For example, in their explorative single case study Tripsas and Gavetti (2000) invoke the dispersed and incoherent findings in capability and managerial cognition research during (technological) change as the argumentative basis for conducting an in-depth study. They purposefully selected Polariod as center stage of their study because the corporation suits the requirements of a research object that is undergoing a radical transition. Furthermore, in his study on innovation-related firm renewal, Danneels (2002) purposefully includes five firms that differ along several ex ante specified criteria (age, size, diversification level) in order to vary his sample as much as possible. The specification criteria are not arbitrarily chosen, but rather reflect a solid theoretical knowledge in a given domain. Summarizing this section, when using purposeful sampling techniques the consideration of theoretical assumptions offers important avenues to select meaningful single and multiple cases in order to avoid arbitrariness.

  1. 5.

    Collecting data: theory provides for the threat to overlook the unconscious

Methods for collecting qualitative data are extensive, including observation, content analysis, surveys or experiments (Flick 2014). Regardless of the fact that theory may play different roles in applying these methods, a lot of communalities exist as well. The advantages of theory integration in the process of data collection will be shown generically for semi-structured interviews and archival data which are frequently used in case study research (Eisenhardt and Graebner 2007: 28), where researchers ask open-ended questions and use guidelines to structure the topic (Flick 2014). Charmaz (2014) refers to the importance to be familiar with the public information about the organization. In addition to this, theory integration can be helpful to ask the ‘right’ questions and to educe meaningful descriptions from the informants, as not all facets of the problem are always in the participant’s explicit awareness. Not mentioning a fact by a participant during an open interview is not a sufficient condition for its unimportance. Theoretical knowledge can help to raise questions about possible ‘problem dimensions’, whose relevance certainly is subject of the participant’s perception. Asking for the relevance of a certain problem facet does not automatically force the participant to agree, but asking not induced the threat to overlook the unconscious.

Asking the right questions basically means to craft an interview guideline on the basis of previous theoretical knowledge. This guideline has to guarantee flexibility and openness according to further problem areas or relevance settings by the participant, otherwise it would be a broad operationalization of the theoretical assumptions. In this regard, theoretical frameworks or process theories contain several broad dimensions, like e.g. context, content, and process dimensions (Burgelman 1983b; Pettigrew 1987), which provide a didactic structure for grouping distinct questions around theoretically important areas and thus facilitate the interview process for both the researcher and the interviewee. Some researchers used theoretical knowledge exactly in this way to structure their guidelines. Bacharach et al. (2000) for example structure the interview questions in their study on boundary management tactics around a theoretical framework, which focuses on the impact of personal and contextual characteristics on support providers’ actions (Bacharach et al. 2000). The formulation of interview questions serves the purpose of making the researchers’ implicit theoretical knowledge more explicit. As a consequence the questionnaire works as a tentative guideline for structuring the entire interview process. During the interview researcher take a more passive role as a dialogue partner and let the experts talk. But in situations where the conversation seems to fizzle out, the prepared questionnaire might help the interviewer to induce new stimuli to keep the conversation going and if needed to direct the interview to theoretically important areas, which have not been covered yet.

Apart from verbal data, qualitative research also benefits from archival data stored in documents, press articles or company reports. In qualitative research different methods exist, which exclusively rely on such secondary data for inductive theory-building, especially in the realm of qualitative content analysis (Mayring 2000). But most often archival data complement, or triangulate interview data and provide published and accessible facts about e.g. interesting industry factors, financial data or company history (Charmaz 2014). Nevertheless in order to select ‘relevant’ documents, especially from huge databases, researchers may operationalize their theoretical knowledge for placing queries. Useful keywords might reflect widely shared and agreed terminologies in a given research area and thus condense the researchers’ conceptual knowledge in order (1) to place theory-related queries and then (2) to filter theoretically relevant documents, or articles, which cultivate interview data and help to enhance the uniqueness of the studied phenomena.

As a result, it is the researcher’s theoretical knowledge and the formulated sensitized concepts, which provide a certain structure in the conduct of expert interviews. However, this form of explicit systematization enables researchers to operationalize their most often rather implicit theoretical assumptions about phenomena and therefore must not be equated with the highly formalized and standardized questionnaires, which are used by quantitative surveys. Although some advocates of more naturalistic inquiries claim that even semi-structured interview guidelines are a perceived threat to openness and thus are hardly compatible with explorative case study designs, we believe that theoretically sensitized interview guidelines still leave room for the discovery of new facets of phenomena. Even if sensitizing concepts are a mould for forming interview questions, they remain preliminary in the conduct of an interview and thus the researchers’ theoretical assumptions may be refined throughout the interview process and grant openness for generating new insights.

  1. 6.

    Coding procedures: theory overcomes the self-boundaries of interpretation

Coding of observations or texts is the basic analytical process in qualitative case study research and it is represented in all the interwoven operations by which researchers collect, conceptualize and form categories out of data. Generally, coding takes place stepwise, e.g. primarily rather close to the native data and later on more conceptional. Coding is a rather subjective activity and to a certain degree bound to the interpreter. This is probably the rationale for the manifold calls for openness. The mistake is, in our view, that the influence of theory during the coding processes is conceived oversimplified as forcing data into a so-called ‘procrustean bed’. On the basis of the coding procedures of Grounded Theory we will demonstrate that the opposite is true: applying theoretical knowledge helps to overcome existing self-boundaries of interpretation by calling ‘societal reason to the witness stand’.

  1. 1.

    Open coding is the first step in an interpretive process by which unstructured data are analytically broken down and merged to concepts and aggregated categories (Corbin and Strauss 1990). The term ‘open’ means here a pure reconstruction of the recipients’ view, partially even in their own words (in vivo). However, open coding is not a random labeling technique, where researchers assign arbitrary concepts to interview passages that are unrelated to existing theory. Every interview consists of ‘fresh’ answers of interviewees that are not influenced by previous knowledge in the field and answers that reproduce labels or terms, which are adopted from the theoretical discourse. An interview about change-related aspects may result in labels like ‘inertia’ or ‘turnaround’, which are theoretically loaded. This perspective already challenges the naïve empiricism view of theory emergence apart from previous knowledge. During the initial coding procedure openness of coding is supported by abductive reasoning (Kelle 2014). Researchers confront their elicited surprising data with pre-existing knowledge in a way that those preconceptions may be revised to become consistent with the empirical data. This is done i.a with the technique of constant comparisons. When constantly comparing categories, researchers comprehensively analyze each category, try to critically reflect their meaning by comparing it with other existing codes and moreover by comparing the emerging categories with the own theoretical assumptions (Strauss and Corbin 1990). In his study of product innovation and competences, Danneels continuously matched and contrasted memos and codings and compared them systematically with existing bodies of work. He deliberately does not code data by “starting from scratch” (Danneels 2002: 1101) but used previous knowledge to refine understanding and code reflectively. Thereupon the formulated sensitized concepts prevent researchers from arbitrarily assigning unrelated codes and hastily cementing preconceived views. By comparing each category, researchers also may rule out alternative theoretical interpretations, which could have been ascribed to different codes. For example, by making use of a so-called ‘flip-flop’ coding technique, individual categories are ‘theoretically’ differentiated and thereby transferred to other contexts of meaning (Corbin and Strauss 2008). In this respect one can argue that comprehensive knowledge of different theoretical variables also increases the creative capacity to abstract from that knowledge and thus facilitates to give initial meaning to data. Reichertz (2007) calls this abductive inference ‘logic of discovery’.

  2. 2.

    After open coding has initially fractured data in order to identify first categories, axial coding is the subsequent procedure that puts data back together in new ways by making connections between the developed categories on the basis of a ‘coding paradigm’ (Strauss and Corbin 1990). The technique of axial coding is essentially linked to theory because the coding paradigm itself represents a theory of action that is rooted in pragmatist and interactionist social theory and relates the phenomenon to its surrounding dimensions (context factors, causes, and consequences) (Corbin 1991; Corbin and Strauss 1990; Strauss 1987). Additionally, the general coding paradigm might be substantiated further by assigning previous knowledge to the respective dimensions. E.g it might be fruitful to divide the dimension ‘context’ into ‘internal’ and ‘external context’. In their study on international entry modes, Wrona and Trapczynski (2012) explicitly incorporate Dunning’s (1988) three forms of advantage in their coding paradigm and use them as theoretical inspiration for analyzing how pharmaceutical companies enter into transition economies. Thereupon they carry on their categories from open coding, iteratively refine and rearrange them in order to generate broader and theoretically more enriched categories that are finally matched and likewise contrasted with the designed coding paradigm.

  3. 3.

    The final stage of coding refers to the identification of the most important, or even most surprising finding of the case analysis that is represented in a core category. This technique is called selective coding and continues the former process of relating categories at a higher level of abstraction (Corbin and Strauss 2008). In order to elicit the study’s core category researchers may rely on existing knowledge to sharpen the idiosyncrasies of the analyzed case and ‘distill’ its empirical essence. Hence, the researchers theoretical knowledge will be constantly rolled out upon, sharpened and contrasted with the empirical findings in order to make sense of the phenomenon’s peculiarities. The development of new concepts based on empirical data is therefore an abductive ‘pinch grip’ of oscillating between the empirical data and previous knowledge (Kluge 2000). Summarizing this section, theory can be used as a means to code data in a reflective way. Certainly, it cannot be ruled out that researcher uses theoretical knowledge to force data in pre-existing categories. But using theory thoroughly as described above creates sort of a ‘think lab’ for alternative meanings of data, hinders hasty interpretation, supports organizing data and sharpens the peculiarities of the case.

  1. 7.

    Outcome presentation: theory unveils the study’s contribution

Theory certainly is an issue at the stage of presenting the results of qualitative case study research, and at this stage, it is the most undoubted. Even proponents of delaying theory application support its use at this stage of investigation (Glaser and Strauss 1967). The role of theory at this stage can be discussed both as outcome and as input. Presenting the outcome of case analyses refers to the process of writing the emergent theory. Theory is the outcome of research here and can result in new categories, theoretical frameworks, typologies, hypotheses, or mid-range theories (Bryman 2008a; Corbin and Strauss 2008). However, in order to demonstrate how own findings make a contribution to the research area theory served as input at this stage, too. Existing theory is used to enter a “synergistic or antagonistic dialogue” (Ridder et al. 2014: 9) to demonstrate the contribution by challenging, changing or fundamentally improving the understanding of how phenomena are perceived in existing literature (Eisenhardt 1989; Stern 2007). This dialogue between case study findings and existing theory strives for seeking either complementarities from the study’s initial theory domain, or rather seeking dissimilarities by drawing upon theories, which go beyond the study’s initial theory domain. Recently, scholars have been working on the development of precise criteria that describe attributes of a substantial contribution (Corley and Gioia 2011; Flyvbjerg 2004; Ridder et al. 2014). While enfolding literature, existing theoretical knowledge might serve as a reference point for judging the quality of the emergent theory and unveils the own contribution.

  1. 8.

    Quality criteria: theory as benchmark for increasing the quality of case study results

Usually, research projects close with a paragraph about the study’s strengths and limitations. Nowadays, a broad discussion takes place in literature about appropriate quality criteria for qualitative research designs (Altheide and Johnson 1998; Cassell and Symon 2011; Easterby-Smith et al. 2008; Gibbert and Ruigrok 2010; Kirk and Miller 1986; Reid and Gough 2000; Steinke 2004). Generally, there are three approaches on quality criteria.

  1. 1.

    Dismissive attitude towards quality criteria because of methodological assumptions such as radical constructivism (Reid and Gough 2000).

  2. 2.

    Use of alternative appropriate quality criteria: qualitative research is basically different compared to quantitative research but should be evaluated anyhow (Kirk and Miller 1986; Lincoln and Guba 1985; Steinke 2004).

  3. 3.

    Use of adapted quality criteria: return to the criteria’s epistemological core and ‘saving’ the criteria regardless of the research method. The idea is to protect results from threats for their trustworthiness (Wrona 2006).

Instead of reviewing different positions concerning these criteria, which is beyond the scope of this paper, we will take the latter approach as a basis here. The idea is to build on what Cook and Campbell (1979) call “threats for validity”, define specific threats for qualitative case study research and discuss ways to assess those threats. Here the epistemological core of the traditional criteria derived from the positivist paradigm is transferred to qualitative research. For example, the core idea of the criterion ‘reliability’ can be seen as ensuring consistency and stability of measurements. In nomothetic research this is restrained by using re-tests, parallel test, etc. to examine if the ‘true value’ remains the same. In qualitative research a ‘true value’ is reconstructed by the researcher—therefore, a quality criterion that is related to ‘reliability’ may be the proximity of the construction to the ‘real’ meaning of the data as well as a disclosure of these interpretations. This perspective corresponds to what King et al. labeled as a common underlying logic in quantitative and qualitative research (King et al. 1994). Following this perspective, the application of theory can increase quality in several ways and serve as a benchmark to evaluate case results.

First, the assertive use of theory can be interpreted as an improvement of ‘reliability’. In the previous section, it was shown how prior theoretical knowledge increases sensitivity for specific issues that guides researchers towards certain outcomes, such as the creation of an interview guideline or the development of a definite code. An extensive disclosure of prior theoretical knowledge and its influence on interpretations can be evaluated as an improvement of reliability because it enables a higher replicability and transparency of the research process and promotes procedural reliability. Second, the theory-laden view of case study research increases ‘validity’. A comparison of the covered scope of a developed construct with possible prior theoretical knowledge in this field will be a fruitful way to determine ‘content validity’ of the measurement, because theory may be seen as a specific form of the content domain associated with the construct. Similarly, ‘construct validity’ of developed categories can be specified with regard to prior theories. Possible threats to construct validity such as mono-operation bias or vague operationalization of the construct can be counteracted by theoretical sensitivity. Furthermore, prior theoretical knowledge may build a criterion variable that represents the newly developed categories. A match between previous knowledge and generated categories will shed light on ‘criterion validity’. Due to the circular and iterative process of qualitative case study research, this theoretical knowledge is not only to be used for the assessment of validity at the end of empirical research but it also influences the validity of interpreting by permanently referring to representative categories or the content domain of categories. Similarly, this process of referring to theoretical knowledge enhances ‘internal validity’ since inferences of the researcher are based on diverse sources of evidence and are disclosed.

Third, taking prior theory as a basis of case study research will increase the possible degree of generalization of the outcomes. Theoretical knowledge helps researchers to match idiosyncratic empirical findings with more general prior knowledge and enables hereby to better understand the novelty of findings vis-à-vis existing knowledge and to broaden the strong situatedness of interpretations in favor of more conceptual categories and a logic of inference. In qualitative research, this is done by analytic generalization where researchers demonstrate how their specific findings base on prior theoretical knowledge and how theory might be applied to comparable contexts in order to create and extract analogous categories (Yin 2014). Summarizing, integrating previous knowledge can serve as an important means to discuss the trustworthiness of the findings at the last stage of the research process.

Referring to the methodological framework depicted in Fig. 1, the previous section described in detail how theory could be integrated in qualitative case study research to increase the repertoire to interpret. On the basis of these explications, Table 2 now substantiates the methodological framework and provides exemplary methodological rules that are associated with the described theory-ladened approach and enable researchers to benefit from theory integration.

Table 2 Methodological rules for theory integration in qualitative case study research
Full size table

4 Conclusions

Qualitative case study research can provide important insights when its methodical strengths such as holistic view of the research object, context-sensitivity or consideration of everyday life settings are combined with a theoretical responsiveness. Qualitative case study designs today vary across different epistemological worldviews and different proponents. This paper introduced a theory-oriented approach, which is mostly based on constructivism incorporating facets of postpositivism. Our framework depicts qualitative case study research as an interpretative endeavor to construct meanings from qualitative data. The basic idea is that these constructions benefit from theory integration as opposed to emerging naturally from data not influenced by preconceptions. As described above, the reasons for this are based on the assumption that extant knowledge will increase a researcher’s awareness of diversity of meanings, broaden the repertoire to act and interpret, enhance the comprehensiveness of research and increase probability of generating scientific progress. The methodical rules in Table 2 demonstrate how researchers could be stimulated to integrate theory in case study research.

This paper contributes to the general methodological debate in qualitative research, and in case study research in particular, in various ways: first, it makes a methodological contribution concerning the debate on openness vs. theory-ladeness or on inductive versus deductive research characteristics. Despite the fact that there is only limited dispute today over the fact that a tabula rasa concept is not viable and the basic idea of theory-ladeness is present in general (Kelle 1995), its explication is still underemphasized in methodical literature. Depending on the research design and epistemological views, theory integration in current methodological approaches mainly takes place at the beginning of the research process (theoretical gaps) and/or in the end (theoretical contributions). The few existing papers that highlight the theory-ladenness of qualitative case study research frequently focus on a practical claim that postponing literature is not effective (Bruce 2007; Suddaby 2006; Hallberg 2010) or emphasize general benefits of theory integration (McGhee et al. 2007; Dunne 2011). Strübing (2007) therefore points out, that it is more important now to assess how to make use of previous knowledge. Current attempts to cover this topic are still too focused on certain stages of qualitative research. Predominantly, scholars refer to the role of previous literature to stimulate coding processes (Coffey and Aktinson 1996; McGhee et al. 2007; Urquhart 2007; Thornberg 2012). Noteworthy exceptions are the papers of Tummers and Karsten (2012) who discussed literature usage at the stages ‘research design’, ‘data collection’ and ‘data analysis’ as well as the constructivist grounded theory approach by Charmaz (2014) in which previous knowledge is entangled. Instead, what is still missing is a comprehensive framework (a) that incorporates the entire research process and does not only focus on coding, and (b) that refers to theory integration and not only to existing literature. Our developed framework for theory integration describes precise steps to broaden the researchers’ repertoire to act and interpret. Here we distinctly move beyond what is already known since we describe avenues of theory integration throughout the entire research process including stages that are not even mentioned so far (e.g. method indication or trustworthiness). Apart from considering existing literature we furthermore explain how grand theories or theoretical frameworks might be used in nascent research fields with low literature coverage. By referring to the repertoire concept we furthermore clarify that theory integration is not inevitably in conflict with openness. Just the opposite is true, for example, if theory is used to produce as many distinct interpretations (‘Lesarten’) of data as possible. Therefore, by developing a comprehensive framework we significantly contribute to the current debate and move it forward.

Second, the paper enhances the methodical discussion concerning the process characteristics of case study research by introducing modernistic elements in terms of process standardization, formalization, or guidelines regarding the integration of theory. Today, qualitative research methods are well established in social sciences and various methodical contributions are available for researchers to navigate through the complex qualitative research process. However, the awareness about possibilities of theory integration is still weak. As van Maanen’s (1998: xxv) puts it: “there are […] rules for writing the persuasive, memorable and publishable qualitative research article but, rest assured, no one knows what they are”. In contrast to other papers on this topic, to our knowledge we are the first to present a set of methodological rules that span over the entire research process and unveil how previous knowledge can be integrated systematically. Additionally, these rules have not been introduced in a ‘confessing’ way but on the basis of our methodological framework.

Third, this paper extends the debate on quality criteria. Currently, the debate consists of various sub-debates on the usefulness or the manner of quality criteria (e.g. Altheide and Johnson 1998; Easterby-Smith et al. 2008; Gibbert and Ruigrok 2010; Kirk and Miller 1986; Reid and Gough 2000; Steinke 2004). Realizing that prior knowledge influences all stages of the case study research process, a disclosure of prior theoretical assumptions and their application throughout the research activities can be seen as an improvement of reliability. Moreover, theory integration can be considered as a means to improve internal and external validity because extant theory serves as a content domain of research. In the current debate, nearly no attention has been paid to the fact that theoretical anchoring will strongly affect the research process and its outcome. Consequently, its application and disclosure has to play an important role in the evaluation of research quality and should serve as an additional quality criterion.

Finally, the introduced theory-laden approach of case study research may be a further step on the way to overcome the unhelpful gap between qualitative and quantitative research (positivism disputes). An important implication of these disputes lies in current efforts to define overarching minimal standards of research that emphasize common features and not differences. “Our view is that these differences are mainly ones of style and specific techniques.” (King et al. 1994: 3). As a consequence of those standards, qualitative research has to address the role of theory more systematically, similar to its central role in quantitative research. Despite all differences between qualitative and quantitative research, the degree of theory integration has several communalities to postpositivistic research. The frequently found perspective arguing that qualitative case study research should refrain from theory oversees important methodical developments. In this sense our paper contributes to the field of developing a minimal research design as it outlines how the entire qualitative research process is permeated by theoretical knowledge. Therefore, this theory-laden approach may help to alleviate the existing debate.

The presented dictum of a theory-laden case study design might be criticized particularly from the angle that theoretical preconceptions repress open interpretation and authenticity of data by ‘contaminating’ the qualitative research process with existing perspectives. Even a rejection of the naïve empiricism approach does not necessarily lead to the conviction that a theory-laden approach is the consequence. It cannot be denied that theory integration might bias interpretation and impose pre-existing structures on data. However, in our opinion, this critique underestimates researchers’ reflective ability and assumes a researcher that reckons up data with matured knowledge. On the contrary, we described ways of theory integration that stimulate sensitivity towards ‘resistant’ data. Furthermore, it has been shown that the ‘bias threat’ mainly applies to coding activities. At other stages (e.g. at the indication or trustworthiness stage) applying theory strongly supports research comprehensiveness. Finally, the point that theory plays a role in our framework does not imply that theory plays the main role. Its ‘contaminating’ power is kept covered by certain techniques (e.g. constant comparisons) and methodical defaults.

Summarizing this section, our paper shows how theory integration can be utilized to multiply prospects and to enrich interpretive horizons of researchers: The one who sees more is more right (according to Husserl). Furthermore, theory integration makes qualitative case study research more reflective and ‘objective’, because the one who sees more becomes more doubtful in his actions as more options are perceived. In this way, theory integration may serve as an ‘antidote to the obvious’  (Reichertz 2004). As shown before, theory integration does not automatically lead to less openness. If theory is used as a sensitizing concept and not as a ‘deposited conviction’ it may have the potential to enrich the entire qualitative case study research process.