Building theories from case study research doi

This stage of the analysis resulted in the initial selection of a sample of 43 journal articles. The review team conducted a more detailed analysis of the 43 qualitative articles within the sample. Or when under closer inspection the article did not address issues pertaining to the organization, management, and performance of networks. A sample of 31 articles remained after refinement.

In stage 3, the review team began the process of analysis, that involved open and inclusive discussion of the 31 articles.


  • Towards More Case Study Research in Entrepreneurship.
  • belonging thesis statements you like.
  • essay writing on news papers;
  • Qualitative Case Study Research as Empirical Inquiry.
  • latex bibliography cite thesis?
  • Using Case Study Research to Build Theories of IT Implementation;
  • Paper statistics.

This technique marked the first stage of a process that applied a series of thematic criteria to underpin analysis. With respect to analytic approach , we anticipated that this would vary depending upon the type of theorizing in play. As previously noted, inductive approaches start with an empirical orientation, deductive approaches start from extant theoretical understandings, and abductive approaches utilize both.

Building Theories from Case Study Research

In some cases data may be coded, while in others they may be analyzed in broad thematic terms and even though these trajectories will differ, all three approaches should facilitate the generation of the conceptual leaps that underpin theorizing. The emphasis on analytic approach in this study is designed to capture how theory informed the analysis, whether the study was inductive, deductive, or abductive, and how data were analyzed by the researcher. However, it should be noted it was not possible to discern this information for all cases due to a level of ambiguity within many studies.

Consequently, we have sought to provide a summary interpretation of the analytic approach adopted in each case. Further, we note that our sample did not include any examples of purely deductive research. Iterating between the studies and the criteria set, the team then proceeded to agree a summary of each article. The 31 studies were then subject to a further stage of narrative review by the research team.

This process involved the identification of a subset of four studies that were selected as they provide neat illustrations of the variation in approaches to theorizing through qualitative research.

The first two studies each adopt an inductive approach but execute this in quite different ways. Bate provides an example of thick description based on ethnography, while the article by Waring, Currie, and Bishop is indicative of interpretive or thematic analyses that have been identified as underrepresented in leading management journals Cornelissen The latter two studies engage in different forms of abductive theorizing, combining deduction and induction. Ferlie and McGivern is noteworthy for its strong upfront utilization of theory and Saz-Carranza and Ospina provide a particularly comprehensive overview of their analytic approach.

Discussion of these cases facilitated the opportunity to develop a broad set of principles to support theorizing in qualitative public administration research. These are presented and reviewed in detail in the closing sections of this article. Having discussed the research design and analytic approach, we employed in the context of this study, the following section of the article reports on the research findings that emerged from the narrative review. It begins by providing a summary characterization and analysis of the full set of qualitative studies.

Detailed analysis of the full set of studies which is reported in full in Supplementary Table 1 indicates that scholars are addressing significant and important theoretical questions on network and collaborative governance through qualitative research. However, the reporting of information on qualitative methods and data collection among these studies is variable and somewhat ad hoc. Some authors afford particular attention to the sampling rationale, when for other studies this could be more systematically addressed. Similarly, on occasions data gathering is extensively detailed, while elsewhere it is pithily described and, in a minority of articles, only briefly mentioned.

Beyond an often broad characterization of analytic approach, there is sometimes a lack of specification of processes underpinning the review, incorporation, and presentation of data. Authors vary in their description of the derivation of themes and codes, and differ in the detail provided regarding the link between data and conceptual claims.

Few studies discuss these in detail and those that do, tend to demonstrate abductive theorizing seldom labeled as such that combines elements of inductive and deductive approaches.

It is interesting to observe that few articles explicitly identify with, and conform to, a pure grounded theory approach, despite the common perception that most qualitative research adopts this perspective Suddaby ; Tummers and Karsten However, in most cases, these choices remain implicit, along with reflections on the wider applicability, implications, and generalizability of findings.

Our overview of the set of 31 qualitative studies on network and collaborative governance provides some support for claims that qualitative studies are developing greater theoretical ambition and sophistication in the specification of their contributions e. However, our findings resonate with others in identifying that all too often aspects of the qualitative research process remain opaque Ospina, Esteve, and Lee This potentially undermines the value of qualitative research and places limitations on the quality of theorizing, especially when there is an inadequate account of the steps taken between data collection, analysis, argument and conclusions.

On the basis of this review, we propose that theorizing in qualitative public administration is likely to be most effective when authors provide transparent accounts for, and outline the connections between, their research design, analytic approach, and theoretical contribution. To further unpack these elements and elaborate examples of the theorizing process, we now move to discuss four illustrative studies derived from the wider set.

Login using

The first of the two studies that exhibit inductive theorizing is Bate Eschewing a detailed literature review, the research problem is set out as one of cultural fragmentation that stymies attempts at networked governance. Indeed, very soon into the article, Bate launches into rich and contextualized description and manages to organize the data in a coherent manner to capture the chaotic nature of organizational life, without drawing upon themes derived from literature.

The article temporally brackets the process of change, providing a clear sequencing that assists the reader, while also ordering data to demonstrate organizational problems, early impressions, subsequent diagnosis, and solution Langley ; Langley et al. In this sense, he describes a networked governance process that is ongoing, rather than linear, and one with no end point yet visible.

He reports that he carried out visits days over a 2-year period, with myriad opportunities for observation, as well as a large number of interviews approximately Given his embeddedness within the organizational setting, there are a larger number of direct interview quotes and observations than would be usual, even within qualitative studies. The overall effect of this is one of vicarious engagement with the research site as the reader feels they are transported into the organization in a way that proves particularly engaging. Senior Nurse. Neutrality and even-handedness were central to this process.

We were positioned in the space between management and workforce, strenuously avoiding being seen as management-centric the traditional bias of organization development and worker-centric … mediating between the different interests and perspectives …. Our second example of inductive theorizing comes from Waring, Currie, and Bishop , a study that differs significantly from Bate as a theoretical gap and rationale is evident from the outset.

Download Limit Exceeded

The strength of theorizing observed within Waring, Currie, and Bishop lies partly in an inductive approach applied to understand and explain the impact of partnership configuration on local implementation. This allows the authors to challenge dominant perceptions of network governance effects and, as a result, specify and present a theoretical model in figure form figure 3 ; Waring, Currie, and Bishop , However, Waring, Currie, and Bishop boost explication of theorizing significantly by detailing the research process through eight paragraphs of text and a supporting figure summarizing coding and thematic analysis see extract in figure 4.

To deliver transparency in research design and facilitate theoretical generalization, they provide background on their comparative cases to situate their analysis and elaborate on their sampling strategy. Finally, Waring, Currie, and Bishop reinforce theorizing through a transparent attempt to connect analysis, the presentation of their data and theoretical contribution—bringing both evidence of the research process, and credibility to the account.

In so doing, supporting extracts from illustrative examples of coded data are provided in tables, while narrative reporting incorporates illustrative quotes.

Related eJournals

Note, Waring, Currie, and Bishop are unusual in actively reporting on a full range of data, including that derived from policy documents and clinical operating procedures see figure 5 , for examples. Document Excerpts from Waring, Currie, and Bishop , The theoretical contributions in our next pair of studies are underpinned by an abductive approach that combines deductive and inductive theorizing techniques.

In the first study, provided by Ferlie and McGivern , theoretical concerns are again apparent from the outset. Labeled as iterative, but displaying abductive characteristics, Ferlie and McGivern puzzle on the long-term shift to indirect governance.

Search form

They identify extant accounts as incomplete, and aim to enrich theorization. Development of an interview proforma deductively informed by governmentality and competing theoretical perspectives enables empirical interrogation in pursuit of their theoretical objectives. Theorizing is strengthened in this study by an unambiguous set of research questions and a highly transparent account of theoretical implications in the discussion section of the article. Although Ferlie and McGivern provide a helpful and clear account of their theoretical contribution, we argue that theorizing also requires a high degree of transparency in research design and analytic approach.

Their ability to provide theoretical generalization is delivered through a comparative case approach, combined with explicit connection with theory to boost external validity. Ideally, the rationale for the research setting and sample could be better explained and connected to subsequent methodological choices. And while their theoretical framework and contribution is elaborated in some detail, little information is given on coding procedures and analytic steps for the two case studies. In particular, the inductive aspects of their analysis are not explicated.

So while Ferlie and McGivern provide a high degree of transparency in their application of a Foucauldian theoretical perspective to enrich traditional theoretical conceptualizations of governance, and deliver helpful clarity on their research design, some steps taken within their analysis remain opaque.

This serves to marginally undermine clarity in their theorizing and somewhat obscures their conceptual leap. Nevertheless, a significant amount of primary data from both cases is reported which allows the reader to engage with the context and gain insight into the effects experienced by those working within the network.

Our second study characterized by an abductive theorizing approach comes from Saz-Carranza and Ospina However, beyond an empirical contribution focused on the behavioral dimensions of network governance, theorizing is explicit and convincing from the outset.

This is striking considering that claims in qualitative articles tend to be more tentatively written, with caveats regarding the transferability of findings. Indeed, the article conveys a degree of specificity in the transparency of research design, analytic approach, and theoretical contribution—and the connection between the three—that is, rarely seen within qualitative research.

Theorizing in this study is especially convincing due to the clarity and consistency of detail offered in the documentation of research design and analytic approach. The authors present a rationale for the adoption of a qualitative approach, while also detailing their sampling frame, both in the text and in tabulated form. Running for almost seven pages, the qualitative methods section is one of the most transparent and informative published in public administration. However, we observe that the article runs to nearly 40 pages, emphasizing the trade-off between length and depth in qualitative research and demonstrating the importance of editorial discretion.

Greater flexibility in this regard, as well as alternative strategies—such as enabling the placement of methodological material in online appendices—may help to facilitate the publication of qualitative research that provides rigor and richness. Of particular note in this study is the established connection between the analytic approach underpinning the two phases of study and the subsequent theoretical contribution.

Ospina et al. Interview data is privileged, relative to the analysis of observation and documentary data, but the explicit and thorough analytic account facilitates a transparent and convincing basis for theorizing and stands in contrast to the fleeting references sometimes found within qualitative studies. Having summarized a set of qualitative studies on network and collaborative governance and illustrated alternative theorizing approaches, we next discuss the implications arising from our review and present a framework that facilitates rigorous theorizing from qualitative research in public administration.

This study presents evidence derived from a two-stage investigation of theorizing in qualitative public administration research, focused on a subset of studies in the field of network and collaborative governance. The review indicates that distinctive sets of research questions had been addressed, including those with an exploratory and explanatory orientation drawn from a range of theoretical perspectives.

However, the systematic review also illustrates wide differences in the accounts of qualitative research provided across the set of studies. This suggests the need for a set of broad guiding principles to aid authors and reviewers in providing and prompting greater transparency in accounts of the qualitative research process. In response, in table 4 , we draw together and summarize these principles relative to three key component elements of the research process: research design , analytic approach , and contribution. Research design is fundamentally concerned with how authors report on how their research has been conducted.

Research context and background

Analytic approach focuses on how authors move from an account of their data collection to offer interpretation and a series of claims. Finally, contribution involves the way authors articulate and highlight the impact of their theorizing, in terms of whether it enables them to test, elaborate, or generate theory. Our evidence indicates that theoretical claims are more convincing when studies demonstrate sufficient levels of transparency both in their approach to, and representation of, each of these three elements of the research process.

Nonetheless, we are mindful of previously noted concern regarding the unintended consequences of formulaic and standardized quality criteria. Reflecting this, our review of four illustrative studies also suggests the need for some further nuance. For example, while Bate and Saz Carranza and Ospina both provide transparent accounts of research design, they adopt highly differentiated approaches to describing their process of analysis.

Specifically, Bate describes analysis of his action-ethnography in broad procedural terms, noting use of a temporal bracketing strategy, for example.