Abstract
Qualitative research for theory building is gaining increasing prominence in management research. With the growing complexity of organizations and their experience of continuous flux, suitable approaches to create action-oriented knowledge are emerging. The underexplored aspects of interpretive ontology and epistemology are being adopted more and more by contemporary scholars. The nascent state of the field is a challenge, on one hand, but an opportunity to become innovative, on the other. The article narrates one such attempt of devising a qualitative research approach by synthesizing from multiple perspectives and addressing the issues of philosophy, methodology and methods. The three issues are taken for granted in quantitative positivist research as the methodologies and methods are built on a positivist philosophy. However, it becomes a challenge in a qualitative research as methodologies and methods have been built for the philosophy of both positivist and interpretive thinking, an issue often not so explicit. Moreover, the existing resources on qualitative research have a partial appeal as they lean in favour of either one of the three issues—philosophy, methodology or methods. One needs to be creative in developing a holistic approach by selecting, integrating and operationalizing all the three issues.
Introduction
Qualitative research for theory building is gaining prominence in management research. With the growing complexity of organizations and their experience of continuous flux, suitable approaches to create action-oriented knowledge are emerging (Donmoyer, 2000). The underexplored aspects of interpretive ontology and epistemology are being increasingly explored by contemporary scholars. The nascent state of the field is a challenge, on one hand, and an opportunity to become creative, on the other. Though there are multiple perspectives prevalent, they are inadequate in one way or the other and few attempts have been made at integrating them together in meaningful ways.
The article narrates one such attempt of devising a research approach by synthesizing from multiple perspectives and extending that synthesis to create knowledge. It highlights how existing perspectives can be leveraged in creative ways. Though the research was started with a question and a framework and methodology based upon popular methodological prescriptions, the final research approach emerged through interaction with data, self and the extant theory. The subsequent sections in the article narrate the journey and the process through which the synthesized approach emerged.
The Research Question and the Initial Approach
The core focus of the research study was to understand the ‘interaction of a CIO 1 (the agency) with the context (the structure) and the consequent implication on the role effectiveness’. It emerged from the literature review that a gap existed in understanding of what determines a CIO’s role effectiveness. CIO role has been widely studied in terms of the functions and the competencies required for performing those functions. Yet, gaps regarding the effectiveness of the CIO in his role existed in terms of understanding the relevance and impact of the organizational context so that the CIO role could be developed holistically (Earl & Feeny, 1994; John, 2006; Lepore et al., 2000; Rockart et al., 1982).
The nature of the question, complexity of the phenomenon and the newness demanded that a case study approach be adopted (Yin, 2009). It was decided to adopt a multiple case method, with a sequential conduct of the case studies. Both within and cross-case analysis to develop the theory were anticipated. A multi-pronged data collection strategy was adopted, which included personal interviews with the CIO and other members of the top management team (TMT), secondary and published data collection and personal observation. In total, there were 56 personal interviews conducted across four case study organizations, with each interview lasting for an average of 60 minutes. The personal interviews and secondary data formed a total of 300 plus pages, which were transcribed and analyzed.
An initial framework to guide the research in terms of design, data collection and analysis was developed based on the works of Giddens (1984), Sewell (1992) and Callero (1994).
Many Perspectives (Yet Little Integration and Operationalization)
The research approach evolved as the research progressed. The evolution happened through three broad phases—‘initial design’, ‘in-course extension’ and ‘full-frame development’.
Initial Design
The initial research approach was guided by the works of Pettigrew (1997), who advocated a recursive interaction between inductive and deductive logic to build the process theory. However, it was soon realized that its operationalization required the support of other perspectives like the methodology of conducting case study research proposed by Eisenhardt (1989) and Yin (2009) and methods proposed by Miles and Huberman (1994), Glaser and Strauss (1967) and Charmaz (2006). Pettigrew (1997) himself confesses at the beginning of his article on processual research that it has been written as a free-flowing expression of his approach to research being employed in his work over three decades. It inspires the reader to adopt a similar approach but does not spell out a clear set of guidelines as to how to go about doing it. The initial choice of perspectives, honestly speaking, was based upon their popularity. Popularity makes defence possible, but considering the perspectives in a law-like manner inhibits any scope for improvisation in research approach (Lee, 2004).
In Course Extension
The need for improvisation soon emerged with the completion of the first case. The first case was analyzed with the support of perspectives, methodologies and methods from Pettigrew, Eisenhardt, Yin, Charmaz, Staruss and Corbin. The inadequacy of them as standalone resources for a qualitative researcher was soon realized as one was confronted and had to deal with the debate of ‘story’ (Dyers & Wilkins, 1991) versus ‘constructs’ (Eisenhardt, 1989). Understanding of a phenomenon (like the one taken for the CIO role study) can be accomplished by adopting two extremes. One, by expressing the phenomenon as a narrative case study and interaction with the narrative which shall bring forth the theory (Dyers & Wilkins, 1991); or, two, it can be reduced into well-defined constructs and the relationships between them (Eisenhardt, 1989).
It was found that both the extreme perspectives are being captured in a single frame by Van de Ven and Poole (2005) as ‘strong process’ and ‘weak process’ approach. The Van de Ven and Poole paper very clearly highlights the interaction of choice of approach as process or variance and the way one views an organization as made up of things, which change with time, or a free flow of events and actions. This is nothing but essentially the ontological and epistemological stance one may adopt. However, even Van de Ven and Poole present a broader philosophical issue without specifying the methodology and methods to be used.
It was soon realized that in order to create an appropriate research approach for processual research using qualitative case studies, one needs to select, integrate and operationalize from the available perspectives. This is a creative process resulting in a unique design based upon unique needs of a particular research.
With the understanding from the popular works as discussed above, it started becoming clear that a qualitative researcher needs to clearly address the issues of philosophy (the ontological and epistemological stand), methodology (clear steps) and the methods (the tools used for data collection, collation and analysis). The three also need to be internally consistent in terms of their underlying assumptions and orientations (Figure 1). None of the perspectives available in the literature provide a holistic integration of the three. It is undoubtedly a challenge, but at the same time it is also an opportunity to improvise and create one’s own approach.
It is this awareness of the lack of a unified and integrated approach to qualitative interpretive research that paved the path for the third phase of full-frame development. Before the third phase of evolution, the full-frame development which is discussed in detail, it is quite pertinent to present the various existing perspectives and approaches being explored from the literature and their relative strengths and weaknesses. An extensive literature review, with respect to the philosophy, methodology and method issues with respect to qualitative research, brought into the fore the works done by Giddens (1993), Weick (1991), Van de Ven and Poole (2005), Tsoukas and Chia (2002), Pettigrew (1997), Eisenhardt (1989), Dyers and Wilkins (1991), Donmoyers (2000), Yin (2009), Strauss and Corbin (1990), Charmaz (2006), Pentland (1999) and Miles and Huberman (1994) (Table 1).
Multiple Perspectives: Relative Strengths and Weaknesses
All of these in their own ways provide support to the approach required for qualitative, processual and case-based research. However, on their own, they lean towards either one of the three issues of philosophy, methodology or methods. For example, Giddens, Weick, Van de Ven and Poole and Tsoukas and Chia highlight the philosophy issue (the ontological and epistemological stance), Eisenhardt and Yin bring forth the methodology issue (as a set of sequential steps to be adopted) and Charmaz, Miles and Huberman, Pentland and Strauss and Corbin give methods of data collection and analysis. Pettigrew straddles across methodology and method whereas Dyers and Wilkins and Donmoyers straddle across philosophy and methodology.
The review of these perspectives highlights bipolarity on several aspects that need to be reconciled. In the current research, dilemma occurred in terms of several bipolarities (as shown in Figure 2). How they are reconciled is a creative process. The dilemma was resolved by either selecting one pole as against the other or by integrating the two poles and further operationalizing the selection and the integration.
Therefore, there are three issues of philosophy, methodology and methods and three decisions of selection, integration and operationalization. The three decisions to select, integrate and operationalize are considered to be relevant for all the three core issues but are more prominent for the issues of philosophy, methodology and methods respectively. The three decisions map closely with the philosophical, methodological and methodical issues involved (as depicted in Figure 3).
One may make different selections based upon one’s inclination and the research question in mind. What one selects from highlights the leveraging of the existing perspectives. But what one selects and the way one integrates and operationalizes indicate ingenuity, improvisation and creativity.
The selection, integration and operationalization for the CIO role research study is as depicted in Figure 4.
The Output of the Reconciliation Process
With this, we return to the third phase of the evolutionary process, the full-frame development.
Full-frame Development
The full-frame development explicates the three fundamental issues of philosophy, methodology and methods and the three decisions of selection, integration and operationalization. The discussion that follows is very specific to the research study on CIO role effectiveness, yet it demonstrates how one can leverage existing perspectives through selection and creative integration and operationalization of them to address a specific research question.
Philosophy: Selection
Let us first start with philosophy, an often-neglected issue in many domains of research (Lee, 2004). A discussion of philosophy essentially involves deliberations around ontology and epistemology. Though the popular notion is that there exists an irreconcilable divide between positivism and interpretivism and hence the need for a clear choice in favour of selecting one, a few attempts at integrating the two have also emerged (Lee, 1991; Van de Ven & Poole, 2005).
A philosophy of social construction of reality and the researcher as an interpretive agent was adopted for this study. This is consistent with Donmoyer’s (2000) methodological perspective of knowledge creation as an enhancement in one’s schema through engaging with the research context. In this philosophy, the researcher interprets based upon the interpretation of the actors involved in the organizational flow of events, their individual experiences and actions (Giddens, 1993; Weick, 1989). The organization evolves along with the flow of events. This is also aligned with the ‘strong process’ approach of Van de Ven and Poole (2005). The ontological stance has been based upon the view of an organization as a flow over time rather than being made up of things.
The key research question ‘how does the interaction of a CIO (the agency) with the context (the structure) impact role effectiveness?’ demanded that the issues of structure and agency be defined as per the interpretive philosophy adopted for the study.
With such ontological stance, the first discussion has to be around understanding the structure. It must be made clear that the clarity of the stance did not emerge by mere perusal of the extant literature. While conducting and analyzing the first case study, the difficulties faced with the abstract nature of rules and resources (Giddens, 1984) and the search of a better conceptualization resulted in the ontological clarity. The knowledge of ‘research approach’ also emerged as a result of the researcher’s empirical experience of the first case study.
This was further built by a deeper appreciation of Giddens’ (1984) new way of looking at structure (as rules and resources), which not only subsumes all the past definitions of structure (Turner, 1987) but also gives them an additional edge to explain the aspects of power, change and agency. In an era where change is the norm, such a theoretical perspective becomes very relevant. However, ‘rules and resources’ existence is so deep that it does not become a part of the human consciousness easily, that is, one cannot easily verbalize what rules and resource combination one is dealing with. Giddens has given another way of understanding them as the social system, a manifestation of the deeper aspects of rules and resources combinations, building over time and empirically verified in space and time (Giddens, 1984; Sewell, 1992).
The experience of the first case study suggested that this empirical existence of structure is in the form of structural conditions of the social system and interactions amongst the agents in the backdrop of those structural conditions. The structural conditions are the institutionalized aspects of the structure (routinization and regionalization in Gidden’s terms), such as, formal structure and positions, practices, processes, norms and beliefs. Structural conditions render resources, give them meaning and define the enabling and constraining nature of the structure. They shape the interaction amongst the actors, based upon the actors’ individual experience, mutual meaning creation and acting upon those meanings. These interactions result into structural conditions at a later point in time. Thus, the mechanism through which structure and agency create each other becomes evident. This is how the recursivitiy between form and flow takes place, thus cocreating each other. Hence, though form is an integral part of the process, it shapes and is also shaped by the flow. This is depicted in Figure 5.
Hence, structure is defined as the mutual creation of form and flow, that is, the ‘structural conditions of the social system’ and the ‘interactional conditions of the social system’, respectively.
Agency, on other hand, is not just a psychological entity or a social entity but a ‘situated social entity’. It experiences the structural conditions, acts and shapes the interactional conditions and hence plays a central role in the flow of events over time. The situated social nature of an agency is the true outcome of the interaction between agency and structure.
Methodology: Integration
The next is the methodology issue. Since knowledge creation was attempted through appreciation of Donmoyer’s (2000) argument based on Piaget’s schema theory, Donmoyer argues in favour of knowledge creation through an individual schema’s capability to assimilate, accommodate, integrate and differentiate information. Therefore, a researcher creates new knowledge by immersing into the research context and building a richer schema over time (about the phenomenon).
The knowledge thus created can then be made available to others in the form of an opportunity for them to immerse in a context (e.g., through a case study or a story). What can also be done beyond this is that the theory can be abstracted in the form of key constructs and their mutual interplay. Hence, an integration of ‘story’ and ‘constructs’ was attempted. The four stories give an opportunity to vicariously experience the cases of the four CIOs in their respective contexts and the abstraction gives a lens through which any organization can be seen through and its story be created. This also integrated tacit learning (reading the stories and experiencing) with explicit learning (understanding the abstracted theory in terms of key constructs and their interplay).
Pettigrew’s perspective of recursive cycle of applying deductive and inductive logic served as a good link between the philosophy adopted and Donmoyer’ methodological requirements. The recursivity between the deductive and inductive logic is a creative process, which weaves new theory out of what is already known and what becomes known through the empirical data (Pettigrew, 1997).
The upper part of Figure 6 explains the cyclic process integrating the deductive and the inductive logics. It started with a broad research question. An exploration of extant literature around that question enabled a guiding framework to take shape. This deductive framework guided the researcher in empirical data collection and induction from the data.
By triangulating the data, the researcher’s inner world (of metaphors, imagery, subjectivity) and the extant literature’s important themes were identified, which helped create the theory as description (the story) and explanation (the abstraction) regarding the emerging phenomenon. Once it was done, new research question(s) emerged. Another cycle of the process refined the theory by either validating it or extending the understanding of the phenomenon.
The lower part of the diagram explains the process as it evolved over the four case studies. This process was repeated until theoretical saturation was achieved at the end of the fourth case. The fourth case, though was unique in many ways, did not throw any new dimensions to the understanding of the phenomenon obtained until the third case. Thereby, this explained the methodology of conducting research, which selected appropriate perspectives from literature and integrated them into a whole.
The various perspectives related to methods as appropriated for synthesizing the methodology shall be highlighted in the subsequent discussion on methods. One aspect that deserves a special mentioning here is that of internal consistency in the three issues in terms of their underlying assumptions and orientations.
The philosophy and the methodology were operationalized through 11 steps using selected methods (Figure 7). The following discussion will clearly demonstrate what methods were selected and operationalized. One has a plethora of options here and a careful search of literature will help one identify the best fit for a given research question and the philosophical and methodological choices made.
Methods: Operationalization
The methods were operationalized in 11 steps. Step 1 dealt with conducting the literature review, stating the research question and developing the initial framework. The framework was developed based upon the works of Giddens (1984), Sewell (1992) and Callero (1994). Figure 8 gives a visual representation of the research question, ‘how does the interaction of a CIO (the agency) with the context (the structure) impact role effectiveness?’
Data were collected through personal interviews with TMT members, researcher’s observations (such as unspoken words during the interviews, the organizational conditions and researcher’s sense-making) and archives (such as emails, published reports and Internet search).
Data from all the sources were transcribed and entered in an Excel sheet. The relevant chunks from the transcribed data were granted codes (Charmaz, 2006; Miles & Huberman, 1994) for them to be processed subsequently. A small piece of verbatim data from one of the interview transcripts and the codes developed from that is shown in Figure 9.
The identified codes were suitably placed under broad heads of structure/context, role, action and role effectiveness (as demonstrated in Figure 10). This was done by referring to the initial research framework with which the research work was started. The codes under each head were clubbed together and the content analyzed for common themes (similar to factor analysis with quantitative data).
The codes under the broad heads of structure/context, role, action and role effectiveness were analyzed for identifying the underlying themes, which would explain the large number of codes.
For example, in case 1 as shown in Table 2, the codes were explained by the theme of ‘adhocism/lack of planning processes’.
Similarly, various other themes were identified, which together described the structural conditions of the social system. These conditions were identified through a ground-up process of data collection, coding and content analysis. The salient themes identified in step 5 (Figure 11) were further developed by referring to the extant literature on the topic and the researcher’s subjective world, thus providing a fuller description and explanation associated with the theme (Figure 12). Weick (1989) calls it ‘disciplined imagination’, which brings forth the importance of subjective sense-making by the researcher as an important element in theory building.
Deriving Salient Theme from Codes
First, the initial framework, and second, the exercise of arriving at the description and explanation based on marrying inductive and deductive approach served the purpose of axial coding (Charmaz, 2006; Miles & Huberman, 1994). The extant literature was also not taken mindlessly, but it was done based on how well it served the purpose of interpretation by the researcher. The researcher’s sense-making of the grounded data, the stories and metaphors he had in mind while making sense of the data played an important role in the process. Descriptions and explanations based on this inductive and deductive interpretation by the researcher provided the inputs for creating the case understating.
However, before doing so, another step was undertaken: Writing memos on different aspects of the phenomenon, which not only served as the analytical handle (Charmaz, 2006) for case description and analysis but also an important aspect of cross-case comparison (Figure 13).
The memos were further built upon each subsequent case for the thread to emerge. In total, six memos emerged related to the phenomenon of agency structure interaction and impact on role effectiveness.
At this stage, it becomes relevant to discuss how data bias was handled. It was done by exploring consistency and inconsistency across data points, both within and across data sources. Consistency would refer to similar perspective from multiple respondents and across different data points. Consistency would also refer to maintaining one line of argument by the same respondent.
An example of consistency across multiple data points was observed in case 1. This was related to ‘business pressures’ mentioned by many respondents, which was also supported by the published figures of business performance highlighting the lack of working capital. Hence, there was consistency across respondents and across data (interviews and published figures).
Similarly, an example of inconsistency was observed around the CIO’s role effectiveness in case 3, where except for two individuals—the heads of human resource (HR) and accounts—all rated it as high. A deeper exploration revealed the personal equations between the CIO and the heads of accounts and HR, their personal ambitions and the related anguish.
Hence, data collection through multiple sources helped determine the level of consistency or the lack of it (Eisenhardt, 1989). Once the ‘consistency’ and ‘inconsistency’ in findings were identified, efforts were put towards consolidating and cohering around consistency and further exploring inconsistency for probable explanations.
Steps 7, 8 and 9 are explained here in brief because in limited space it may not be possible to give a detailed version. At this stage, a grounded up view of the themes most relevant to the phenomenon under study were identified. Their identification was based on integrating the three aspects of grounded data, self as an interpreter and the existing theory. These themes can be termed as ‘loosely defined constructs’, which served as the handles for narrating the story, that is, producing the case description. The story involved the actors, their experiences, their relations and interactions, organizational outcomes, temporal progression of events and amidst these was embedded the CIO’s role and its effectiveness.
Case description, analysis and explanation, though shown as distinct sequential steps, cannot be separated by a clear line of divide in reality. They both support each other recursively and ideally and should be seen as a whole rather than distinct parts. The case description, analysis and explanation provided the basis for bringing forth the refinement in the initial framework and abstraction of the theory. A sequential approach to refinements and abstraction over the four case studies helped identify the key dimensions of the phenomenon and provided the opportunity to validate and extend it by comparing and contrasting the phenomenon over the four case studies. For example, the role of the apex (the CEO or the managing director (MD)) that remained subdued in the first case came to the forefront in the second case being central to the phenomenon and got further reinforced in the third and the fourth case. Similarly, the presence of change mechanisms as resource was conspicuous by its absence in case 1 but became more apparent in the subsequent cases.
For the sake of brevity, the case descriptions, analysis and explanations are not presented here; only the refinements in the initial framework over the four cases (Table 3) and abstracted theory are presented.
In step 10, once the four case studies were conducted and theoretical saturation achieved, the final abstracted theory was developed. The abstraction of theory included identification and definition of three important constructs: (a) structural resourcefulness, (b) interactional functionality and (c) role outcomes and explanation of their interplay (Figure 14).
Subsequent Refinements in the Initial Framework
The explication of the phenomenon of agency structure interaction and its impact on role effectiveness was followed by a discussion highlighting the research study’s important theoretical contributions such as giving power and change more explanatory edge in role theory than given earlier; integrating action in role theory by giving the agency more space to play and; the chief executive officer’s (CEO) importance in building the structural resourcefulness whereas the CIO’s importance in determining the interactional functionality for effectiveness in role and developing a diagnostic framework of the CIO role in its context so that appropriate interventions can be designed for enhancing its effectiveness.
The discussion of theory here may give an impression of a weak process, that is, focus on constructs rather than the story. In reality, descriptive case studies provide for the strong process side of the research. Various other outputs, which can be further created based upon the study (e.g., teaching cases and books) can provide the opportunity to the reader to immerse into the context and vicariously experience the phenomenon. In the limited space of a research article, they cannot be demonstrated.
Conclusion
Organization researchers are increasingly adopting qualitative methods for studying newer dimensions of the phenomenon emerging amidst constant change. Unlike the positivist stream of quantitative methods, the progress in the approaches for conducting qualitative research suffers from lack of holistic frameworks and issues of clarity on ontology and epistemology. Moreover, the proof of a good theory in qualitative research is in designing and adhering to the process of research. Unlike quantitative studies, one cannot rely on the statistical tests to prove the reliability and validity of the study.
The real challenge while designing the process or the approach for processual, qualitative case study research is in grappling with multiple perspectives, each one of which lean towards limited issues to be addressed. Through this article, it has been highlighted that three issues need to be explicitly dealt with before one selects the specific methods or tools for the research: The issue of philosophy or the ontological and epistemological stance, the broad methodology for research and the specific methods or steps to be followed.
There are umpteen numbers of perspectives, methodologies and methods or tools available in the extant literature but one needs to select them based on their fitment with the research question and their consistency with each other. The real task while doing this is to select between alternatives, integrate the seemingly opposite poles and operationalize the implicit concepts.
The article demonstrates one such attempt where the research approach was developed by suitably addressing the issues of philosophy, methodology and methods. The perspective presented in the article should not be limiting in terms of using it as a ‘prescription’ for doing processual, qualitative case research for studying organizational phenomenon, but it should inspire researchers to devise their own methods by leveraging what is available and by creatively synthesizing them into a unique approach.
The real test of good qualitative research is inherent in the rigor with which one devises, defines and adheres to the process of research. The article presents an attempt to do so for producing a strongly defensible research.