Abstract
Introduction
Qualitative research has a dual purpose. On one hand, it seeks to uncover what reality may be, while on the other hand determines how that reality has been arrived at. This dual process provides opportunity for the application, as well as the adaption, of various methods. These enable interpretation of wide-ranging data and information, sourced from multiple perspectives, and assessed through multiple methods to be applied, enabling researchers to understand meaning in the context of life setting scenarios. A key outcome of such research is to understand the what, the why, and the how within research settings, and to apply this to the broader research agenda of both extant and emerging theory.
Grounded theory represents a key qualitative research method. It identifies a range of essential elements that, when combined, offer a consolidated framework within which wide-ranging data are gathered, assessed, and subsequently used in developing theory, based on what has been observed. These essential elements include the coding and categorization of data, concurrent data collection and analysis, the writing of memos, theoretical sampling, constant comparative analysis using inductive and abductive logic, the application of theoretical sensitivity, the development of intermediate coding practices and routines, the selection of core categories from the data, and the application of theoretical saturation.
Applying grounded theory in the manner identified in this article results in the hypothesis that change management in the nonprofit sector may necessitate the inclusion of four key considerations that should be built into planned organizational change programs. In the context of a single case study approach, these include that reflection for both the change agent and the change recipient should be accounted for, that preexisting confidence and trust levels in management are a necessary element, that a balanced focus on both the individual and the organization must be evident, and the sequencing of specified events before, during, and after the change, impact change outcomes (Rosenbaum, More, & Steane, 2016). The grounded theory methodology applied to this research suggests that the inclusion of these elements in existing change management models may, with the aid of further research, support the development of specific change management models for application in this sector.
Qualitative Methodology and the Grounded Theory Context
Many aspects of qualitative research continue to contest the notion of what good research is and what it is not. In feeding this ongoing dilemma, qualitative research appears to have divided itself into two quite broad camps, one which seeks to link its own legitimacy to the positivist world, and the other which seeks to specifically extricate itself from this potentially restrictive set of barriers and be differentiated by focusing on an interpretivist approach. The former seeks to legitimize itself in the eyes of quantitative researchers, and the latter seeks to substantiate qualitative research as a valid alternative, supportive rather than competitive, motivated to achieve additional, and valuable, research outcomes. Researchers have, over the years, identified the challenges of such broad epistemological variations (Angen, 2000; Jardine, 1990; Sandelowski, 1993), and, while qualitative research has grown in application and use (Cooper & White, 2012; Cummings, Daellenbach, Davenport, & Campbell, 2013; Kathleen M. Eisenhardt & Graebner, 2007; Locke, 2011; Morse et al., 2009), debates concerning the methodological paradigms between approaches, and questions about the broader practical use of qualitative research continue (Bansal, Bertels, Ewart, MacConnachie, & O’Brien, 2012; Denyer & Tranfield, 2009).
Recognizing the important distinction between the
While grounded theory has developed along different lines of execution and use since its first application (B. G. Glaser & Strauss, 1970), there has been recognition of the ongoing development and adaptation that each application of such methodology makes to the original theory (Morse et al., 2009). While variation in method, as distinct from methodology, is being applied in the research described in this article, commonly accepted aspects of the recognized methods remain as cornerstones (Tummers & Karsten, 2012). Consequently, the research method applied here ensures the application of true grounded theory as distinct from a descriptive or exploratory research approach (Birks & Mills, 2011).
This article identifies a specific application of grounded theory method to study change management in the nonprofit sector, and in doing so, provides input into the development of a range of prescriptive processes that may guide future researchers and future applications of the methodology. This is consistent with the views expressed by Strauss and Corbin (Denzin & Lincoln, 1994) regarding the potential development of grounded theory in the course of applying it to different research settings: As with any general methodology, grounded theory’s actual use in practice has varied with the specifics of the area under study, the purpose and focus of the research, the contingencies faced during the project, and perhaps also the temperament and particular gifts or weaknesses of the researcher. (p. 276)
Process Overview
As described in more detail throughout this article, the specific application of grounded theory to the study of change management in the nonprofit sector, via a longitudinal case study, was structured in a manner consistent with the principles developed by its founders (B. G. Glaser & Strauss, 1970), as well as those that underpinned variations to the original application (B. G. Glaser, 2001).
Figure 1 identifies an overview of the processes developed and applied by the principal researcher where specific procedures become individualized through ongoing applications of grounded theory to new research settings. This further defines the range of tools that, when applied, may lay another foundation stone in the bridge-building exercise between a process that provides little by way of prescription, and the broad criticism of grounded theory as a means of escaping theory testing (Goldthorpe, 1997, 2000; Mjøset, 2005). To some extent, such criticism may also be fueled by poor appreciation of the role of qualitative research and its importance in identifying new interpretations (Gadamer, 2004; Peshkin, 1993; Sanjek, 1990), which a well-developed and well-applied grounded theory research may offer.
Process overview.
This process of moving from the design phases to the collection and analysis phases is detailed further in Tables 1 to 3.
Research Design Processes.
Data Collection and Construction Processes.
Data Analysis Processes.
Research Design
Table 1 identifies the various subprocesses involved during research design and identifies the activity level for each as well as the linkages back to the broader grounded theory methodology and methods.
The design of the research focused on contributing to change management theory through the use and application of grounded theory, observing that such an approach focuses the research on how change is perceived by those that are affected by the process and recognizing that existing change management models in use do not adequately capture the complexity of the change process from the perspective of change recipients (Balogun & Jenkins, 2003). The original impetus for the research was the need to move beyond traditional methodology used in understanding change management, given that anecdotal evidence revealed how a range of characteristics of people working in this sector may affect how change unfolds and is therefore managed.
This sees the principal researcher applying a constructivist approach to the application of grounded theory, where the emphasis on data is predicated on personal attachment to the research, the role that previous knowledge and experience plays during the research process, and emphasizing the researcher’s role and actions (Bryant & Charmaz, 2007; Denzin & Lincoln, 1994; Goulding, 2009; O’Reilly & Marx, 2012), the reality that data and analysis are social constructs rather than pure objective facts (Bryant & Charmaz, 2007), and the epistemological underpinning regarding how data are constructed and interpreted (Fendt & Sachs, 2008).
The importance of understanding change through the lens of those experiencing it was viewed as a pivotal approach, especially in the context of studying change management where historical approaches have viewed change from an organizational perspective, with limited research regarding the views, thoughts, and feelings of actors immersed in the process (Fox-Wolfgramm, Boal, & Hunt, 1998; Oreg & Berson, 2011; Oreg, Michel, & By, 2013; Smith & Graetz, 2011). In addition, in order to account for the impact of change on organisational actors over time, a longitudinal approach (van den Broek, Boselie, & Paauwe, 2013) was adopted to enable views of their changing environment, and their feelings toward these events, that would impact on the success or otherwise of the change program to be assessed (Bartunek, Rousseau, Rudolph, & DePalma, 2006).
Existing literature and its use remains one of the outstanding issues in the ongoing debate between Glaser and Strauss’ original approach to grounded theory, and the approach subsequently developed by Strauss (B. G. Glaser & Strauss, 1970; Strauss, 1987). Grounded theory espoused an approach that suggested no engagement with existing literature on the research topic prior to any forms of data collection. It was reasoned that such engagement would inhibit the natural emergence of categories from the empirical data (Dey, 2007; Dunne, 2011). Subsequent repositioning by Strauss (1987) and supported by other researchers (Corbin & Strauss, 1990; Wiener, 2007) ensured an ongoing, well-defined debate with grounded theory purists (Holton, 2007; Nathaniel, 2006) who place the literature review at the end stages of the research.
The approach adopted here was to enter the field with a strong knowledge base of existing literature in the substantive area of the research, namely, change management, supported by an in-depth exposure to broad management issues within the nonprofit sector, obtained from the principal researcher’s consulting activities in that sector. This base knowledge was further developed during the data analysis phase as specific research-related issues developed, which focused the attention of the researcher to different areas of literature on related subject matters, designed to expand knowledge in areas considered relevant. This also supported the research question of what can be learned from a nonprofit longitudinal qualitative case study regarding the management of change, which points to key differentiating features of existing, commercial-based models. An extensive literature review was undertaken on grounded theory as a methodology to enable the researcher to effectively design and initiate an appropriate grounded theory method to answer this question.
A single case study (Kathleen M. Eisenhardt, 1989; Orlikowski & Hoffman, 1997; Raelin & Catalado, 2011) was identified as relevant, given the need to source an organizational change program that enabled longitudinal analysis on a before-the-change, during-the-change, and after-the-change basis. This accounts for the temporal aspects of the change program (Beer & Nohria, 2000; Buchanan et al., 2005; Maimone & Sinclair, 2014; Pettigrew, Woodman, & Cameron, 2001; Van de Ven & Poole, 1995; Whittle & Stevens, 2013; Wilson, 1992).
The research was undertaken over a 3-year period in a large nonprofit general hospital undergoing the implementation of an in-house designed E-Pathways system that sought to replace an existing paper-based patient records process with an integrated online pathways-based platform. The research involved data analysis from 56 structured and semistructured interviews (Mossholder, Settoon, Harris, & Armenakis, 1995; Rowley, 2012) and, for triangulation purposes (Kathleen M. Eisenhardt, 1989), a range of hospital-based documentation, as well as information gleaned from attendance at a number of staff meetings, which the principal researcher attended as a nonparticipant observer (Denzin & Lincoln, 1994).
The software program Nvivo 10 was used as a tool to analyze the transcripts. As has been suggested by some researchers (Bryant & Charmaz, 2007), the software application was not the central focus for developing the grounded theory. Rather, the software was merely an aid, or tool, in the process. Theory development resulted from intimate involvement with the data, avoiding the possible risk of displacing personal immersion in the data through a cognitive process, with a detached software driven process that, if fully applied, “extracts” themes from word usage and frequency patterns. A heavy focus and reliance on the software application, for purposes other than maintaining easier access and tracking of large data volumes and cross referencing, could jeopardize the quality of the developed grounded theory.
In this manner, the research deemphasized the role of software in the process and stressed the fluidity and dynamic nature of qualitative analysis (Morse et al., 2009). This was further emphasized by appreciating that the outcomes of each interview were dependent on a number of key attributes of the interviewer, including preexisting knowledge and how that would be applied, levels of sensitivity that are brought to the interview, and the ability to apply empathy toward the interviewee during the conduct of the interviews (Kvale, 1996), thereby gaining their trust and eliciting meaningful data. This underpins what has been described as the “discovery tradition” of field research and stressing the creativity involved in data interpretation (Busi, 2013; Locke, 2011).
Data Collection Supported by Ongoing Data Construction
Table 2 identifies the various subprocesses involved during data collection and identifies the activity level for each, as well as the linkages back to the broader grounded theory methodology and methods.
Interviews were undertaken at the hospital’s premises as nurses and allied health staff were either commencing their shifts or completing them. All interviews were conducted in staff rooms, offices, hospital cafes, or vacant ward rooms, depending on the shift that the interviewee was completing or about to commence. With permission of each of the interviewees, a recording device was used from which transcripts were created and used for purposes of data analysis. In this manner, rapport with interviewees ensued, and while the pure neo-positivist approach to interviews was largely rejected, in line with the principal researcher’s epistemological view of data gathering and data generation, the straddling between a “romantic” and “constructionist” approach was more evident (Rowley, 2012).
Interviews were semistructured, which maximized the breadth of interactions between the researcher and the interviewee (Denzin & Lincoln, 1994), and well supported in a grounded theory methodology (Goulding, 2002). This approach remains consistent with approaching interviewer and interviewee interaction from a constructionist perspective (Rowley, 2012). In addition, in researching such an organizational process as change management, interview-based approaches have been widely applied in qualitative studies (Mossholder et al., 1995).
Interviewees included representatives from all ward nursing staff, ward nursing managers, allied health professionals, managerial staff involved in designing the E-Pathways system as well as those tasked with implementing the system, including the internally designated change agent. Members of the hospital’s executive team, including the chief executive officer, were also interviewed. This wide source of interviews ensured that interview data were obtained from an array of those who were impacted by the changes, directly and indirectly. In keeping with theoretical sampling principles, these interviews developed an iterative framework as data analysis informed ongoing interviews.
The areas covered by the contents of the semistructured interviews, focused on the phases through which the organizational changes associated with the E-Pathways implementation went through, as well as the areas within the hospital from which the interviewees originated. In this manner, there were a series of semistructured questions that reflected the longitudinal nature of the research and related to the stages of “before-the-change,” “during-the-change,” and “after-the-change.” In addition, there was a different focus of interview questions for different interviewee groups, namely, executive-level team members, manager-level team members, and staff-level team members. Each group’s semistructured interviews reflected their different levels of responsibility and experience with regard the implementation, and therefore the associated change program. In this manner, purpose and structure were closely interlinked (Cassell, 2009).
Recognizing interviews as being the predominant source of data (Denzin & Lincoln, 1994), in keeping with the traditions of grounded theory, the concept of data was widely applied (B. G. Glaser, 2001) and related to data extracted from interview transcripts, from organizational documentation, from attendance by the researcher at numerous hospital staff and working party meetings as a nonparticipant observer, and from general observation of hospital staff interactions throughout the hospital during the period of the change program. In this manner, data triangulation (Trent, 2012; van den Broek et al., 2013) supported data analysis, which also directed ongoing interviews.
Data Analysis
Table 3 identifies the various subprocesses involved during data analysis and identifies the activity level for each, as well as the linkages back to the broader grounded theory methodology and methods.
All interviews were recorded and transcribed using a professional transcription service to cope with the volumes of data that were obtained during the course of the interviews, which resulted in 360 pages of interview data, all of which provided the depth of rich descriptions and explanations to support the emerging theory and much of the raw material necessary in the discovery process (Bryant & Charmaz, 2007; Miles & Huberman, 1984; R. Walker, 1985). Interview data also included the researcher’s attendance at the numerous staff and other in-house meetings attended as a nonparticipant observer, where recordings were also made of all proceedings (with approval of participants and meeting chairs). These transcripts were also prepared in a similar fashion to the one-on-one interviews. Transcripts were further enhanced by the inclusion of documentation reviews, with the latter incorporated into WORD documents for ongoing analysis. When combined, the interview data, the meeting notes, and the documentation review notes resulted in more than 400 pages of raw data to be applied in the overall interpretation and theory development process.
Coding of data was undertaken using Nvivo software versions 9 and 10 (updates applied as available throughout the research process). Using Nvivo terminology, Nodes (Codes) were identified a posteriori, and in the process, descriptions applied for each newly identified Node. These descriptions laid the initial foundations for detailed memos that were used in redefining the Nodes as more and more data were coded that underpinned the constant comparison of data and Nodes. This led to Nodes being initially created in large numbers as more and more data were coded, followed by Node changes and rationalization as previously coded data were reviewed, Node relationships were identified, themes extracted and the formation and definition of levels of Nodes that enabled theoretical sampling and, finally, theoretical saturation, leading to theory development.
While memos provided the framework and glue that enabled the process to evolve, as identified in recognized grounded theory method (Birks & Mills, 2011; Bryant & Charmaz, 2007; Corbin & Strauss, 1990), there was a heavy reliance on the development of a detailed research log, referred to by some as a research diary (Newbury, 2001), and in other cases extended to be known as a reflective journal (Ortlipp, 2008), as distinct from participant diaries which often support data collection and interpretation (Jacelon & Imperio, 2005). The principal researcher applied more of a reflective component to its development and maintained this through the theory development stage of the research, applying the research log in ways that identified his own experiences and values. This approach supported the way chosen to represent the research findings (Harrison, MacGibbon, & Morton, 2001), while also utilizing the log to effectively “think-through” the more challenging aspects of qualitative research, and grounded theory more specifically, in areas of constant comparison, theoretical saturation, theoretical sampling, validity, and transparency.
The overall process has been detailed diagrammatically in Figure 2 and focuses on a range of process issues. These were designed to inform a prescriptive aspect to the “doing” of grounded research, while maintaining the cognitive component. Such a structured and integrated approach ensures differentiation between grounded theory and a descriptive exploratory research perspective (Birks & Mills, 2011).
Node interrelationship map.
Figure 2 identifies the bottom-up approach that was developed in applying grounded theory (remembering the use and application of Nvivo descriptors which were mandatory). An explanation of terms developed and used in this process are as follows:
LRM = Linkage Review Memos, being the memos identified in grounded theory method. The term
RL = Research Logs, being the document created that identified an activity or observation for which detailed ideas were documented with high levels of detail, and supported by their levels of importance and action points that required consideration.
NIML1 and NIML2 = Node Interrelationship Map Level 1, being the assessment of the first level Nodes that resulted from the first level coding of interview transcripts and other data gathering sources. These were structured using a purpose designed Excel spreadsheet format. When combined with the LRMs and the RL, these developed into NIML2 (Node Interrelationship Map Level 2), which operated at various levels including, supporting constant comparison and theoretical saturation, given the emphasis on bio-direction between them. These essential components of grounded theory were, accordingly, applied in this formal circular movement to eventually evolve to a higher level, being the HLDC.
HLDC = Higher Level Descriptive Characteristics, being the result of the above interactions which sought to identify the characteristics identified in the data, structured in the form of Node Groupings (NG), informed by the LRMs and the RL, and developed into key aspects of theory which evolved inductively from the data.
The process began with the Nvivo coded interview transcripts (“NCIT”), which informed the Linkage Review Memos (“LRM”). The LRMs were created from the beginning of data collection, document reviews, and observational opportunities. As informed by grounded theory application, these became the linchpin of the research. The LRMs were constantly updated by, and referenced in, the Research log (“RL”). The LRMs, supported by the RL evolved into two levels of Node Interrelationship Maps (“NIML1” and “NIML2”). Level 1 maps identified the first level of Node relationships that were identified in the data and supported by detailed comments made in the LRMs. These LRMs were created around a four-level structure which sought to develop the researcher’s views around general thoughts regarding the Nodes, the relationships between other existing Nodes, the potential creation of new Nodes, and considerations regarding further analysis in existing literature.
The NIML1 and supporting comments in the RL led to the flow of information into the NIML2, which grouped Nodes into broader categories that was supported by the ongoing analysis of the data. Continued assessment of the LRMs in the context of the RL identified the linkages between Nodes, which became the foundations of the NIML2. This ongoing iterative process enabled the creation of both higher level and lower level Nodes, which underpinned theory development through a foundational structuring process of Node Groupings (“NG”) that linked a range of Nodes to their Higher Level Descriptive Characteristics (“HLDC”). A secondary attachment to a further list of potential Nodes in the NIML2 were also identified in this process, and were used to reinforce the theory through both positive and negative correlations. This was identified in the LRMs throughout the process.
The structure in Figure 2 illustrates the links between data and the construction of theory, via the use and application of memos (LRMs and RL) and the iterative process of comparing data with categories and categories with categories (NIML1 and NIML2), in recognition of the importance of constant comparative analysis as a cornerstone in true grounded theory. As categories are developed and refined, theoretical sampling provides further input into the process to support theory creation (NG and HLDC). At this point in the process, theoretical saturation of categories is achieved when no further developments can be identified in category linkages (NG) or category refinement (HLDC), supporting the essence of grounded theory as developed, and reinforcing the utility of the method.
This research method identifies the unique attributes of a constructivist approach to grounded theory. It highlights variations in the position of the researcher, where reality is not independent of the researcher, and where multiple realities and nuances of data construction exist. This is distinct from data collection, as well as the construction of theory rather than the discovery of theory, and the earlier recognition of extant literature around the subject area (Aminian, Kirkham, & Fenn, 2013).
Use of Literature
While grounded theory method is the focus of this article and the processes identified in Figure 2, the use and timing of literature in grounded theory research has been one of the issues that has loomed large as a focal point of dissention among the three broad grounded theory “camps” of its original founders (B. G. Glaser & Strauss, 1970), its refiners (Corbin & Strauss, 1990), and its revisionist (Charmaz, 2006). The use of existing literature in grounded theory has been described as a “polemic” and “divisive” (Dunne, 2011) issue among experienced researchers, and a potential for confusion for less experienced researchers and those considering its use in PhD research.
This is a challenge of timing (Cutcliffe, 2000; McGhee, Marland, & Atkinson, 2007), not of application. However, it results in a range of theoretical and practical considerations. In the context of the founders of grounded theory, the key issue was the extent to which a review of existing literature at the onset of the research would effectively inhibit the emergence of categories naturally from the empirical data, owing to the impact of extant theoretical frameworks and their related hypothesis (Dunne, 2011). The approach adopted by Corbin and Strauss (1990) placed less emphasis on the process issues of grounded theory and viewed procedures as tools that help the researcher to build theories grounded in data. In this manner, the literature could support the development of the research question. The approach adopted by Charmaz (2006) was even more practical in that it considered researchers having ideas about the literature around the subject area as a possible vantage point which would add value to the process of theory construction.
The approach adopted in the present research was impacted by a range of factors including the professional background of the principal researcher which involved practical management experience in change, as well as a practical history in the management of change within the nonprofit sector, and, specifically, within a hospital setting. This led to the proposition of using literature during the process of data analysis as well as during theory construction. As referenced earlier,
Recommendations for Further Research
As a single case study, the hypothesis developed should be tested in further nonprofit settings in order that a more generalized application of the findings can be tested. To support such an approach, further research, which could be undertaken and is currently being considered, would include the following elements:
A longitudinal grounded theory study in a second nonprofit hospital undergoing similar change in order, magnitude, and type as was undertaken in this research. Such a study would seek, in a similar manner to the present study, to understand change from the perspective of those experiencing it, with outcomes being compared with those achieved in the present study. This would provide clarity as to the hypothesis derived in the present study and determine both its veracity as well as potential other intervening factors that could expand it.
A longitudinal grounded theory study in a nonprofit organization outside of the hospital sector in an organization that is undergoing major change of the order of magnitude experienced in this present study. This would provide clarity as to the extent to which the hypothesis developed here is applicable to a broader nonprofit context or the possibility that it may be more specific to a hospital setting.
Finally, a longitudinal grounded theory study in a for-profit organization within the hospital sector in an organization undergoing similar change. Such a study would seek to determine the extent to which the hypothesis identified in the present study is in fact unique to the nonprofit sector through a process that has had limited application in previous change management case study research.
Conclusion
While this article identifies a specific approach to grounded theory research in a unique application, it does so from the perspective that
The historical roots of grounded theory evolved from the early dominance of research settings that reflected the strong quantitative ideological framework. This dominance saw qualitative methodologically based research viewed in adverse terms (Johnson, Long, & White, 2001) and relegated as subsidiary to more scientific based quantitative methodologies. Responding to this positivist paradigm, grounded theory developed as a means of generating theory from real-world data, evolving over time, to incorporate a wider application of methods, in a world that has since come to value the addition to knowledge that qualitative research can bring, resulting in part, from the rigor of analysis as well as the richness and depth of interpretation (D. Walker & Myrick, 2006).
The grounded theory approach to both collecting and analyzing interview and related data supported an understanding of how change recipients as well as change agents, involved themselves in, and responded to, a wide range of organizational processes and dynamics that characterized the changes at the hospital. As a recognized qualitative research methodology, its unique inductive underpinning supported a depth of understanding through the thoughts, feelings, responses, attitudes, and emotions of those involved in, and affected by the processes, over its duration. In the context of a growing appreciation in recent change management literature as to the importance of the individuals in change, as distinct to just the organizational focus, grounded theory supports the depth of analysis and understanding to be achieved by seeing change through the eyes of those who are experiencing it, planning it, and managing it, and doing so in the context of an iterative data analysis approach that is sensitive to the gradation within that data and its meaning.
The research outcomes resulting from the application of grounded theory methodology, and supported by the specifics of the method identified in this article, to the study of a single case study in the management of change in the nonprofit sector, supported the hypothesis identified earlier regarding the identification of four key characteristics that necessitated an increased focus in planned organizational change. These were the inclusion of formal reflection time for change agents and change recipients during the change program, the development of trust and confidence in the organization prior to the actual change commencing, ensuring an appropriate focus on the individual experience of change rather than an entirely organizational focus, and the identification of a range of sequencing events necessary throughout the change program. By outlining specific grounded theory design characteristics focused on researching change from the perspective of change recipients, this study has further developed grounded theory as a research method, while identifying a number of change management activities that may underpin successful change management within the nonprofit sector, thereby contributing to the theory and practice of change, in a sector that has not been the subject of such research to date.