leadership Week 2 Part II

EXPLORING THE USE OF GROUNDED THEORY AS A METHODOLOGICAL APPROACH TO EXAMINE THE 'BLACK BOX' OF NETWORK LEADERSHIP IN THE NATIONAL QUALITY FORUM A. BRYCE HOFLUND University of Nebraska at Omaha

ABSTRACT

This paper describes how grounded theory was used to investigate the “black box” of network leadership in the creation of the National Quality Forum. Scholars are beginning to recognize the importance of network organizations and are in the embryonic stages of collecting and analyzing data about network leadership processes. Grounded theory, with its focus on deriving theory from empirical data, offers researchers a distinctive way of studying little-known phenomena and is therefore well suited to exploring network leadership processes. Specifically, this paper provides an overview of grounded theory, a discussion of the appropriateness of grounded theory to investigating network phenomena, a description of how the research was conducted, and a discussion of the limitations and lessons learned from using this approach. Keywords: grounded theory, network leadership, health care, network organization, collaboration

470 JHHSA SPRING 2013

It is a capital mistake to theorize before one has the data.

- Sherlock Holmes

The task of scientific study is to lift the veils that cover the area of life that one proposes to study.

-- Blumer (1978)

Generating a theory involves a process of research.

--Glaser and Strauss (1967)

In The Rise of the Network Society (2000), the first

in a trilogy of books about the social, economic, and cultural impacts of the Information Age, sociologist Manual Castells documents the rise of the Information Age. A defining feature of this new age is interconnectedness, which is manifested through the complex networks that are a ubiquitous part of the Information Age. Networks are everywhere; there are, among other things, global business networks, cellular networks, television networks, social networks, the Internet, and computer networks.

In the public sector we also are witnessing the movement away from bureaucratic, hierarchical organizations toward networks. Rubin (2005) argues that the three-branch metaphor for government is outmoded and that the network metaphor more accurately describes government and intergovernmental relations today. Goldsmith and Eggers (2004) note that this shift has occurred for a number of reasons, including an increase in cross-agency and cross-government initiatives, an increase in public-private collaboration, and the growth of the Digital Revolution, which allows for increased citizen demand for and input in service delivery options.

JHHSA SPRING 2013 471

In 1999 the health care industry created the National Quality Forum (NQF), a network organization, whose founding mission was to improve American healthcare through endorsement of consensus-based national standards for measurement and public reporting of healthcare performance data that provide meaningful information about whether care is safe, timely, beneficial, patient- centered, equitable and efficient.

The NQF was created because of all of the failed attempts in health care to make some headway in quality improvement. The NQF is representative of a network organization because it was created to address issues of health care quality in a new way by bringing together organizations from the public and private sectors and providing them with a forum to discuss and debate measures of quality, and ultimately, to effect change. The NQF thus represents a major administrative experiment in addressing health policy issues.

Dr. Kenneth W. Kizer, as the NQF’s first leader, was tasked with building the network. This makes network leadership important. Since there were so few empirical studies into the phenomenon of network leadership, a grounded theory approach seemed to be the most appropriate way to study the “black box” of network leadership. The purpose of this paper is to provide some insight into how to conduct an empirical study of network leadership using a grounded theory approach and considers some of the strengths, limitations, and lessons learned from this application of grounded theory. I argue that grounded theory offers a powerful and promising approach way of studying such social phenomena as network leadership. It is important to note that this paper does not present the findings from this study of network leadership since they have been published elsewhere (see Hoflund & Farquhar, 2008; Hoflund, 2012a, Hoflund, 2012b).

472 JHHSA SPRING 2013

This paper is organized into three sections. First, I discuss the research design and strategy for this study, including a brief overview of the history and nature of grounded theory and its usefulness as an approach for studying networks and leadership. Secondly, I discuss data collection and data analysis. In Part Three, I conclude by discussing some of the limitations and lessons I learned from conducting research using a grounded theory approach.

RESEARCH DESIGN

Since I was interested in exploring the phenomenon

of network leadership, I used a qualitative research design and, more specifically, a grounded theory approach to conduct this study about network leadership during a network organization’s formative stages. A qualitative research design is most appropriate for this study because it provides the best means to explore complex processes and investigate “little-known phenomena or innovative systems” such as network leadership, and it is useful when “relevant variables have yet to be identified” as is the case with the critical tasks related to network leadership (Marshall & Rossman, 1999, p. 57).

A qualitative approach also allows one to describe the “naturally unfolding program processes and impacts” and allows for a certain richness in the research—the participants’ thoughts, opinions, and experiences are captured in their own words—that one may not be able to get through the use of another approach (Patton, 1987, p. 14). That is, a qualitative approach allows one to “lift the veils” surrounding an area of study.

But this does not mean that this type of research does not follow a process. In fact, it is quite the opposite. Grounded theory allows researchers to follow a process that allows for creativity in discovering and understanding

JHHSA SPRING 2013 473

social processes and phenomena (Glaser & Strauss, 1967). I first discuss the origins and philosophical underpinnings of grounded theory and, in the following sub-section, the benefits of a grounded theory approach and justification for why I employed this approach to study network leadership.

Grounded Theory’s Origins and Philosophical Underpinnings

Glaser and Strauss (1967) state that grounded theory is the “discovery of data systematically obtained from social research” (p. 2). Creswell (2003) elaborates on their definition by noting that grounded theory is a strategy “in which the researcher attempts to derive a general, abstract theory of a process, action, or interaction grounded in the views of participants in a study” (p. 14).

Glaser states that grounded theory is useful to “researchers and practitioners in fields that concern themselves with issues relating to human behavior in organizations, groups, and other social configurations” (Glaser, 1992, p. 13). The nature of grounded theory is to ensure that the theory being generated will “fit” the situation being studied and that it will “work” in terms of describing the behavior being observed (Glaser & Strauss, 1967, p. 3). It follows from this, then, that for theory to be useful for understanding social phenomena and behavior, the best way to develop theory is to “ground” it in data.

In using the grounded theory method to develop theory, one begins with an area of study and allows what is relevant to that area to emerge from the data. Two key characteristics define grounded theory: a de-emphasis on the verification of theory and an emphasis on the generation of theory. Glaser and Strauss (1967) proposed grounded theory as a way to counteract the preoccupation with the verification of theory in both qualitative and quantitative research that had dominated social science since the 1940s, to address some of the weaknesses of

474 JHHSA SPRING 2013

qualitative theory, and to allow for the development of theory that would be meaningful to both practitioners and scholars.

Glaser and Strauss (1967) argued that scholars were too concerned with verifying the “grand theories” bestowed on us by “great men” such as Marx, Weber, and Durkheim. After World War II, there was significant growth in the development and distribution of quantitative methods (e.g., survey research) that could be used to test and verify these theories. In The Discovery of Grounded Theory, Glaser and Strauss (1967) offer a polemic against Robert Merton and the positivist approach:

His reasoning necessarily leads to the position that data should fit the theory, in contrast to our position that the theory should fit the data [emphasis in the original] (p. 261).

While grounded theory acknowledges that verification of theory is important, it argues that this task should be subordinate to the generation of theory.

Glaser and Strauss (1967) also proposed grounded theory as a way of strengthening qualitative research. They argued that qualitative approaches suffered from an overemphasis on verification, but more importantly were increasingly labeled as “impressionistic” and criticized for not being rigorous or systematic enough. On the other hand, quantitative methods were seen as rigorous and “more scientific.” As a direct result of this, over time, quantitative methods gradually usurped qualitative approaches to studying and gaining insight about social phenomena.

With the publication of The Discovery of Grounded Theory, however, Glaser and Strauss tried to formalize and systematize “grounded” theory, and qualitative methods

JHHSA SPRING 2013 475

more generally, as a legitimate form of inquiry into social phenomena. What was unique about their approach, however, was that they did not discount the importance and benefits of scientific rigor that had been so lauded in quantitative research:

It is vital to note that the fundamentals of Grounded Theory, the underlying analytic methodology, are in very large measure drawn from the analytic methodology and procedures of inductive quantitative analysis laboriously discovered by researchers and students in the Department of Sociology and the Bureau of Applied Social Research at Columbia University in the 1950’s and 1960’s (Glaser, 1992, p. 7). Perhaps the most important difference to note

between grounded theory and other approaches to qualitative research is grounded theory’s emphasis on theory development. Glaser and Strauss (1967) argue that the growth of positivism and the emphasis on verification of theory rather than generation of theory resulted in a significant gap between theory and empirical research. Theory that was “grounded” in data, they proposed, would contribute toward “closing the embarrassing gap between theory and empirical research” (Glaser & Strauss, 1967, p. vii).

During the past thirty-five years, researchers from a variety of different disciplines, including psychology, information science, education and health care, have used grounded theory as a means of exploring social relationships and phenomena.

476 JHHSA SPRING 2013

Grounded Theory as a Means of Exploring Network Leadership Scholars argue that new methods are required to research and understand new organizational forms such as network organizations. Daft and Lewin (1993) note the trend away from bureaucratic, hierarchical structures toward more loosely coupled, flexible structures that emphasize learning. They contend that in order for managers to function in this new environment and for researchers to understand this new environment, these emergent forms of organization, which include network organizations, require new forms of empirical investigation (Daft & Lewin, 1993). This new form of investigation, according to Daft and Lewin (1993), “will be characterized by midrange theory and method, grounded research, and research that does not presume to test hypotheses” (p. ii). Daft and Lewin (1993) argue that the primary benefit of a grounded theory approach to emergent organizational structures is:

A midrange, grounded study of some part of a new organizational form would enable a scholar to learn firsthand about it and provide new theory. We are proposing a role for organizational scholars that is primarily one of developing new variables and theories to describe new phenomena, not to test hypotheses. If done well, the emerging knowledge will advance both organization theory and the practice of management (p. iii).

More recently, other scholars, including McGuire and Agranoff (2007) and Agranoff (2004), have explored the relevance of grounded theory as an approach to studying networks, noting that employing grounded theory will help

JHHSA SPRING 2013 477

answer some of the “big questions” about network management by allowing researchers to delve more deeply into the “black box” of networks and examine them from the inside out.

Scholars also contend that leadership theory would be enhanced by the generation of theories that are “grounded” in what leaders are actually doing (Parry, 1998). The applicability of grounded theory to leadership has been demonstrated in a number of studies (Hunt, 1991; Hunt & Ropo, 1995). Hunt and Ropo (1995) argue that grounded theory can be effective as a means of studying social processes, such as leadership because “grounded theory emphasizes dynamism, whereas mainstream analysis emphasizes static structure” (p. 381). Therefore, grounded theory allows one to understand the dynamic of “change” as it relates to leadership, as opposed to traditional approaches, which study leadership at one point in time. As a result of the focus on quantitative methods, there have been increased calls issued for more qualitative work on leadership (e.g., Bryman, Stephens, & a Campo, 1996; Parry, 1998).

For this research, the choice of grounded theory as a strategy of inquiry was appropriate for several reasons. First, this is an exploratory study in that its purpose is to generate theory about network leadership strategies that is grounded in empirical evidence. Eisenhardt (1989) notes one of the strengths of grounded theory is it “produces theory which closely mirrors reality” (p. 547). The nature of grounded theory is to move from observations to the development of concepts then to theory development (Locke, 2001). Theory building grounded in empirical evidence promises to contribute to the scholarly literature in public administration and organizational theory, but at the same time be “useful to practitioners in the settings studied, providing them some understanding and control

478 JHHSA SPRING 2013

over situations they encounter on a daily basis” (Locke, 2001, p. 18).

The use of grounded theory also was appropriate because the network leadership literature is underdeveloped; the theories related to network management that do exist were not systematically obtained from observations and may lack validity. Thus, more empirical studies need to be conducted (Agranoff & McGuire, 2001). The emergent theory is more likely to be empirically valid because the theory building process is so closely linked with empirical observations.

DATA COLLECTION AND ANALYSIS

Grounded theory is an iterative process during

which there is interplay among data collection, analysis and theory generation. Glaser and Strauss (1967) refer to this as the constant comparative method of analysis. The idea behind the constant comparative method is that a researcher gathers data, analyzes the data, and compares them against previously collected data in order to determine variables and uncover emerging relationships between variables and categories.

The overlap of data collection and data analysis serves several purposes. First, it allows the researcher to move ahead with data analysis during the data collection stage. Second, it permits researchers to be flexible with regard to things that might emerge from the data. For example, Eisenhardt (1989) notes that it enables one to make adjustments to the data collection instruments that “allow the research to probe emergent themes or to take advantage of special opportunities which may be present in a given situation” (p. 539). Finally, the constant comparative method serves as a source of validity because the process generates further data and knowledge, leading

JHHSA SPRING 2013 479

to theory that is more reliable because it is more clearly defined and less abstract (Parry, 1998).

In the following sub-sections, I discuss how I conducted this research. In keeping with the tenets of grounded theory, I have not separated data collection and analysis into separate sections; instead, I discuss them jointly. I begin by discussing the types of data or “data slices” that Glaser and Strauss (1967) recommend collecting to develop grounded theory. I then discuss the six phases of this research in light of the techniques Glaser and Strauss (1967) discuss and Strauss and Corbin (1998) elaborate on for collecting, organizing and analyzing these data slices and developing theory that is truly “grounded” in the data.

Data Slices: Interviews, Field Notes, Observations and Documents

Glaser and Strauss (1967) advocate gathering “slices of data”—which others refer to as “triangulation”— as a means of understanding conceptual categories from different vantage points. Caudle (1994, p. 89) defines triangulation as “the combining of methods, data sources, and other factors in examining what is under study” in order to determine whether or not they are congruent and/or complementary. The nature of this research was to uncover recurring patterns and to describe the administrative processes, activities, and resources involved in the development of standards in a network setting. Grounded theory, as distinguished from other forms of qualitative research such as phenomenology, demands that researchers consider multiple forms of data (Suddaby, 2006). In order to develop theory that takes into account multiple perspectives and different types of data, I collected data from a variety of sources, including one-on-one interviews, field notes, observations and NQF-related documents.

480 JHHSA SPRING 2013

Interviews. Thirty-nine interviews informed this research. Of these, I conducted semi-structured interviews with 21 individuals who were active in the NQF during its formative years. Of the 25 people I contacted, 19 agreed to an interview, three said no (one gave no reason for saying no, the other two individuals said that s/he was too busy and his/her agency would not allow him/her to be interviewed), two did not respond, and one referred me to another person, who I interviewed. A colleague also was conducting her research about the NQF. Since both of our studies were investigating aspects of the NQF as a network organization and our lines of questioning and interview protocol were similar, she and I shared the interviews that we had conducted and transcribed. She shared 18 interviews with me. I analyzed and coded these as I did my own interviews. Strauss and Corbin (1998) refer to this type of sharing and coding data sets “secondary analysis” and state a “researcher building theory can code these materials as well, employing theoretical sampling in conjunction with the usual coding procedures” (p. 213).

Glaser and Strauss (1967) recommend theoretical sampling of different groups to maximize the similarities and differences of information. Therefore, the interviewees were representative of the diverse organizations that belong to the NQF and consisted of individuals who were involved at all levels of the NQF, including NQF staff members, and those who served on the Board of Directors, the Never Events Steering Committee and each of the four Member Councils (Consumers, Purchasers, Providers, and Research and Quality Improvement Organizations).

The interviewees initially were contacted by phone or e-mail about participating in the study. Whether by phone or e-mail, I introduced myself, provided them with information about the project, and asked them if they would be willing to participate in an interview. The interviews occurred either by phone or in person. Each interview

JHHSA SPRING 2013 481

lasted approximately one hour, and, in many cases, quite a bit longer. In order to ensure anonymity, I have not divulged the names of the individuals I interviewed or mentioned their names or positions in the text. I gathered additional contacts using snowball or chain sampling. In snowball sampling, interviewees are asked to provide other names of individuals who know about the issue (Caudle, 1994).

Different stages of theory development demand different interview techniques (Polit & Beck, 2004; Wimpenny & Gass, 2000). Glaser and Strauss (1967) state that during the initial period of data collection and analysis, interviews may take the form of unstructured conversations and, as the theory begins to emerge from the data, the interviews will become more focused and structured. During the initial phases of my data collection, the interviews tended to be more conversational and broadly focused and I developed an initial interview guide that reflected this. As the theory began to emerge, I developed an interview guide in which the questions were more focused than they were initially. The interviews themselves became more structured. Fielding (1994) notes that some of the strengths of semi-structured interviews are that they allow the researcher to ask questions in the same way each time, while allowing for flexibility in the sequence of questions and the depth of exploration.

Throughout the process, I taped and transcribed each of the interviews after asking a subject for his or her consent. All of the interviewees consented to being taped. After I completed each interview, I transcribed it into an MS Word document and uploaded it into QSR N6. Richards and Richards (1994) maintain that software such as QSR N6, NVivo, and Atlas/ti is essential to maintaining precision and rigor in data analysis. After I uploaded the interview into QSR N6, I began coding it; I elaborate on this process in later sections.

482 JHHSA SPRING 2013

The one-on-one interviews not only allowed me to gather information about the specific management strategies the NQF used to manage the development of standards but also permitted me to observe the body language and tone of voice of the network managers and members and the physical setting of the NQF. From the interviews I gleaned quotes and gathered opinions and information about how the participants interacted within the network. I used “memoing” to record thoughts, interpretations, questions and directions for further data collection (Strauss & Corbin, 1998). These memos were written to explore what was emerging from the data, what I was learning from the literature, and how I linked the two in developing an interview guide and theory.

Field notes. Field notes are an important part of grounded theory research because they allow a researcher to record observations and thoughts about the research process and topic as the research progresses. Eisenhardt (1989) recommends writing down impressions and asking such critical questions as “What am I learning?” and “How does this case differ from the last?” after interviews and observations.

I kept two types of field notes: a set for interviews and a set of notes outlining what I observed at the two NQF Annual Meetings I attended. As part of the interview process, I kept records of notes that I took during the interviews. I also took time immediately after I completed each interview to record my impressions and thoughts about what I learned from the interview. While attending the NQF Annual Meetings, I took notes about the issues discussed at the meetings, differing opinions and who raised them, Kizer’s representation of the NQF, and my reactions to and thoughts about various events and topics discussed.

JHHSA SPRING 2013 483

Written documents. I also gathered and analyzed documents related to the NQF, including working papers about the NQF’s first project—the “Never Events” project—minutes from committee meetings, and briefing materials. In order to gain access to these documents, I contacted the NQF’s staff members and executive officers involved in overseeing the “Never Events” project as well as individuals involved in the development of the consensus report. A confidential source close to the NQF also provided me with many financial and other documents pertaining to the NQF’s creation and the Never Events project. I also collected data from public sources, including newspaper and journal articles, speeches Kizer gave that were available on the Internet and information from the NQF’s web site.

One can learn a great deal about the organizational structure, operations, history and philosophy of an organization through the examination of written documents, and these documents provided me with a strong sense and appreciation of the NQF as an organization. I used these documents to provide me with information about reports I might want to obtain and the individuals I might want to contact for interviews during the initial stages of my research.

Observation. I also attended and observed the proceedings of two NQF Annual Meetings. I attended the two meetings for several reasons. First, they gave me an opportunity to observe first-hand how the NQF conducts business. Second, I was able to meet people and question them informally about the NQF at these meetings. Third, as the research progressed and leadership became the focal point of the study, the meetings gave me a chance to observe Kizer in action and determine whether what I was observing matched with what I was hearing from the interviewees. Finally, the annual meetings gave me a

484 JHHSA SPRING 2013

chance to learn about and keep up-to-date on the various issues affecting the NQF and its operations. The Research Phases: Collecting, Analyzing and Developing Theory

Glaser and Strauss (1967) do not prescribe how to conduct research using the grounded theory method in their seminal work. Strauss and Corbin (1998), however, elaborate on the original work and outline some steps for conducting research using grounded theory: open coding, axial coding, and selective coding. Strauss and Corbin (1998) state that the process is “a free-flowing and creative one in which analysts move quickly back and forth between types of coding, using analytic techniques and procedures freely and in response to the analytic task before analysis” (p. 58). Since I was new to grounded research, I followed the approach outlined by Strauss and Corbin (1998). I conducted this research in six phases and delineate the tasks I undertook for each phase of the data collection and analysis process in the following sections.

Phase one: Initial contact with the NQF and immersion in the health care literature. I began the project by establishing contact with NQF staff and Dr. John Eisenberg, then Director of the Agency for Healthcare Research and Quality (AHRQ)—which provided some of the initial funding for the NQF. I read widely about the health care system, including the history of health care in the United States (Starr, 1984; Millenson, 1997); the history of quality improvement efforts (Brennan & Berwick, 1996); quality problems in health care (President’s Advisory Commission, 1998; Institute of Medicine, 1999, 2001), including research and information about medical errors and patient safety (Bogner, 1994) and quality initiatives underway in the health care industry. I also read about high-reliability systems and human error

JHHSA SPRING 2013 485

(Perrow, 1999; Reason, 1990) and error-reporting systems and quality improvement efforts underway in other industries, including the aviation, nuclear and chemical industries (Rees, 1994; Gunningham & Rees, 1997). Phase two: Literature review and development of initial research questions. During Phase Two, I narrowed the focus of my study to the NQF and its role as a network organization charged with coordinating quality improvement efforts in the health care industry. I also began to read literature about networks and network management. One of the common misperceptions about grounded theory is that a researcher should come into the research as a “blank slate” with no prior immersion in or knowledge about the literature (Suddaby, 2006, p. 634). However, Glaser and Strauss (1967) argue for a link between substantive theory, or the theory associated with a particular subject area, and the generation of grounded formal theory:

We believe that although formal theory can be generated directly from the data, it is more desirable, and usually necessary, to start the formal theory from a substantive one. The latter not only provides a stimulus to a ‘good idea’ but it also gives an initial direction in developing relevant categories and properties and in choosing possible modes of integration. Indeed it is difficult to find a grounded formal theory that was not in some way stimulated by substantive theory (p. 79). From this literature review, I identified the broad

questions that guided the research questions during the preliminary phases of this project: (1) how are these new

486 JHHSA SPRING 2013

organizational forms (i.e., networks) managed? and (2) what is the U.S. national government’s role in this process? While these two questions outlined the broad purpose of this study, I developed the following questions to guide me initially as I gathered specific information about the NQF:

1. How and why was the NQF created? 2. How is the NQF organized? 3. How does the NQF manage the development of standards? 4. What is the federal government’s role in this process? 5. What lessons does NQF’s administrative experiment hold for students of public administration? From my initial literature review and document

collection efforts, I developed a preliminary interview guide that I used for conducting the initial exploratory interviews. Since the process was exploratory, the interview questions served as probes to generate data that I later tied-back to the existing literature during Phase Three.

Phase three: Initial interviews and identification of the emergent themes through open coding. During Phase Three, I conducted, coded, and began to analyze the initial exploratory interviews and continued collecting documents related to the NQF. I interviewed seven individuals involved in the “Never Events” project.

Strauss and Corbin (1998) state that the first step in the process of theory building is the development of concepts. The initial interviews allowed me to begin the process of organizing and interpreting the data. Open coding is the “process through which concepts are identified and their properties and dimensions are discovered in data” (Strauss & Corbin, 1998, p. 101).

JHHSA SPRING 2013 487

During open coding, “data are broken down into discrete parts, closely examined, and compared for similarities and differences” (Strauss & Corbin, 1998, p. 102). The process enables researchers “to group similar events, happenings, and objects under a common heading or classification” (Strauss & Corbin, 1998, p. 103). Researchers can analyze documents by line, by sentence or paragraph, or as a whole (Strauss & Corbin, 1998). From there, categories are identified and their properties and dimensions are specified (Strauss & Corbin, 1998).

To assist me with the coding process, I used QSR N6. One of the strengths of using qualitative software is that it allows the themes to emerge from the interviews. As I coded the interviews sentence by sentence, I began to pull common themes from them and group them into broad categories. As I coded these interviews, the broad theme of leadership and, more specifically, Kizer’s leadership in creating and building the NQF, consistently emerged.

Since the NQF was a relatively new organization, I began to think about the role of a leader in building a network organization. I developed the following question to guide me: “What are some of the key tasks a leader engages in to create a network organization?” Since I had initially started the research with a broad area for investigation, network management, asking these questions effectively allowed me to narrow the scope and focus of my research and to develop a more finely tuned research question. This is in keeping with Strauss and Corbin (1998), who state, “Although the initial question starts outs broadly, it becomes progressively narrowed and more focused during the research process as concepts and their relationships are discovered” (p. 41).

One way to investigate phenomena and develop sensitivity or insight into the data and the concepts being developed is to examine the literature for relevant information (Strauss & Corbin, 1998). Eisenhardt (1989)

488 JHHSA SPRING 2013

explains the importance of looking at a broad range of literature when developing theory:

An essential feature of theory building is comparison of the emergent concepts, theory, or hypotheses with the extant literature. This involves asking what this similar to, what does it contradict, and why. A key to this process is to consider a broad range of literature (p. 544).

Furthermore, Eisenhardt (1989) argues: “While linking results to the literature is important in most research, it is particularly crucial in theory-building research because the findings often rest on a very limited number of cases” (p. 545). It is important to look at two types of literature— those that conflict with the findings and those that agree with the findings. The former allows one to be more creative and groundbreaking, and “the result can be deeper insight into both the emergent theory and the conflicting literature, as well as sharpening the limits of generalizability of the focal research” (Eisenhardt, 1989, p. 544). I therefore looked at the existing leadership and network management literatures to assist me with the initial conceptualization of “network leadership” and the possible tasks a leader engages in to create a network organization.

Comparing the findings to extant literature in a different area with similar findings allows a researcher to tie “together underlying similarities in phenomena normally not associated with each other. The result is often a theory with stronger internal validity, wider generalizability, and higher conceptual level” (Eisenhardt, 1989, p. 544). I started looking at the leadership literature in order to determine which area of the literature fit with my project. Since the interviewees discussed extensively the tasks that Kizer engaged in to get the NQF up-and-running, I decided

JHHSA SPRING 2013 489

to focus on and pull concepts from studies that outline the tasks, activities, and roles of leaders; that is, what it is that leaders actually do and the skills that are needed to accomplish their goals. For example, Selznick (1984) in his classic work, Leadership and Administration, delineates some of the critical tasks a bureaucratic leader might undertake in order to build an organization and its institutional character and culture, and Doig and Hargrove (1987) examine public sector leadership and discuss the leadership tasks undertaken by leaders during the formative stages of organizational development.

Phase four: Refinement of the interview guide and the development of subcategories through axial coding. During Phase Four, I conducted and analyzed additional interviews and attended an annual meeting. After I conducted the initial interviews and started to extrapolate themes, I turned to the literature to help me with the development of a more focused interview guide. The guide I developed covered the following topics: the role of the member organizations in the NQF, the Never Events project, key organizational actors in the NQF’s environment, the NQF’s Board of Directors, the Member Councils, the NQF’s staff, the role of professional expertise in the NQF, Kizer’s role in the NQF, and the NQF’s challenges and accomplishments. I used the in-depth interview instrument of open-ended questions as a guide when interviewing participants.

During this phase, I engaged in axial coding, which is the “process of relating categories to their subcategories, termed ‘axial’ because coding occurs around the axis of a category, linking categories at the level of properties and dimensions” (Strauss & Corbin, 1998, p. 123). The purpose of axial coding is “to begin the process of reassembling data that were fractured during open coding” (Strauss & Corbin, 1998, p. 124). Strauss and Corbin (1998) identify several tasks associated with axial coding:

490 JHHSA SPRING 2013

1. Laying out the properties of a category and their dimensions, a task that begins during open coding 2. Identifying the variety of conditions, actions/interactions, and consequences associated with a phenomenon 3. Relating a category to its subcategories through statements denoting how they are related to each other, and 4. Looking for cues in the data that denote how major categories might relate to each other (p. 126).

I began to ask questions about the larger categories (i.e., the three critical tasks a network leader engages in to build a network organization) that were emerging from the data: defining the mission, building a social base, and managing diverse interests through the consensus development process. These questions allowed me to develop subcategories that explain each category in greater detail. According to Strauss and Corbin (1998), “subcategories answer questions about the phenomenon such as when, where, why, who, how, and with what consequences, thus giving the concept greater explanatory power” (p. 125). Phase five: Refinement of the theory through selective coding. During Phase Five, I attended another annual meeting and engaged in selective coding. Selective coding is the “process of integrating and refining the theory” (Strauss & Corbin, 1998, p. 143). Integration involves organizing categories “around a central explanatory concept” (Strauss & Corbin, 1998, p. 161). Strauss and Corbin (1998) outline several tools that can be used to assist with integration: telling or writing the storyline, using diagrams, sorting and reviewing memos, and using computer programs. After integration, the researcher begins

JHHSA SPRING 2013 491

to refine the theory. “Refining the theory consists of reviewing the scheme for internal consistency and for gaps in logic, filling in poorly developed categories and trimming excess ones, and validating the scheme” (Strauss & Corbin, 1998, p. 156). In this case, I wrote about the leadership tasks interviewees maintained that Kizer engaged in, using the memos I had written to assist me with developing the narrative about the leadership tasks involved in creating an NAO. As part of the integration phase, I compiled my findings into a conference paper. The conference presentation enabled me to obtain reactions to my findings and refine my data further. Phase six: Closure. Strauss and Corbin (1998) encourage the researcher to consider three things when deciding to conclude data collection and analysis: time, money and, most importantly, theoretical saturation. Although the first two issues are self-explanatory, the third deserves an explanation. In order to reach closure, Eisenhardt (1989) maintains that researchers should constantly ask themselves two important questions. “When should I stop adding cases?” and “When should I stop moving between data collection and analysis?” The answer to both is theoretical saturation, which Strauss and Corbin (1998, p. 143) define as “The point in category development at which no new properties, dimensions, or relationships emerge during analysis.” There is nothing new that can be added through further sampling; that is, collecting further information will not enhance the categories and their properties any further. Glaser and Strauss (1967, p. 224) maintain that closure should occur “When the researcher is convinced that his [sic] conceptual framework forms a systematic theory, that it is a reasonably accurate statement of the matters studied, that it is couched in a form possible for others to use in studying a similar area, and that he can publish is results with confidence, then he is near the end of his research.” I

492 JHHSA SPRING 2013

stopped collecting data when I started to hear the same stories and examples from interviewees. I also had a well- developed theoretical framework and found it difficult to collect information that would shed additional light on it.

LIMITATIONS

One concern associated with grounded theory and this study is whether the findings are transferable. While Glaser and Strauss (1967) do not discuss this issue directly, they discuss credibility and state:

“The reader’s judgment of credibility will also rest upon his assessments of how the researcher came to his conclusions. He will note, for instance, what range of events the researcher saw, whom he interviewed, who talked to him, what diverse groups he compared, what kinds of experiences he had, and how he might have appeared to various people whom he studied” (p. 231).

Locke (2001), however, notes that by gathering diverse data observations, the general applicability or analytic generalizability of the theory can be extended.

Another concern is the subjectivity of the researcher. That is, the researcher becomes the primary measurement instrument in the investigative process, in contrast to that of quantitative research where the researcher tries to stay removed from the process (Caudle, 1994). In grounded theory, one must let the theory emerge from the data. This is not an easy task, especially considering that researchers bring their own sets of biases and expectations to research, but an astute grounded theorist recognizes and is sensitive to bias. In order to counteract researcher bias, a researcher needs to present

JHHSA SPRING 2013 493

evidence that corroborates the data (Caudle, 1994). One way to do this is to gather multiple perspectives and documents about the same incident (Eisenhardt, 1989). By doing so, validity is enhanced because one is relying on more than one person (and more than one document) to provide an understanding of the events that occurred. Lincoln and Guba (1985) also recommend that the researcher find someone to examine the research findings and play “devil’s advocate.” In order to address these issues, I asked several individuals to serve as my devil’s advocates.

A third concern relates to the interview process and document analysis. Once interviews are granted, there is a concern with being able to move beyond “scripted” responses in order to get the “real” story. Potential problems related to document analysis include identifying the relevant documents and, once identified, gaining access to those documents. Another concern is whether or not the documents reflect reality. That is, do they accurately reflect decision processes and decisions or were they written to protect individuals? For example, one concern might be that the minutes might not have been written in a way that reflects the actual discussions and debates that occurred. Another concern is whether minutes and memos contain more than cursory information. In order to address these concerns, I collected as many documents as possible and spoke to a wide variety of individuals to verify that the stories I had heard were indeed accurate.

A final concern is with the reliability or dependability of the research (Neuman, 2003). The concern with a study’s reliability can be remedied with replication. In qualitative research, however, nothing remains static; that is, reality is constantly changing, making replication difficult. Furthermore, it is impossible to replicate such things as semi-structured interviews. Qualitative researchers argue that because processes are not stable over

494 JHHSA SPRING 2013

time and the research process itself is supposed to be dynamic the preoccupation of “positivist” researchers with regard to replication is unfounded (Neuman, 2003; Chenitz & Swanson, 1986; Denzin, 1970). Indeed, Chenitz and Swanson (1986) point out that replication is not important to grounded theory. They maintain it is more important that researchers be able to use the grounded theory to explain, understand and predict phenomena in similar situations.

CONCLUSION AND LESSONS LEARNED

Janesick (1998) uses the metaphor of dance to

describe qualitative research. Grounded theory also exemplifies the metaphor of dance in that it is an iterative, creative process, which lends itself to experimentation and exploration of concepts and ideas. During this process, I learned five lessons about using grounded theory to investigate the “black box” of network leadership.

First, grounded theory is not easy to master. There are few prescriptions for how to conduct grounded theory research. In my experience, Suddaby (2006) is correct in observing: “The seamless craft of a well-executed grounded theory study…is the product of considerable experience, hard work, creative and, occasionally, a healthy dose of good luck” (p. 639). Furthermore, many researchers have found that competence in using grounded theory techniques improves over time and with experience (Suddaby, 2006). Learning to use grounded theory techniques requires patience, flexibility, the ability to tolerate ambiguity and time. The constant movement between data analysis and data collection requires patience. Developing grounded theory also demands that the researcher be able to remain flexible and responsive to emerging themes. It requires flexibility in the sense that one must be willing to follow the data’s recommendations and pursue an unintended line of inquiry. A researcher therefore

JHHSA SPRING 2013 495

also must be comfortable with ambiguity in the research process. Since the data drive the direction of the research and the lines of inquiry, grounded theory cannot be “mapped” in advance. Researchers who must “map” the research path ahead of time may have some difficulty conducting research using a grounded theory approach. These characteristics also mean that grounded theory research is time-consuming.

Secondly, grounded theory research requires a process. One of the benefits of conducting grounded theory research is that it leads to fresh insights about the social phenomenon under investigation. Achieving this requires researchers to be intuitive, flexible, and open-minded. This does not mean, however, that when conducting grounded theory research that “anything goes” (Suddaby, 2006; Jones & Noble, 2007). Although I certainly found that there is tension between creativity and the rigorous application of formal rules in conducting grounded theory, the perception that grounded theory is an excuse to throw methodological rigor out the window is wrong. Suddaby (2006) notes that in evaluating grounded theory research, he checks that a researcher has followed the core analytic tenets of grounded theory, including theoretical sampling, constant comparison, theoretical sensitivity, and the technical language a researcher uses to describe the research process is accurate, because he believes “there is a clear connection between rigor in language and rigor in action” (p. 640). Through this research I learned that being transparent about how I collected, coded and analyzed my data is as important in qualitative research as it is in quantitative research.

Thirdly, qualitative software programs are helpful in conducting grounded theory research. A grounded theory approach can leave one feeling inundated by tons of data that can be characterized as thematically diverse. I found that using a software program, in particular QSR N5 and

496 JHHSA SPRING 2013

later N6, helped to counteract the feeling that I was “drowning in data” by providing me with the tools to organize and analyze the data efficiently. Although I ultimately decided how to interpret the data and which categories to focus on, QSR N5 and N6 allowed the categories and themes to emerge from the data. The programs, however, had too many “bells and whistles” that I did not use and was a bit complex for my research needs.

Fourthly, transcribing my own interviews was essential to understanding the data. The more exposure one has to the data, the more familiar it becomes and the more likely the researcher will be able to “listen to” and “hear” what the data are trying to tell her. When one is conducting an interview, one is more focused on asking the questions and guiding the interview than on analyzing what is actually occurring during an interview (tone of voice, body language, etc.). Similarly, if one does not transcribe their own interviews but reads a transcription, one misses “hearing” the interview and the subtle cues and insights that might be conveyed by listening to the interview.

Finally, and perhaps most importantly, grounded theory has contributed substantially to my personal growth as a scholar and researcher. Strauss and Corbin (1998) outline the characteristics of a grounded theorist and emphasize that these skills do not need to be developed prior to engaging in grounded theory research:

• The ability to step back and critically

analyze situations. • The ability to recognize tendency toward

bias. • The ability to think abstractly. • The ability to flexible and open to helpful

criticism. • Sensitivity to the words and actions of

respondents.

JHHSA SPRING 2013 497

• A sense of absorption and devotion to the work process (p. 7).

To this list, I would add intuition. I have always been very intuitive and able to identify themes, and grounded theory enabled me to draw on these strengths. As a new researcher, conducting grounded theory research refined and sharpened my ability to identify and ask broader research questions and connect these questions to the broader scholarly literature in the areas of network management and leadership.

Strauss and Corbin (1998), however, neglect to mention one important aspect of the research process that the grounded theory approach, and qualitative methods more generally, help new researchers develop: developing and designing interview questions and guides and conducting interviews. As a new researcher, this process, with its emphasis on constant comparison between data collection and analysis, helped me to develop and fine-tune relevant questions. Furthermore, when I began this research, I found it difficult and stressful to conduct interviews. With more experience, I became more comfortable with the interview process.

498 JHHSA SPRING 2013

REFERENCES Agranoff, R. (2004, March). Inside the Network Operation.

Paper presented at the annual conference of the American Society for Public Administration. Portland, OR.

Agranoff, R., & McGuire, M. (2001). After the network is

formed: Process, power, and performance.” In M. P. Mandell (Ed.), Getting results through collaboration: Networks and network structures for public policy and management (pp. 11-29). Westport, CT: Quorum Books.

Blumer, H. (1969). Symbolic interactionism. Englewood

Cliffs, NJ: Prentice Hall. Bogner, M. S. (Ed.). (1994). Human error in medicine.

Hillsdale, NJ: Lawrence Erlbaum Associates. Brennan, T. A., & Berwick, D. M. (1996). New rules:

Regulation, markets, and the quality of American health care. San Francisco: Jossey-Bass Publishers.

Bryman, A., Stephens, M., & a Campo, C. (1996). The

importance of context: Qualitative research and the study of leadership. The Leadership Quarterly, 7, 353-370.

Castells, M. (2000). The rise of the network society (2nd

ed.). Malden, MA: Blackwell Publishers. Caudle, S. L. (1994). Using qualitative approaches. In J. S.

Wholey, H. P. Hatry, & K. E. Newcomer (Eds.), Handbook of practical program evaluation (pp. 69- 95). San Francisco: Jossey-Bass Publishers.

JHHSA SPRING 2013 499

Chenitz, W. C., & Swanson, J. M. (Eds.). (1986). From practice to grounded theory: Qualitative research in nursing. Menlo Park, CA: Addison Wesley Publishing Company.

Creswell, J. W. (2003). Research design: Qualitative,

quantitative, and mixed methods approaches. Thousand Oaks, CA: Sage Publications.

Daft, R. L., & Lewin, A. Y. (1993). Where are the theories

for “new” organizational forms? An editorial essay. Organization Science, 4, i-vi.

Denzin, N. K. (1970). The research act: A theoretical

introduction to sociological methods. New York: McGraw-Hill.

Denzin, N. K. (1994). The art and politics of interpretation.

In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of Qualitative Research (pp. 500-515). Thousand Oaks, CA: Sage Publications.

Doig, J. W., & Hargrove, E. C. (Eds.). (1987). Leadership

and innovation: A biographical perspective on entrepreneurs in government. Baltimore, MD: The Johns Hopkins University Press.

Eisenhardt, K. (1989). Building theories from case study

research. The Academy of Management Review, 14, 532-550.

Fielding, N. (1994). Varieties of research interviews. Nurse

Researcher, 1, 4-13.

500 JHHSA SPRING 2013

Glaser, B. G. (1992). Basics of grounded theory analysis: Emergence versus forcing. Mill Valley, CA: Sociology Press.

Glaser B. G., & Strauss, A. L. (1967). The discovery of

grounded theory: Strategies for qualitative research. New York: Aldine Publishing Company.

Goldsmith, S., & Eggers, W. D. (2004). Governing by

network: The new shape of the public sector. Washington, DC: Brookings Institution Press.

Gunningham, N. & Rees, J. V. (1997). Industry self-

regulation: An institutional perspective. Law & Policy, 19, 363-414.

Hoflund, A. B. (2012a, forthcoming). Designing a

Decision-Making Process for a Network Administrative Organization: A Case Study of the National Quality Forum's Consensus Development Process. Public Organization Review.

Hoflund, A. B. (2012b). Building a Network

Administrative Organization’s Social Base: Lessons From the National Quality Forum, Administration & Society, first published on February 16, 2012 as doi:10.1177/0095399711433695.

Hoflund, A. B., & Farquhar, M. B. (2008). Challenges of

Democratic Experimentalism: A Case Study of the National Quality Forum in Health Care. Regulation and Governance, 2, 121-135.

Hunt, J. G. (1991). Leadership: A new synthesis. Newbury

Park, CA: Sage.

JHHSA SPRING 2013 501

Hunt, J. G., & Ropo, A. (1995). Multi-level leadership: Grounded theory and mainstream theory applied to the case of general motors. Leadership Quarterly, 6, 379-412.

Institute of Medicine. (1999). To err is human: Building a

safer health system. Washington, DC: National Academy Press.

Institute of Medicine. (2001). Crossing the quality chasm:

A new health system for the 21st century. Washington, DC: National Academy Press.

Janesick, V. J. (1998). The dance of qualitative research

design: Metaphor, methodolatry, and meaning. In N. K. Denzin & Y. S. Lincoln (Eds.), Strategies of qualitative inquiry (pp. 35-55). Thousand Oaks, CA: Sage Publications.

Jones, R., & Noble, G. (2007). Grounded theory and

management research: a lack of integrity? Qualitative Research in Organizations and Management: An International Journal, 2, 84-103.

Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry.

Newbury Park, CA: Sage Publications. Locke, K. (2001). Grounded theory in management

research. Thousand Oaks, CA: Sage Publications. Marshall, C., & Rossman, G. B. (1999). Designing

qualitative research. Thousand Oaks, CA: Sage Publications.

502 JHHSA SPRING 2013

McGuire, M., & Agranoff, R. (2007, October). Answering the big questions, asking the bigger questions: Expanding the public network management empirical research agenda. Paper presented at the 9th Public Management Research Conference. Tucson, AZ.

Millenson, M. L. (1997). Demanding medical excellence.

Chicago: University of Chicago Press. Neuman, W. L. (2003). Social research methods:

Qualitative and quantitative approaches. (5th ed.). Boston: Allyn & Bacon.

Parry, K. W. (1998). Grounded theory and social process:

A new direction for leadership research. The Leadership Quarterly, 9, 85-105.

Patton, M. Q. (1987). How to use qualitative methods in

evaluation. Newbury Park, CA: Sage Publications. Perrow, C. (1999). Normal accidents: Living with high-risk

technologies. Princeton, NJ: Princeton University Press.

Polit, D. F., & Beck, C. T. (2004). Nursing research:

Principles and methods. New York: Lippincott, Williams, & Wilkins.

President’s Advisory Commission on Consumer Protection

and Quality in the Health Care Industry. (1998). Quality first: Better health care for all Americans. Retrieved October 30, 2011 from http://www.hcqualitycommission.gov/final/

JHHSA SPRING 2013 503

Reason, James T. (1990). Human error. New York: Cambridge University Press.

Rees, J. V. (1994). Hostages of each other: The

transformation of nuclear safety since Three Mile Island. Chicago: University of Chicago Press.

Richards, T. J., & Richards, L. (1994). Using computers in

qualitative research. In N. K. Denzin & Y. S. Lincoln, (Eds.), Handbook of qualitative research, (pp. 445–462). Thousand Oaks, CA: Sage Publications.

Rubin, E. L. (2005). Beyond Camelot: Rethinking politics

and law for the modern state. Princeton, NJ: Princeton University Press.

Selznick, P. (1984). Leadership in administration: A

sociological interpretation (Rev. ed). Berkeley, CA: University of California Press.

Starr, P. (1984). The social transformation of American

medicine. New York: Basic Books. Strauss, A., & Corbin, J. (1998). Basics of qualitative

research: Techniques and procedures for developing grounded theory. Thousand Oaks, CA: Sage Publications.

Suddaby, R. (2006). From the Editors: What grounded

theory is not. Academy of Management Journal, 49, 633-642.

504 JHHSA SPRING 2013

Wimpenny, P., & Gass, J. (2000). Interviewing in phenomenology and grounded theory: is there a difference? Journal of Advanced Nursing, 31, 1485- 1492.

Copyright of Journal of Health & Human Services Administration is the property of Southern Public Administration Education Foundation and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.