Case
Abstract
Classic grounded theory involves the generation of a theory from data, while remaining open to the ideas emergent from the data in question. This case study details a classic grounded theory study into countering violent extremism practitioners. After providing some context for the study in question, this case study details the significant elements that mark a grounded theory study (here specifically a classic grounded theory as opposed to another version of grounded theory), in particular coding, constant comparison, theoretical sampling, and memo writing. Outlining the ways in which these procedures were approached in this study, some of the difficulties along the way are also considered. Ultimately, grounded theory offers a creative, open, and flexible set of procedures for a researcher to follow. It is focused on participant perspectives, conceptualizing rather than describing their main concern and the manner in which they go about resolving this concern. As such, it moves beyond individual perspectives to the identification of underlying patterns, raising these perspectives to an abstract level of conceptualization. It is the benefits of grounded theory, its openness in particular, that can cause difficulties for the novice researcher. Some of these are detailed here, along with practical advice and suggestions for dealing with these issues.
Learning Outcomes
By the end of this case, students should be able to
- Demonstrate an understanding of the principles underlying the classic grounded theory methodology, and the manner in which these are applied within a classic grounded theory study.
- Identify and evaluate classic grounded theory research.
- Differentiate classic grounded theory from other research methods, and compare the different approaches.
- Discuss the benefits of using classic grounded theory and identify some of the possible difficulties that may be encountered over the course of a classic grounded theory study.
Project Overview and Context
The research project in question focused on individual practitioners working to counter violent extremism at grassroots level. The idea that it is possible to intervene and prevent people, particularly young people, from being influenced by violent extremist organizations to commit acts of violence has led to the growth of what has been described as a “countering violent extremism (CVE) industry.” Those seeking to intervene in or reverse the “radicalisation process” range from government officials and offices to Silicon Valley corporations, from educators, non-governmental organizations (NGOs), and private companies to individuals seeking to exert some influence. In this study, the focus was on those working at grassroots level, individuals who work independently of government and policy makers; they are not part of the “official” system.
From the outset, the research in question was exploratory. I wanted to use a methodology that allowed for issues to arise, as opposed to approaching the project with a framework, or hypothesis in mind. Grounded theory is not a methodology typically associated with studies in the area of international relations, so it was important in the write up to outline the reasons for choosing this approach. The rationale for using grounded theory was influenced by the research topic itself and the aims of the study. Glaser and Strauss (1967/2008), the founders of grounded theory, suggest that the approach taken should be based on methods “best suited to the socially structured necessities of the research situation” (pp. 233–234). In this case, grounded theory suited the study itself, and I also felt that grounded theory suited me, as a researcher; it is open ended and flexible, leaving the researcher to follow leads, engage with a variety of data sources and have issues of importance emerge through the research process. Coupled with the need to bridge the gap between policy and practice, wherein CVE is firmly rooted, and academia, classic grounded theory, by focusing on the concerns of the participants, felt like a good step in this direction.
Grounded Theory: A Journey Through Discovery and Emergence
Early data analysis played a significant role in shaping the research focus of the study in question, highlighting the importance of discovery and emergence within classic grounded theory, with the researcher being led by what is happening in the data. It is through this “mandate to remain open to what is actually happening” (Glaser, 2004), coupled with the systemic procedures offered by grounded theory, that the researcher can “discover the main concern of the participants in the field and how they resolve this concern” (Glaser, 2004—emphasis in original). The goal of grounded theory, notes Glaser (1978), “is to generate a theory that accounts for a pattern of behaviour which is relevant and problematic for those involved” (p. 93). Via the researcher’s engagement with the systematic procedures of grounded theory—data collection, coding, constant comparison, memo writing, theoretical sampling, and arrival at theoretical saturation—the theory emerges and is built. Further data collection is driven by the emerging concepts that will form part of the final theory. This happens throughout the research process; as ideas emerge from the data, that is, as the researcher engages with the data and pulls it apart seeking to understand what is going on in the data, further data sources are identified based on the likelihood that they will contribute to the development of the emerging categories and theory. Ultimately, good ideas “earn” their way into the theory through emergence or emergent fit (Glaser, 1978, p. 8).
A number of key decisions in this study were made based on initial data analysis. In early conversations around this time, it soon became clear that one of the significant occurrences in the area of CVE was the ever-increasing number of actors becoming involved in this area and the idea of the growth of a CVE industry. The growing emphasis on the use of social media for work in this area was also apparent. The decision was taken to interview a variety of these actors, the aim being to explore and compare the experiences of a variety of actors working to counter violent online extremism.
These early interviews, along with some informal meetings and conversations I had with those working in the area of CVE highlighted the large gap between the different actors. While they may all claim to be working to counter violent extremism online, their main concerns were very different. This was coming thorough strongly in the data, and was apparent very early on. Around the time of these interviews and discussions, I attended a Grounded Theory seminar with Barney Glaser in Mill Valley. The discussion there echoed the concerns I was having regarding the participants for the study in question being too disparate; Dr. Glaser suggested that I focus on one group of participants. Glaser and Strauss (1967/2008) suggest that “when beginning the generation of a substantive theory, the sociologist establishes the basic categories and their properties by minimising difference in the comparative groups” (p. 56).
In addition to drawing on early interview data and informal conversations, CVE policy documents were reviewed and analyzed at this stage of the research. Policies in both the United Kingdom and United States promote the idea of localism and refer to the unique position of local communities to recognize the threat of violent extremism. Dolnik (2013, p. 3) observes that much of the research within the field of terrorism studies relies on a government perspective which brings its own biases, with research skewed by the comparatively easier access to government data and the one-sided nature of research funding. The comparison of interview data with policy data and relevant literature at this stage of the research also contributed to the decision to focus on grassroots/community-based CVE practitioners.
Another issue that came up repeatedly in the data at an early stage was the narrow focus within CVE policy and the reporting of CVE issues on the Muslim community. Based on this analysis, I felt that those working to counter varieties of extremist messaging from a grassroots or community-based level should be included in the study. This would provide the opportunity to see if the codes and categories coming up thus far would be similar, or if the variety of extremism involved would see changes in the patterns within the data. Glaser and Strauss (1967/2008, p. 47) explain that “with theoretical sampling the question is: what group or subgroup does one turn to next in data collection? And for what theoretical purpose?” Here, I felt that turning to those working in the same area, but tackling a different “extremist ideology” would add to the dimensions of the categories that were emerging from the data. Having followed the early codes and concepts and considered where to go to next, the participants of this study emerged as those working to counter violent Islamic extremism, right-wing extremism, and nationalist/loyalist extremism in the context of Northern Ireland.
Research Practicalities
Ahead of interviewing any participants and starting this project, I had to apply for ethics approval. Given that grounded theory requires being led by the concerns of the participants, it is difficult for a researcher, at the outset of a study, to be sure about where they will need to go for the data they require and what they will need to ask. In this instance, the participants of the study were not a vulnerable population. They were adults being asked about their day-to-day work. In fact, in this study, the participants were very keen to go on the record about their work. They wanted to be identified and have their words ascribed to them in the write up. Initially, I applied for ethics approval solely for the first interview and, as the research project evolved I applied again. This time, the approval sought was to cover any of the participants I might end up interviewing. This was not an onerous task given that I did not require approval from outside the university, and there was no great risk to the participant group in question. In this second application, I included a schedule of possible questions, noting that the interviews would be open-ended and unstructured. Interview questions in a grounded theory study are changed and revised based on the emergent ideas, evolving as the study progresses.
Having received ethics approval, I began to contact those who best fit within the parameters of the study: grassroots CVE practitioners. The initial interviews offered an entry point into the process of data collection using purposive sampling. Through the analysis of data as it was collected, decisions about where to go next for further data collection were made. These connections were made in a number of ways. Some of the participants in question are increasingly becoming well known for their work in this area and have websites and social media pages through which it was easy to contact them. In other instances, I met people at various events, such as conferences or other meetings to which I was invited who put me in touch with people they knew. Participants introduced me to colleagues and acquaintances with whom they felt I should speak. In all, I interviewed 30 practitioners based in the United Kingdom, Northern Ireland, the Republic of Ireland, Sweden, and Denmark.
In addition to contacting and interviewing the individuals in question, there were practicalities around travel and interview locations. As many of those I interviewed don’t work in typical office jobs, interviews were conducted in a variety of public venues. These included a variety of coffee shops, a park bench (Minneapolis, Minnesota), a skate park (Des Moines, Iowa), and a McDonald’s (Chicago, Illinois). Getting to the venues was something that I had to consider. In the United States, for example, I did not have access to a car and, on one research trip it was snowing. In a couple of instances, I had to rely on participants to collect me and bring me to the interview venue. While this worked well for me and I had no significant issues, who you are meeting and where, depending on the study undertaken, is important to consider, particularly where safety might be a concern.
Method in Action
Grounded theory is, as Glaser (1998, p. 13) describes, a “package.” Within this package, the method is clearly laid out, offering a systematic process for the generation of a relevant theory. It is through data collection, coding, and analyzing through memoing, theoretical sampling, and sorting to writing, using the constant comparative method that a theory with fit and relevance is arrived at (Glaser, 1998, p. 13). Key is to use the complete package of grounded theory procedures as an integrated methodological whole (Glaser, 2004). Remove any of the elements listed and the product cannot be called a grounded theory. This section outlines briefly these procedures.
Open Coding to Selective Coding
After the first interview took place, I began open coding. This involved asking the following questions of the data (Glaser, 1998, p. 140): What is this a study of? What is the participants’ main concern? How do they resolve their main concern? and What is happening in the data? These questions are important in focussing the coding onto behavior, which is the unit of analysis in grounded theory. The emphasis should be on what people do, not what they think. This process is not simply about labeling the pieces of data that are there, but it is a way of capturing the data; codes conceptualize the patterns within the data.
Initially, this proved difficult and I found I was “fracturing” the data, breaking it up too much and ending up with hundreds of codes at a time. I learnt that this is not an unusual issue for the novice grounded theory researcher (Holton, 2007, p. 276). By revisiting the data and remaining focused on the coding questions as opposed to worrying about missing something I soon began to notice patterns in the data which I could label appropriately. The process also became easier once I had further interview data to draw on.
Grounded theory is built on concepts, these concepts being the ideas that emerge through the coding process. Concepts, in grounded theory, should be analytic—“sufficiently generalised to designate characteristics of concrete entities: not the entities themselves,”—and also sensitizing, “yield[ing] a meaningful picture, abetted by apt illustrations that enable one to grasp the reference in terms of one’s own experience” (Glaser and Strauss, 1967/2008, pp. 38–39). By comparing incidents and memoing around them, initial codes in this study were grouped together into concepts. As the study progressed, following the procedures outlined here, some concepts took on a higher-level and became categories. A category captures the underlying patterns in the data. The properties of a category serve to outline the category and its dimensions.
Selective coding in grounded theory takes place once the core category has been identified. The core category accounts for the way the main concern of the participants is continually resolved. In this study, the core category was mining the personal. Selective coding offers a way of focussing the analysis and delimiting the study.
Constant Comparative Analysis
Constant comparative analysis is a key element of grounded theory, adding to the rigor of the methodology and ensuring that the systematic procedures are adhered to. At the beginning, I was comparing incidents (an example of the participants’ behavior in seeking to resolve their main concern) within the early interview data, which limited me to some extent. I left a very lengthy gap between the first interviews conducted. While on the one hand this allowed me to learn more about using grounded theory, it also slowed the ability to start comparing incidents across different pieces of data. As this analysis continued and further interviews were conducted, emerging concepts were compared to new incidents which allowed me to elaborate on the categories that I was creating. Holton (2007) summarizes the purpose of constant comparison as follows: “to see if the data support and continue to support emerging categories. At the same time, the process further builds and substantiates the emerging categories by defining their properties and dimensions” (p. 277).
Memo Writing
Helpful throughout the coding process and the constant comparison process, and central to any grounded theory study is the use of memos. Memo writing is, suggests Glaser (1978) “the core stage in the process of generating theory, the bedrock of theory generation” (p. 83). It proved in this study to be a very valuable way of engaging with the data. While memos are also used in general inductive qualitative methods, in a grounded theory study memos are more focused on “theoretical” rather than “substantive” richness (Hood, 2007, p. 161). Here, they are used to develop theoretical categories and their properties. It is the type of memoing rather than existence of memoing that sets a grounded theory study apart from other qualitative studies (Hood, 2007, p. 161).
Throughout the process of coding, constant comparison, and theoretical sampling, I wrote memos based on anything and everything that was emerging from analysis of the data. This continued throughout the final write up of this work. It was, to quote Glaser (1978, p. 86), never over. At the beginning of the study I felt, surprisingly, somewhat burdened by the idea of memo writing. This was, in part, due to the vast number of codes I was generating at the very beginning of open coding; to memo every code in the first interview would have resulted in a very large number of memos, and I was assuming the same for subsequent interviews. Moving to thinking about memos as capturing a moment, and the fact that a memo can be anything (it can take on different forms) or be about anything was freeing—freedom being a central goal in memo writing (Glaser, 1978, p. 85)—and allowed me to focus on ideas rather than keeping track of a process. “Memos,” notes Lempert (2007), “especially early ones, are often messy and incomplete, with undigested theories and nascent opinions. Ideas may be represented in fragmented phrases, in weird diagrams, half sentences, or long treatises” (p. 249).
Theoretical Sampling
Theoretical sampling works closely with constant comparison, and is the way that the grounded theory researcher decides where next to go for further data collection. This is guided by the emerging theory, and is focussed on what data will add to its further development. Through theoretical sampling, emerging codes, concepts, and categories are expanded. This includes identifying their properties and relations up to the point where a code is “saturated, elaborated and integrated into the emerging theory” (Glaser, 1978 p. 36).
In this study, emerging codes led to the focus on those working to counter a variety of extremisms and working in different policy contexts. Theoretical sampling also influenced the interviews as the study progressed. While I continued to start out in the same way, once codes, concepts, and categories were emergent, I asked about these in further interviews. In October 2016, a number of interviews were conducted in Denmark and Sweden. These were used to verify emergent categories and to add to them, but also to consider any further variation in these categories. In addition to using theoretical sampling to drive data collection and gather new data, I also spent a considerable amount of time going back to data that had been collected earlier in the study. Given the significant time lapse between interviews, I returned to early data as categories became more elaborate, and ways that they might fit together became clearer, allowing for further verification of fit.
Theoretical Saturation
Theoretical saturation marks the point where no further variations on a category are emerging form the data. It is important to note that this is not the same as repetition, rather, no additional data are found to develop the properties of a given category. As the researcher sees similar instances over and over again, the researcher becomes empirically confident that a category is saturated (Glaser & Strauss, 1967/2008, p. 61). With a grounded theory study, the adequate theoretical sample is judged based on how widely and diversely the groups for saturating categories were chosen, according to the type of theory under development (Glaser & Strauss, 1967/2008, p. 63). Saturation and “knowing” is always provisional, and the potential that next participant will add some new property always remains (Cutcliffe & McKenna, 2002).
Theoretical Coding
Theoretical coding involves identifying and conceptualizing the relationships between substantive codes. It is this interaction between substantive and theoretical coding which “characterises GT as an analytic inductive research method rather than empty journalism” (Glaser, 1998, p. 164). A chapter in Theoretical Sensitivity is devoted to the discussion of possible theoretical codes. Adding to this is the 2005 Grounded Theory Perspective III: Theoretical Coding. Here, Glaser lists additional “New Theoretical Codes,” (pp. 17–30). It is important to remember that while these works offer lists and examples of theoretical codes, a theoretical code cannot be forced on the data; the theoretical code should emerge and fit the theory being developed. There are, states Glaser (2005, p. 17), “literally hundreds of TCs that are yet to be found, named and available to GT generating.” The theoretical code is, in most cases, implicit within the emerging theory, bringing “the fractured substantive story turned into substantive concepts back into an organised theory” (Glaser, 2005, p. 11).
Practical Lessons Learned
The use of classic grounded theory in this study, a methodology that is not widely used in the fields of terrorism studies or international relations more generally, offered a way of looking at the area of CVE from a different perspective. Through this focus on grassroots CVE practitioners, the study offers insights into what is happening on the frontlines of CVE practice. In detailing the main concerns of those working in the area much can been gleaned about the field as a whole. In this case, the actions of the participants come about in response to the frustration—their main concern—they experience as a result of what they believe to be poor policy and practice on the part of those operating at governmental level.
The methodology itself did allow for a very open and flexible approach. The path of the study altered in response to the data collected. Rather than focusing on what I, as a researcher, believed to be of significance in the area of CVE, I was able to focus on the concerns of those working in this area. While this openness exists, there is also a set of principles to adhere to and, in doing so, there is in built rigor within the research outcomes.
In some ways, those elements that make grounded theory so useful can also be difficult to navigate for the novice research. It is certainly a very experiential approach; to understand grounded theory fully, you need to do grounded theory. This involves a lot of trial and error and a willingness to deal with much confusion along the way. It is important to understand that this confusion is part of the process and happens with any grounded theory study. Grounded theory is not a linear process and, as such, it can be difficult to know where to go next. That is something important to be aware of from the outset.
The confusion can also make it difficult to explain your research along the way, particularly those unfamiliar with the methodology. This can be an issue in postgraduate research where often you are expected to present your research to colleagues and at conferences. Because grounded theory is not typically used in my field, I found that one way around this was to present on the methodology itself and what it offers, finishing by detailing some of the emerging categories yet emphasizing that these might change considerably. In other instances, for example, at conferences where the focus was on the substantive area, I used the wealth of interview data I had to present the themes that were common to the participants. This allowed me to present on my research, while not having to worry about detailing the theory itself, which was still a long way from coming together. This is typical in grounded theory, and it is often only in the write up that the theory comes together.
On a practical level, I found that throughout the confusion returning to the data and to the memos helped. Often looking back, I would either see that yes, where I was heading did work and were a good fit. Looking back on the data in light of the developing categories provided a new way of looking at incidents within the data and helped confirm the properties of categories. I also used different ways of putting things together, for example, diagrams, index cards and color coded charts. I found that looking at what I had in different ways—even simply a visually different way—helped me make sense of things.
Perhaps one of the most difficult steps in a grounded theory study is moving beyond description to the level of theory. In the case of the study in question, I felt that the participants all had very rich stories about their own lives and how they came to be involved in CVE work. I wanted to include this in the write up. However, this meant a trade-off between the amount of description included and the level I was able to bring the theory up to. This was a decision I made that had consequences, but one that I believed suited the way in which I wanted to write up my research. There are choices such as these to make along the way, and ultimately, as a researcher, you need to understand why you have made this choice and the purpose it serves to the research in question. Grounded theory offers space for this, particularly given there is no set way to write your research up.
Central to completing a grounded theory study is learning to trust the process. The steps are clearly laid out. If you stick with these you will have a grounded theory study. However, while muddling through the confusion, this can be difficult to remember. I was slow to commit to ideas around the main concern and core category. Had I felt more confident in identifying the main concern of the participants, selective coding would have helped in this process. Rather than delimiting the study at an earlier stage, I was left with large amounts of data to go through throughout the process. The fact that I was transcribing lengthy interviews did not help; I stopped doing this and moved to field notes once I felt more sure that I would not miss anything by doing so. While delimiting this study happened later than is probably ideal, a reflection of the uncertainty associated with the process when undertaking a grounded theory study as a novice researcher, in an area in which it is not widely used, in hindsight, it is clear that it works and that trusting a little more in the procedures would help to streamline the data at an earlier stage and avoid the data overwhelm that can happen so easily.
One of the key pieces of advice when undertaking a grounded theory study is memo, memo, memo. The writing of memos proved to be so important and helped in various ways, whether understanding issues I was having with the methodology, or elaborating ideas based on initial coding. In addition, it is important to pace yourself. Coding happens immediately after each interview, and you need time to think in between interviews about where the data are leading. Of course, this has to be balanced against practical issues. For example, I traveled to conduct many of my interviews, and had to do them while I was there. This sometimes meant having more than one interview in a day. While not ideal, it did not cause any major problems for the study.
Finally, it is worth remembering that theory, including any that might be generated through the grounded theory process, is not a final product. A theory can mean different things across different disciplines and for different people. For Glaser and Strauss (1967/2008), theory offers a means of handling data, enabling prediction, describing behavior, and offering understanding for practitioners and is a process, an ever-developing entity as opposed to a “perfect product.” Taking a grounded theory approach, the concern is not with discovering reality, or an element of reality, or a particular individual’s or group’s reality. “Grounded theory is not findings, but rather is an integrated set of conceptual hypotheses. It is just probability statements about the relationship between concepts” (Glaser, 1998, p. 3). As such, I continue to consider and develop my work continuously.
Conclusion
Classic grounded theory offers an open and flexible approach to research, while ensuring rigor through the clearly outlined principles of the methodology. Once you adhere to these principles, and trust the approach, you will arrive at a theory that is grounded in data. Trusting the process can be difficult for the novice researcher; there is a lot of confusion throughout a grounded theory study. Learning to tolerate this confusion is part of doing a classic grounded theory study, and, once that is achieved it is a very rewarding approach to take. Classic grounded theory offers a way of moving from description to abstraction, theorizing based on the data collected. As such, it offers a different perspective to many other methodologies and is very useful when trying to see what is actually happening in a given area, and for those working in the area, as opposed to preframing what could or should be happening.
Exercises and Discussion Questions
- In your own area of interest, can you identify how or where classic grounded theory and theory generation might have something significant to offer? What are the benefits of theory generation, what does theory offer?
- How does the conceptualization associated with classic grounded theory offer a different way of considering a problem or issue?
- What are the benefits of using classic grounded theory in a research project?
- Consider ways to approach the completion of funding proposals/obtaining ethics approval when using grounded theory. How is classic grounded theory different from other methodologies when completing the associated paperwork?
- What are the strengths and weaknesses of classic grounded theory?
- Can you identify similarities and differences between classic grounded theory and other approaches?
- What are the core elements that make a study a classic grounded theory study? How are these used?
Further Reading
Web Resources
Grounded Theory Online: http://www.groundedtheoryonline.com/
Grounded Theory Institute: http://www.groundedtheory.com/
The Grounded Theory Review. An International Journal: http://groundedtheoryreview.com/
The Literature Review in Grounded Theory: http://www.youtube.com/watch?v=7S1kJ0k3yHk
Grounded Theory is the Study of a Concept: http://www.youtube.com/watch?v=OcpxaLQDnLk
