Discovering Glaser: My Experience of Doing Grounded Theory

Lesley Margaret Piko, Australian National University

Abstract

For my investigation into how general practitioners (GPs) experience their medical careers, I used a grounded theory methodology based on the early work of Glaser and Strauss (1967) and Glaser’s subsequent work (Glaser, 1978, 1998). Glaser (1998) suggested you need to do grounded theory in order to understand the methodology. I found as I engaged in the process I began to understand the meaning of Glaser’s teachings and to discover Glaser. In this article, I share my experience of discovering my theory of optimising professional life. This theory explains how GPs develop their professional life in response to their need for sustainment—a need that encapsulates self-care to sustain wellbeing, work interest to sustain motivation and income to sustain lifestyle.

Keywords: Career development, sustainment, optimising professional life, general practitioners, grounded theory.

Introduction

Glaser (1998) suggested you need to conduct grounded theory (GT) in order to understand the methodology. I found that as I engaged in the process, I began to understand the meaning of Glaser’s teachings and to discover Glaser. In this article I share my experience of how I discovered a new career development theory, the theory of optimising professional life. This theory explains how general practitioners (GPs) develop their professional life in response to their need for sustainment—need that encapsulates self-care to sustain well-being, work interest to sustain motivation and income to sustain lifestyle. I used GT based on the early work of Glaser and Strauss (1967) and Glaser’s subsequent works (Glaser, 1978, 1998).

Sampling

My early sampling was based on a perspective of GP careers, beginning with young GPs, to hear about the formative years of medical life. I then interviewed experienced and mid-career doctors working in different professional roles and locations. Their need to self-care, be interested in their work, and earn income emerged as catalysts for career events. Guided by Glaser’s writings, I used a sample of GPs for career events that added to the properties and dimensions of these concepts. The sampling continued until similar incidents were repeated and no new concepts were emerging from which I concluded a category was saturated and sampling for that concept could stop.

Data Collection

Sampling and data collection for this study occurred through interviews with thirty GPs and seven managers of general practice businesses. For each doctor I interviewed, I also explored public information from websites, publications, government registration, biographies, Google searches, and conference proceedings. Any information that supported or contradicted a participant‘s account was noted, for further analysis. Data were collected over several months. Six study participants had second interviews and one GP also has a third interview.

Throughout this data collection process, I was concerned with keeping the interview conversation open and guided by the participant (as required by Glaser’s teachings), and wanting to capture depth and consistency in the data. I managed to ease the concern through developing of an interview guide that served as a prompt. I wanted to collect sufficient data to compare what GPs said and to explore new lines of enquiry. This guide was a useful tool that ensured I followed Glaser’s research method appropriately.

I would begin each interview with an open question so that issues were allowed to emerge more freely. Subsequent data collection was informed by these issues. Typically, the first question was as follows: Could you tell me about your work history since graduating in medicine? Other open questions I used were as follows: Tell me about your life as a GP. And what is it like being a GP? I used techniques to encourage the individual to continue talking such as repeating a question or reply. I also used phrases like: Go on. Tell me more about… or, why did you do that? It was important to remember to be open and flexible therefore the specific wording of questions and the order in which they were asked was influenced by the participant’s replies.

The interviews evolved in line with the concepts being developed. For example, the questions asked in the first interview with Monica (aged 44) followed the interview guide. However, my second interview with Monica, which occurred 12 months later, contained specific questions around freedom to make her own decisions and autonomy. My third communication with Monica was by email, shortly after the second interview, clarifying aspects of the doctor-patient relationship.

Even though Glaser (1998) wrote that there is no need to tape interviews in order to generate concepts and hypotheses, I was required to make digital recordings as evidence for my Doctor of Philosophy thesis; I was being asked to provide descriptive illustrations of my concepts. I also made field notes during and post interview, and compared these with the recorded interview. Initially, I analysed the transcribed data but as I gained experience, I coded field notes made from the recording rather than the transcription. On reflection, as a researcher who was inexperienced in conceptualisation, these recordings were an advantage. I was able to revisit, verify, and re-code text as new concepts emerged and I detected patterns.

Coding and Analysis

Inspired by the goal of Glaser’s methodology to find the core category, which explains how the participants resolve their main concern or problem, I conducted my analysis. This required coding incidents into categories, comparing incidents and categories, and trying to identify the main concern being articulated.

I began coding by fracturing the data, line by line, pulling the GP’s storytelling format apart, with each fragment becoming a very small episodic story. I analysed through asking Glaser’s questions: What are these data a study of? What is actually happening? What category does this incident indicate? What property of what category does this incident indicate? What is the participant’s main concern? I compared and grouped the fragments, eventually giving them an abstract name. My thoughts at the time can be found in the following descriptive memo:

Fracturing the data

I’ve highlighted the behaviour by looking at the action in the GP’s story in words such as exercise, work, bill, see, and take. Then I have looked for meaning in motivation, attitude and feelings by looking at words such as like, why I left, think, didn’t like, and didn’t feel comfortable. I compared these statements and conceptualised them as indicating control and autonomy. I have omitted dictated to because it is the participant’s observation. However, should an incident emerge which links this observation to the GP’s behaviour then I would code this text with that behaviour to help me understand the behaviour.

Coding 11 interviews yielded 367 codes (that is, 367 conceptual categories). This large number of codes grew quickly, as a consequence of over-coding and over-fragmenting the data, something which Glaser (2011) warns against (Holton, 2007). The process of comparing incidents became more difficult therefore recognising vital patterns in the data was at risk. With further analysis, I was able to redistribute the 367 codes into 10 categories and 36 sub-categories.

As the study progressed, I found that conceptualisation seemed to automatically take over from detailed description as codes were consolidated or set aside in accordance with GT. Categories were established that captured multiple incidents under the one concept. Examples of this are as follows:

• I found the problems and motivations for GP behaviour in the categories of triggers and coping mechanisms.
• Incidents I had coded in the categories of description of general practice, organisational context, and attitudes could be grouped together to form a new, higher-level category-named context with the two properties of external context and internal context.
• When autonomy emerged as the work value of most importance, all other sub-codes of work values were be set aside while I continued to code incidents to autonomy (thus delimiting the scope of the study).
• Incidents I had coded in the categories of career pathways, satisfaction, and participation appeared to indicate the outcomes of vocational behaviour.

I continued my analysis of the category I had named progression until I discovered the core category of optimising, after which codes and incidents were redistributed within this framework of optimising.

Discovering the Main Concern

For a long time I was unable to identify the main concern or problem that GPs were talking about and processing. There was tension around job changes and categories for triggers, work adjustment, and coping mechanisms. It was possible to see working lives being shaped by concerns about motivation, work-life balance, coping with demanding work and earning income. It appeared the problem was multi-faceted, yet it felt vaguely similar in each case.

I made a decision to examine the categories again, with fresh eyes. A quote taken from an interview with GP Monica (aged 44) helped focus my attention: “the goal from the beginning of my career was to be the best GP that I could be”. Soon after, I read a field note from the twelfth interview, with GP Meg (aged 35): “she says she couldn’t do just general practice without having other outlets and still be a functioning healthy person”.

Comparing these two fragments of data and asked the crucial questions: What is this a study of? What category do these incidents indicate? What property of what category do these incidents indicate? I realised both incidents could be seen as different perspectives of the same category. To pursue this idea, I took a theoretical sample of three doctors who had left general practice (Mary, Alan, and Natalie). My comparison of these cases highlighted, that while GPs wanted to achieve results in their work and progress in their careers, they also expressed concern about their own welfare. I named this new category sustainment—a name that captured both the aspirational character of the first incident referred to by Monica, combined with the protective and sustainable character of Meg’s incident. Sustainment integrated these elements, in response to changing needs and circumstances, in a way GPs could retain their ability to fulfil their primary career ambition.

I proceeded to compare the existing categories—triggers, work adjustment, tension and coping mechanisms, and their subcategories—with the new concept of sustainment. It seemed that the existing concepts could be linked to sustainment; because they related to decisions to enter the medical field, taking breaks, dissatisfaction, searching for experience or a more enhanced work role, financial factors, isolation, and personal or family life. At this point, it was important that I resist any temptation to engage in logical elaboration, so I examined the concept indicators again, asking: Is this a problem for the study participants and are they resolving it? and recoded if necessary. When I completed this process, the need that these GPs had for sustainment was retained as the main concern. However, it now had three subcategories: need for self-care to sustain well-being, staying interested in the work to sustain intellectual motivation, and need for financial reward to sustain lifestyle.

Discovering the Core Category

I needed to find the core category that showed how this problem of sustainment was being resolved. At first, I identified a pattern of behaviour where GPs encountered a problem—which they remedied by seeking a better work situation—by changing jobs or changing some aspect of their current job. This created various career pathways, with different amounts of participation in clinical general practice. After 11 interviews, I was not confident I had found the core category. Initially, I named the core category career shaping for own needs and continued data collection and analysis. I worked on several categories until it became clear that career shaping for own needs had the most explanatory power.

Autonomy and, to a lesser extent, recognition were present in this process of career shaping for own needs—however it was difficult for me to conceptualise the link between them. Autonomy seemed to have a stronger link to a GP’s search for a more satisfactory professional situation than did recognition. Autonomy had the internal perspective, of providing a means for career shaping, whereas recognition held the external perspective of how the public regard GPs. Using the prevailing internal perspective, this study focused on autonomy as a key element in being able to make career shaping changes. Following Glaser’s rules for delimiting, my work continued using those variables that related to this notion of career shaping.

Over time I came to the view that whilst the name career shaping for own needs did fit the core behaviour pattern, it lacked conceptual grab and was limited in how well it explained the observed behaviour. I considered other names: sculpting a professional life and me-shaping for autonomous needs. An individual GP may be juggling multiple issues, and yet they were generally thoughtful and structured in their behavioural responses.

My eureka moment came whilst reading an operations research piece that described an algebraic technique for solving linear problems. The concept of optimal solution was defined as “the point (in the solution space), which maximises the value of the objective function”. I understood that to optimise was to make the best or most effective use of a situation or resource—relevant when there was an imbalance between needs and the limited resources to resolve them. It became clear that the study participants were looking for the most satisfactory or optimal solution within a set of constraints: that they were optimising their professional life.

While career shaping could refer to any solution that resolved their problem in a satisfactory way, these GPs sought the optimal solution, that is, the special case that is most satisfactory. Also, they sought this optimal solution subject to their individual set of internal and external constraints.

I concluded that optimising professional life explained how GP make trade-offs between competing needs to optimise their personal situation and develop a satisfying career. Glaser supports connections of this kind; where the researcher is open and theoretically sensitive to how other fields conceptualise data: “By familiarity with ways of constructing variables in other fields he [the researcher] may imbue his theory in a multivariate fashion that touches many fields” (Glaser, 1978, p. 3).

Integrating Concepts in a Theory

Once the data were analysed, to aggregate incidents into patterns, I turned to Glaser’s theoretical codes for a way to bring these concepts together into a theory that explained what was happening in the data. I saw the connection between sustainment and optimising wherever an individual was leaving or changing work, along a trajectory of professional life. This resembled a turning point or critical juncture—important because “they indicate where the difference occurs” (Glaser, 1978, p. 76). In a descriptive memo I wrote:

She found that her health was not as good as it used to be and she was getting very tired. She resolved the problem by reducing her work hours, thereby exhibiting self-care behaviour. However, reducing her participation in clinical work also reduced her income, which could potentially cause concerns about financial reward. She said that reducing her clinical hours was more important compared with the loss of income.

I used the theoretical code of process to bring together the generated categories of exploring, selecting, implementing, adjusting and testing as the action of the optimising process—into the stages of discomfort, assessment and resolution.

Sorting

The next key step in Glaser’s GT was to sort my ideation memos into an order guided by the theoretical codes found in the data. Holton (2007) described the process: “as the researcher sees similarities, connections, and underlying uniformities, the theoretical decision about the precise location of a particular memo is based on the theoretical coding of the data grounding the idea” (p. 284). By following this GT step I sought “internal integration of connections among the categories” and to avoid developing a theory that was “linear, thin and less than fully integrated” (Glaser, 1978, p. 116).

During the research project, I produced a Word document of 122 pages, containing 212 memos grouped by category name. My memo sorting began with rereading and reconceptualising each memo, breaking down some that dealt with several ideas, and setting aside others that were descriptive rather than conceptual. Where a connection or relationship between concepts was seen, I created a linkage memo (a hypothesis) to discuss the connection.

Glaser (1998) recommended choosing a large table for manual sorting of printed memos. I summarised each linkage memo on a separate Post-it note sticking it to the glass of a large window, in the position where it related theoretically and substantively to other Post-it linkage memos. This process continued—sorting, comparing, resorting, and adding Post-it note memos about other categories and properties—for several weeks. I reflected on the multi-coloured Post-it notes stuck to the windows until the final theoretical framework became visible to me. When the properties of optimising were arranged, beside the sorted linkage memos, I could see how the categories fitted into a theory underpinned by seven propositions (listed below), which described how these categories interacted.

• The main concern being addressed by the professionals in this study is the need for sustainment.
• The aim of the optimising process is to achieve the most satisfactory solution (i.e. to satisfy the need for sustainment to the greatest extent possible) within a set of constraints.
• The process of optimising has three stages: stage of discomfort, stage of assessment (a cross roads), and stage of resolution.

The solution space holds a range of possible satisfactory and very satisfactory solutions that meet the GP’s need for sustainment. Solutions are found in four dimensions: treating patients (control over work content and how the work is done), structuring the work day (administrative structure of the day), integrating work and personal life (balance between personal and professional life), and adapting oneself (building resilience and skills, changing perceptions and attitudes).

• GPs are aware of, and respond to, the constraints present in the internal and external environments.
• Having personal autonomy enables GPs to make desired changes in order to optimise their situation.
• Optimising is a psychological process that recurs throughout a professional career.

Examining the Existing Literature

The literature review for this project was driven by the concepts found in the data. It occurred after the new theory of optimising professional life was developed (Glaser, 1998). This ordering allowed me to commence the work “with as few predetermined ideas as possible” (Glaser, 1978, p. 3) and to “remain open to what is actually happening” (Glaser, 1978, p. 3).

In career development theory, a common theme occurs in that humans strive to meet innate psychological and biological needs through engagement in work. Since the early 20th century, the study of careers has been dominated by psychological-based theories.

I analysed five of these theories for concepts that were relevant to my new grounded theory: Holland’s theory of vocational personalities and work environments (Holland, 1997), Dawis and Lofquist’s (1984) theory of work adjustment, Super’s (1980) self-concept theory of career development(Super, 1980), Gottfredson’s (1981) theory of circumspection and compromise and Lent, Brown, and Hackett’s (2002) social cognitive career theory. These theories are underpinned by the notion of a linear upward trajectory of work positions, usually within a particular profession or organisation. In contrast, participants in my study most often forged their careers in small, non-hierarchical, private business organisations with few opportunities for advancement once they became fully registered.

I examined the medical literature finding studies relating to career choice and entry to the field, satisfaction, stress, work-life balance, and early career—although studies of mid and late career for GPs were scarce. These studies provided factors that could trigger career events. Concepts were analysed using constant comparison and incorporated into the grounded theory when appropriate.

This research discovered sustainment—a concept not found in the career literature. Sustainment conceptualises the dynamic tension between self-care, staying interested in the work and income. These needs are identified in the literature (Blustein, 2006; Kilmartin, Newell, & Line, 2002). However, it is the trade-off between them that uniquely locates the new theory of optimising professional life in the literature. This balancing act was alluded to in a qualitative study of female GPs working in Australia, where the researchers identified “juggling the complexities of competing priorities in one’s professional life” (Kilmartin et al., 2002) as an important issue.

The theory of optimising professional life is a career development theory. This theory is based on the behaviour of GPs working in Australia, with relevant concepts integrated from existing literature through constant comparison. Within the career literature, the theory contributes a dynamic explanation of person-environment fit and pays attention to multiple environments that may simultaneously influence how a GP’s professional career is shaped.

Conclusion

For my investigation of how GPs experience their medical careers and continue to be GPs, I chose Glaser’s theory-building research method. As this was a relatively new topic, the exploration benefited from an approach not preconceived with a priori assumptions that could force misaligned assumptions upon the analysis (Glaser, 1978). My conclusion, from doing GT, is that if Glaser’s teachings are followed, the researcher will be rewarded with a parsimonious theory that fits, works, is relevant, and modifiable. I felt my experience of doing GT enabled me to discover Glaser and the meaning of his teachings.

References

Blustein, D.L. (2006). The psychology of working: a new perspective for career development, counselling, and public policy. Mahwah, NJ: Lawrence Erlbaum Publishers.

Dawis, R.V. & Lofquist, L.H. (1984). A psychological theory of work adjustment. Minneapolis, MN: University of Minnesota Press.

Glaser, B.G. (2011). RE: Personal communication to L. Piko at Grounded Theory Troubleshooting Seminar, New York.

Glaser, B.G. & Strauss, A.L. (1967). The discovery of grounded theory. Chicago, IL: Aldine de Gruyter.

Glaser, B.G. (1978). Theoretical sensitivity: Advances in the methodology of grounded theory. Mill Valley, CA: Sociology Press.

Glaser, B.G. (1998). Doing grounded theory: Issues and discussions. Mill Valley, CA: Sociology Press.

Gottfredson, L.S. (1981). Circumscription and compromise: A developmental theory of occupational aspirations. Journal of Counseling Psychology, 28, 545-579.

Holland, J.L. (1997). Making vocational choices: a theory of vocational personalities and work environments, 3rd edition. Odessa, FL: Psychological Assessment Resources.

Holton, J.A. (2007). The Coding Process and Its Challenges. In: Bryant, A. & Charmaz, K. (Eds.) The Sage Handbook of Grounded Theory. London: Sage Publications.

Kilmartin, M.R., Newell, C.J. & Line, M.A. (2002). The balancing act: key issues in the lives of women general practitioners in Australia. Medical Journal of Australia, 177, 87-89.

Lent, R.W., Brown, S.D. & Hackett, G. (2002). Social cognitive career theory. In Duane Brown and Associates (Eds.) Career Choice and Development 4th edition. San Francisco, CA: Jossey-Bass.

Super, D.E. (1980). A life-span, life-space approach to career development. Journal of Vocational Behavior, 16, 282-298.