The Importance of Epistemology When Defending a Doctoral Thesis: The Research Philosophical Nature of Classic Grounded Theory

Daniel P. Ash, University of Gloucestershire


This article sets out a conceptual discussion based on a lively epistemological debate that took place during a doctoral viva voce examination for a classic grounded theory (GT) study (an exploration of police behaviour during domestic abuse incidents). The discussion uncovered conflicts regarding how methodology is viewed from different research traditions, and the potential impact of this problem on how research (like classic GT) is received and understood by scholars from different research philosophical perspectives, such as positivism. It also revealed how challenging it is, for a novice researcher, to adequately convince others of the rigorous nature of classic GT without reference and comparison to the research philosophical ideas that underpin other research traditions, more broadly. I claim that research philosophy must be addressed more fully by classic GT scholars if they are to encourage the uptake of classic GT by more doctoral candidates.

Keywords: Classic grounded theory, epistemology, research philosophy, police research, positivism, abductive research


This article explores a lively methodological debate that took place during a doctoral viva voce examination for a classic grounded theory (GT) study. The discussion uncovered problems with how classic GT is understood by scholars from different research philosophical traditions, and how challenging it is, as a novice researcher, to adequately convince others of the rigorous nature of classic GT without reference to research philosophical ideas. I argue that there are fundamental research-philosophical differences under consideration during the development and deployment of classic GTs, as compared with natural scientific theories. Recognising these differences is the key to understanding the contribution to knowledge being claimed within a classic GT study, and a reason why classic GT studies do not routinely include tests of reliability or validity using the natural scientific method. However, arguing these points to a doctoral committee or examiner can be problematic because there is a general lack of development or agreement in the academy relating to the research philosophical positioning of classic GT. This makes it challenging for other scholars to situate classic GT among their understanding of research philosophical approaches, which is problematic because situating a method within a philosophical position is often how the academy evaluates the appropriateness and effectiveness of the approach being taken within a doctoral study. In this paper, I claim that research philosophy must be addressed more fully by classic GT scholars if we are to encourage the uptake of classic GT by doctoral candidates.

The classic GT thesis that underpins the discussion in this paper was an exploration of police behaviour during domestic abuse incidents (Ash, 2021). The external examiner, a positivist and police forensic psychologist, challenged the reliability of the classic grounded theory method as an explanation for social behaviour because the thesis did not discuss nor demonstrate how the theory could be falsified using natural scientific techniques. They argued that without using natural scientific validation approaches, the thesis was not rigorous enough viz. the work did not adhere to natural scientific principles. At the core of the scientific method, and this discussion was the positivist concept of falsifiability, which according to Popper (1968) means that before any theory can be considered scientific (and therefore, according to the positivistic perspective, be reliable as a source of explanatory value) (Mingers, 2004), it must be testable using scientific techniques by way of experimentation.

Defending the use of classic grounded theory

Such natural scientific tests are not needed for a classic GT to be of value to practitioners (Glaser, 1978), which is the ultimate destination of any classic GT theory. We must also make a distinction between the way that natural scientific and classic GT theories are placed on the deductive/inductive spectrum. Natural scientific theorising usually adopts a deductive approach whereby theory-building commences with an abstract theoretical idea, which is then empirically tested and developed as a set of hypotheses whereby natural scientists continually aim for correctness and accuracy. In the viva under discussion, the candidate argued that classic GT is an abductive approach, whereby theory is developed based on incomplete knowledge about phenomena; “a creative inferential process aimed at producing new hypotheses and theories” (Tavory & Timmermans, 2014, p. 167). Furthermore, as grounded theorists, we accept that during all stages of GT development, we have an incomplete picture that can be improved with the collection and analysis of more data, but it will always be an approximation–never complete. This concept of “best-fit” explanations of phenomena can be problematic from a positivist perspective because they can be difficult to test experimentally.

There will always be outlying examples of social events that are not predicted by a classic GT because it is not intended to provide complete coverage of the area being examined or a thick description of some social phenomena (Glaser, 1978); a classic GT is not intended to provide conceptual completeness (Glaser, 2005). It follows, therefore, that most classic grounded theories might be falsified (by way of reason or logic) under some set of specific conditions. However, this does not negate their value as a theory of social practice. This is because social events, representing interactions between people, necessarily involve the exercise of human agency–choice (Crewe, 2013). Therefore, people can, and do, behave in ways that are not predicted by social theories, but most of the time they do not. Hence, patterns emerge from social interactions, which support abductive reasoning and can be identified and conceptualized using the classic GT approach.

The classic GT method includes elements that naturally verify the unfolding grounded theory as being an accurate representation of the social events being studied. These elements include the interchangeability of indices and the constant comparison of indicators, which both ensure that GT development is continual because data are iteratively compared with concepts and data that have already emerged during data collection and analysis (Glaser & Strauss, 1967). Continually checking conceptualizations against different parts of the data set and the emerging GT helps to attenuate the impact of any “outlying” datum on the emerging theory, ensuring that the GT being developed is fully integrated, and is conceptually capturing the essence of what is happening (Glaser, 1978).

A classic GT could be understood as being explanatory most of the time because it has been developed from a broad sample of “culturally typical” practice events, and is therefore sufficient for making improvements to practice in most practice situations that are associated with the original parameters of the GT study. The typicality of the examined social practice events can be further verified as “culturally typical” by triangulating data sources, relying on the researcher’s professional experience (if they are researching within their field), and through comparison with the literature after the discovery of the core category.

Therefore, we could argue that the explanatory or predictive value of a classic GT emerges from the method, while also being further developed through a culturally defined knowledge framework in a relativistic approach to theory development based on incomplete data that is good enough for explaining praxis (Bernstein, 1983) in support of practice change, viz. an abductive approach to theory development, without resorting to natural scientific tests of objectively defined accuracy, such as falsifiability.

It is not necessary to establish, via falsifiability, the surety or precision of a classic GT before practitioners can operationalise it as an effective way of improving their practice, and even if a classic GT had been falsified using the natural scientific approach, this would not negate the value of the theory. Classic GTs diverge from natural scientific theories in how they are developed and “tested” because they fundamentally have two different purposes—a classic GT is an interpretation of practice, intended to provide practitioner understanding, not a way of establishing objective practice or social “truths”. Importantly, a classic GT, as a form of social scientific inquiry, is not claimed to operate in all observable circumstances; it is a set of probability statements, which represent social, not natural objects of inquiry, which are only ever claimed to operate most of the time, ceteris paribus (Glaser, 1998).

Practitioners can still use a classic GT that has been falsified (or where falsification has not been attempted) to make improvements to their practice by dealing with most, if not all, problematic elements of their practice. Experienced practitioners can assess the relevance and fit of a classic GT to their practice activities without the need for the objective confirmation of accuracy provided by formal scientific methods (Glaser, 1978); they “try it out” (logically or practically), and it either makes/or is likely to make improvements, or not. Stated another way, a classic GT is “tested” through a form of judgmental rationality (Bhaskar, 1979), not natural-scientific objectivism, because within a social science paradigm (where social events occur in an open, not a closed system), “practical adequacy” is the test being applied to establish the value of such a social theory (Sayer, 1984) i.e. does the theory produce knowledge that adequately explains or predicts social events?–classic GT fits with this “test” of the value of social theory.

In the viva being examined in this paper, despite some of these arguments being made, there persisted disagreement about the “nature” of classic GT, which emerged from a lack of philosophical agreement or understanding between examiner and candidate. In this regard, it may be that some of the challenges of reconciling views between natural and social scientists on the nature of classic grounded theory might, in part, be explained by more fundamental problems of not understanding each other’s specific vocabulary (which necessarily develops within disciplines). For example, a different lexical interpretation of the word “theory” appeared to be a point of contention: a problem arose with the term ”grounded theory.” The positivist examiner took this to mean “scientific” theory that is grounded in social research, rather than as perhaps Glaser intended, a complete system of methods for producing a plausible and operationalizable explanation for some, but not all, elements of social practice. This misalignment of vocabulary caused confusion. More fundamentally, there seemed to be a general lack of shared vocabulary between examiner and candidate affecting understanding of each other’s positions, more broadly. I argue that this situation may have arisen unintentionally because of a lack of clarity around the research philosophical positioning of classic GT, which then made it challenging to present, as a rigorous way of doing social research.

Classic grounded theory and research philosophy

According to Glaser and Strauss (1967), GT is a “stand-alone” research method, rather than being a research method combined with a specific research philosophy–a chameleon that can be adapted to a wide range of philosophical approaches. Glaser (2001) argued that GT is a general method and therefore should not be claimed/privileged as belonging to any epistemological perspective or approach. He went further by providing a spirited defence of the a-philosophical nature of classic GT; writing books that specifically argued against the philosophical appropriation of classic GT, including the dangers it represented for the remodelling of the method (e.g. Glaser, 2003). Specifically, Glaser appeared concerned that the proponents of different philosophical approaches might remodel the GT method to be more congruent with their perspective, and precipitate a dilution of the GT “recipe” through the removal of the unique elements of GT that make it rigorous (e.g. theoretical coding) (Glaser, 2005).

This position has been the subject of much debate in the years following the first coining of the term “grounded theory.” The ostensible problem inherent with Glaser’s “a-philosophical” position was summarised by Nathaniel (2011):

[u]nfortunately, neither Glaser nor Strauss articulated a philosophical foundation for the method. So, through the years various authors have proposed piecemeal explanations of the method’s ontological, epistemological, and methodological underpinnings, thus promoting erosion and remodelling of the grounded theory method and creating a variety of notions about the method’s philosophical foundation [however], the first principles, assumptions and beliefs of a given philosophy contribute the ontology and epistemology to a methodology [emphasis in the original] and hold it together. (p. 187)

It seems that defending the integrity of the classic GT method may have unintentionally created a vacuum of philosophical reasoning, which was then naturally filled by other scholars as they sought to define where GT was positioned concerning their perspectives. As such, Glaser’s fears may have become self-fulfilling. Arguments about the philosophical basis of GT persist in a polarised and often polemic fashion (Kendall, 1999), which has led to different versions of GT being developed independently rather than collegiately–the remodelling that Glaser was probably hoping to avoid and a schism in the GT community. This problem has led to classic GT, as a method, being difficult to learn for a novice researcher because different texts on GT confuse and conflate classic GT concepts in a way that makes it necessary to understand the historical development of GT before one can make an informed choice about which approach to take (O’Connor, Carpenter and Coughlan, 2018).

More recently, authors have recognized the need to discuss the relationship between GT and research philosophy as part of a requirement of most research studies (Bryant & Charmaz, 2007). By engaging with research philosophical debates regarding the nature of GT, it might be possible for classic GT researchers to more confidently position and defend their work and the GT method when trying to extend their research post-doctorate, such as when seeking to generate impact by implementing research findings within a practice context. This is especially important whenever their field of practice is closely associated with deeply entrenched political ideologies that are associated with particular ontological or epistemological schools of thought (DiCristina, 1994), for example, Evidence-Based Policing’s close associations with positivism and its use as a “technology of power” (Lumsden & Goode, 2018).

Some GT scholars have described GT as having flexible epistemological assumptions (Holton and Walsh, 2016), which is an entirely defensible position, but perhaps lacks the precision that is expected or needed by novice GT scholars when they are trying to understand the nature of classic GT. While the concept of epistemological flexibility when discussing GT’s foundations is an advance on Glaser’s position, it remains challenging for classic GT novices to use GT in a PhD study because they are still required to justify its basic methodological assumptions to a PhD committee (Lowe, 2017), a problem which Glaser also recognized to some extent (Glaser, 2015). Unfortunately, arguing for epistemological flexibility as a basis for GT does not necessarily provide the surety of philosophical foundations that are demanded within many research institutes. For example, with flexibility might come confusion for the novice researcher, PhD candidate and even PhD committee members who seek to clearly understand the ontological and epistemological position of examined work (Ahmed & Haag, 2016; O’Connor et al., 2018).

It was certainly the case in the viva under discussion in this paper, that there was a “dialogue of the deaf” (Johnston & Shearing, 2009, p. 415) taking place, where both candidate and examiner lacked the shared research philosophical vocabulary to argue or understand the differences between their positions. Perhaps it is time for a more developed debate to take place about the ontological and epistemological position of classic GT that draws together practitioners from all contemporary forms of “GT” in a collegiate exchange that advances knowledge of GT as a method of research inquiry.


Ahmed, S. and Haag, M. (2016). Entering the field: Decisions of an early career researcher adopting classic grounded theory. Grounded Theory Review, 15(2).

Ash, D. P. (2021). Balancing value and effort: A classic grounded theory of frontline police practice. (Doctoral dissertation, Keele University).

Bhaskar, R. (1979). The possibility of naturalism. Humanities Press.

Bernstein, R. J. (1983). Beyond objectivism: Science, hermeneutics and praxis. University of Pennsylvania Press.

Bryant, A. & Charmaz, K. (2007). The Sage handbook of grounded theory. Sage.

Crewe, D. (2013). Becoming criminal: The socio-cultural origins of law, transgression, and deviance. Springer.

DiCristina, B. (1995). Method in criminology: A philosophical primer. Harrow and Heston.

Glaser, B. G. (2015). Choosing grounded theory: A GT reader of expert advice. Sociology Press.

Glaser, B. (2005). The grounded theory perspective III: Theoretical coding. Sociology Press.

Glaser, B. (2003). The grounded theory perspective II: Descriptionʹs remodeling of grounded theory methodology. Sociology Press.

Glaser, B. (2001). The grounded theory perspective: Conceptualization contrasted with description. Sociology Press.

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

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

Glaser, B. G. & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. Sociology Press.

Holton, J. A. & Walsh, I. (2016). Classic grounded theory: Applications with qualitative and quantitative data. SAGE.

Johnston, L., & Shearing, C. (2009). From a ‘dialogue of the deaf’ to a ‘dialogue of listening’: Towards a new methodology of policing research and practice. Police Practice and Research: An International Journal, 10(5-6), 415-422.

Kendall, J. (1999). Axial Coding and the Grounded Theory Controversy. Western Journal of Nursing Research, 21(6), pp. 743–757.

Lowe, A. (2017). Intellectual autonomy of PhD researchers who use the grounded theory methodology. Grounded Theory Review: An International Journal, 16(1).

Lumsden, K., & Goode, J. (2018). Policing research and the rise of the ‘evidence-base’: Police officer and staff understandings of research, its implementation and ‘what works’. Sociology, 52(4), 813-829.

Mingers, J. (2004). Re-establishing the real: Critical realism and information systems. In Willcocks, L. P., & Mingers, J. (Eds.), Social theory and philosophy for information systems. John Wiley & Sons.

Nathaniel, A.K. (2011). An integrated philosophical framework that fits grounded theory. In Martin, V.B. and Gynnild, A. (Eds.), Grounded theory: The philosophy, method, and work of Barney Glaser. Brownwater Press, pp. 187-200.

O’Connor, A., Carpenter, B. & Coughlan, B. (2018). An exploration of key issues in the debate between classic and constructivist grounded theory. Grounded Theory Review, 17(1).

Popper, K. R. (1968). The logic of scientific discovery. Routledge.

Sayer, A. (1984). Defining the urban. GeoJournal, 9(3), 279-284.

Tavory, I., & Timmermans, S. (2014). Abductive analysis: Theorizing qualitative research. University of Chicago Press.

Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.

© Daniel P. Ash 2022.