GT v Constructivist grounded theory

As noted in the previous section, Grounded theory methods emerged from the fruitful collaboration of sociologists Glaser and Strauss (1965, 1967, 1968; Strauss and Glaser, 1970) during the 1960s. Several variations of grounded theory have evolved over time depending on the researcher’s ontological and epistemological beliefs (Glaser and Strauss, 1967; Charmaz, 1996, 2017; Nagel et al., 2015; Mills, Bonner and Francis, 2017; McCann and Polacsek, 2018; Thornberg and Dunne, 2019).

The methodology of GT was first described by Glaser and Strauss (1967) as a positivist research paradigm, compatible with objectivist underpinnings of the study of human behaviour. It has continued to develop through the work of researchers such as Strauss and Corbin (1998), and Charmaz (2000; 2006), and is now widely considered to be a valuable method for researching the socially constructed world. In Glaser and Strauss’ (1967) original approach, it was assumed that an objective observer discovers data that is subsequently reduced to manageable research problems, and analysed objectively, thus the researcher is detached from the data. However, Charmaz (2006 cited in Powell, 2016, p.67) advocated that GT assumes that people create and maintain a meaningful world in which they construct meaning and act accordingly, which concurs with the critical constructivism stance in the present study.

Strauss and Corbin’s (1998) later version of GT expresses a view in relation to a constructivist approach to inquiry (Mills, Bonner and Francis, 2017). Their work demonstrates a mixture between post-positivism and constructivism language, with terms such as recognizing bias and maintaining objectivity when describing the position, the researcher should assume in relation to the participants and the data. Nevertheless, they mix these ideas with observations such as “we emphasize that it is not possible to be completely free of bias” (Strauss & Corbin, 1998, p. 97). They also acknowledge the importance of a multiplicity of perspectives and “truths” (Strauss, 1987; Strauss & Corbin, 1990, 1994, 1998) and as such have “extended and emphasized the range of theoretically sensitizing concepts that must be attended to in the analysis of human action/interaction” (MacDonald,

2001, p. 137). This enables an analysis of data and a reconstruction of theory that is richer and more reflective of the context in which participants are situated. They insist that theirs is “interpretive work and . . . interpretations must include the perspectives and voice of the people who we study” (Strauss & Corbin, 1994, p. 274; emphasis in original). Such a position aligns with this study’s approach and clearly implies that this perspective includes relating participants’ stories to the world in which the participants live.

According to the literature GT uses are diametrically contested between traditional and constructivist grounded theorists (Mills, Bonner and Francis, 2017). Strauss and Corbin (1998) argued that it is advantageous to establish a ground of knowledge regarding the subject in hand, because previously acquired experience and knowledge are key for guiding the researcher in detecting challenges and concerns in the data, and permit a researcher to seek different explanations, and to identify the dimensions and properties of emerging aspects (Charmaz, 2017). Similarly, the present study sought to achieve this via an extensive review of the extant literature, which was studied, analysed, and synthesised prior to the data collection stage. However, the founders of grounded theory, Glaser and Strauss (1967) proposed that data collection and analysis should occur before conducting a literature review, as researchers may see their data through the lens of earlier ideas and become biased (Mills, Bonner and Francis, 2017). Glaser argued that the literature related to the researched area should only be read in later stages of a study (Glaser, 1992, p.31). He claimed that if a literature review were conducted before data collection and analysis, existing theories could impose themselves on the analysis and the resulting theory, and thus, prevent it from being truly grounded in, and emerged from, the data (Ramalho et al., 2015).

This perspective has been heavily disputed (see Charmaz, 1996, 2000, 2006, 2017). Strauss and Corbin’s (1998) perspective proposes a middle way, reasoning that when developing a grounded theory, it is important to consider how a review of literature can be used as an analytical tool to enhance conceptualisation, rather than constrain theory development. Therefore, they stated that literature read before data collection could not necessarily hinder the emergence of the theory,

Corbin, 1990, p.56). That is, as long as the researcher "maintains an attitude of scepticism" (Strauss and Corbin, 1990, p.45) and does not allow the literature read to impose itself on the theory (Ramalho et al., 2015).

For traditional grounded theorists, coding from the data is the fundamental analytic tool that will uncover an emergent grounded theory from the field of inquiry. Three forms of codes are used: open, theoretical, and constant comparative (Glaser, 1992). Open coding is the initial step of theoretical analysis, developing codes from the data. This form of coding ends when it locates a core category. Theoretical codes are “conceptual connectors” that develop relationships between categories and their properties (Glaser, 1992, p. 38). Constant comparative coding describes the method of constant comparison that imbues both open and theoretical coding. In contrast, Strauss and Corbin (1998) have used complex coding methods as strategies to examine the interface between structure and process. Strauss (1987) and then Strauss and Corbin (1990) followed on from this but focused on one particular coding family, the “Six Cs,” which identifies the causes, consequences, and conditions affecting categories identified by the researcher. Later, Strauss and Corbin (1998) simplified their paradigmatic framework to ask questions about the conditions, actions/interactions, and consequences of categories, thus making links between the ideas being conceptualized from the data. This form of coding is called axial coding and is used in this research (section 3.6.5.2)

The above comparison discusses several key points of difference between the research method of a constructivist grounded theory (CGT) study and that of a traditional grounded theory approach. These are: theoretical sensitivity, treatment of the literature, and coding (Mills, Bonner and Francis, 2017). Charmaz has contended that a constructivist approach to grounded theory is both possible and desirable, because, “Data do not provide a window on reality. Rather, the “discovered” reality arises from the interactive process and its temporal, cultural, and structural contexts” (Charmaz, 2000, p. 524).