Compare and Contrast Grounded Theory and Discourse Analysis
The debates and the doctrines regarding what is legitimate knowledge and how knowledge should be acquired have been extremely influential for the development of Psychology (Camic, Rhodes and Yardley, 2003). In the last half a century the positivist epistemology has prevailed. It asserts that there is a direct correspondence between the world and how it is perceived and therefore it is possible to gain objective knowledge about the world by observing it. This epistemology places great importance on controlled experiments and quantitative data and its enthusiastic adoption lead to a "shift in the emphasis from oral to written, from timely to timeless and from the particular to the universal in psychological research (Eisner, 2003). However by the 1960s and 1970s, social scientists were questioning the value of research which was purposely removed from the phenomena which it attempted to understand and which seemed to discount the depth of human experience (Eisner, 2003). This was compounded by a growing acknowledgement that the positivist-empirical approach was unable to produce purely objective truths. Over the last twenty years, qualitative methods for research have been successfully developed and reintroduced into mainstream Psychology (Stainton-Rogers and Willig, 2008). Within qualitative psychology there are many different approaches, based on different epistemological assumptions (Lyons, 2007). Two important approaches are Grounded Theory (GT) and Discourse Analysis (DA). This essay will compare GT and DA with reference to I)philosophy, II)goals, III)methodology, IV)the role of the researcher and V)criteria for evaluation.
Implicit in the question is that GT and DA are singular perspectives. In reality, they are characterised by a diversity of forms and debates. The original version of GT was described by Glaser and Strauss (1967). It has since been developed in different ways by Glaser (1978, 1998) and Strauss and Corbin (1990). In addition, there have been contributions from Charmaz (2006), in the form of a social constructivist version, Rennie (2000), in the form of a methodical hermeneutics version and Clarke (2005), in the form of a postmodern version (Payne, 2007). Similarly, within DA it is possible to distinguish between discursive psychology, critical discourse analysis and Foucauldian discourse analysis among others. A description of all the different forms of GT and DA is beyond the scope of this essay. This essay will focus on comparing the more interpretive version of GT as espoused by Strauss and Corbin (1990), which will be termed Straussian GT, with Foucauldian DA. In order to clarify the similarities and differences, where possible this essay will use examples from current qualitative research on eating disorders, one paper which uses GT to provide an explanatory schema about recovery (D'Abundo and Chally, 2004) and the other which uses Foucauldian DA to analyse how the "eating disordered patient is constructed (Malson, Finn, Treasure, Clarke and Anderson, 2004).Philosophy.
GT developed from research by Glaser and Strauss into the experience of dying patients in American Hospitals during the 1950s and 1960s. It was advertised as an alternative to the hypothetico-deductive method. Glaser and Strauss questioned the benefits of enforcing an artificial separation between theory and data collection and they proposed a method for generating theory inductively through close examination of data (Payne, 2007). GT is based on a positivist epistemology because it regards the data and its subsequent analysis as being based on an objective external reality. Straussian GT has since adopted a post-positivist epistemology (Payne, 2007) because although it continues to assume an objective external reality, it acknowledges the importance of interpretation and stresses that knowledge is positioned within a particular context (Payne, 2007).
DA is interested in how discourse constructs both people and the world around them. Foucauldian DA views discourse as reflecting and perpetuating socially and culturally specific power relations (Willig, 2003). It is based on a social constructionist epistemology which asserts that the world does have a structured reality but "discourses facilitate and limit, enable and constrain what can be said by whom, where and when (Parker, 1992 c.f. Willig, 2003). Thus, the way that a person perceives and experiences the world is dependent on the discourses that are available to them (Willig, 2003). Madill has suggested that current epistemology is a continuum between a naive realist position and a radical constructivist position (Lyons, 2007). Foucauldian DA and Straussian GT are on opposite sides of this continuum but both lie nearer to the middle than the extremes.Goals and Research Questions.
The principle aim of Straussian GT is to elucidate theories which stay close to the data, are compatible with the experiences and the understanding of the research participants and have the potential to guide action. Grounded theorists are particularly interested in making sense of processes and understanding how and why particular outcomes might occur (Lyons, 2007). D'Abundo and Chally (2004) sought to use GT to understand the process of recovery from an eating disorder from the perspective of someone with an eating disorder. The research concluded with a model of recovery which explained the cyclical nature of eating disorders with reference to a concept called the circle of acceptance and advised on ways of promoting recovery.
The principle aim of a Foucauldian Discourse analyst is to investigate what discourses exist and how this affects the experiences of people living in a particular culture (Willig, 2003). Foucauldian Discourse analysts argue that dominant discourses "privilege versions of reality that reinforce existing social structures and networks of power relations (Arribas-Ayllon and Walkerdine, 2008). It is hoped that an increased awareness of discourse and its relation to power structures will enable "de-stabilising of the status-quo by encouraging the emergence of alternative discourses (Arribas-Ayllon and Walkerdine, 2008). Malson et al (2004) sought to investigate how the "eating disordered patient is constructed by those individuals with an eating disorder as well as by health care professionals. In turn the Researchers were interested in how such constructions affected the experience of treatment. One of the conclusions of this research was that traditional power relations between the patient and the health care professionals were being perpetuated by a construction of the patient as purely "pathologised. When this position was adopted by the eating disordered individual, it naturally limited the potential for recovery.
Both approaches emphasise that their research can be used to guide action and improve the situation of the people who are the focus of the research. The difference is that the aims of Foucauldian DA tend to be more abstract and more political than those of Straussian GT.Role of the Researcher.
Straussian GT encourages Researcher's to try and discard any preconceived knowledge or assumptions, so that they are able to collect and analyse data with an open mind (Lyons, 2007). Most theorists acknowledge that this is not possible and therefore advocate revealing the perspectives of the researcher at the outset. D'abundo and Chally (2004) reveal that they have pre-existing knowledge about recovery from eating disorders through both personal experience and reviewing the literature. The reader is enabled to consider the influence that their preconceived knowledge and experiences might have had on their analysis. Such concerns are related to the debate in GT about the extent to which a literature review should be performed before data collection commences (Payne, 2007). Early versions of GT emphasised that the literature surrounding the research area should not be consulted until after the data has been collected and analysed, so that the Researcher is more perceptive to the concepts and theories emerging from the data (Payne, 2007). This debate does not exist within Foucauldian DA because it views the analysis that it provides as socially constructed and therefore there is no attempt to separate the researcher from the data. The fact that the knowledge and experience of the Researcher will undoubtedly influence the analysis is not seen as a threat to the validity of the research because an objective position is considered unachievable (Lyons, 2007).Methodology.
The main methodological features of Straussian GT are the initial coding and constant comparison technique, theoretical resampling and axial coding, data synthesis, theory development and testing of the emergent theory. Initial coding is commenced following some familiarisation with the text. With the intention of being as inclusive as possible at this early stage, meaningful parts of the text are determined and sorted into categories. Data continues to be coded until no more new categories are revealed. Parts of the text are then compared to the established categories in order to achieve further insights. If insights are gained, taking them into account may require a change in the coding. Definitions of each category are written and further participants are selected in order to test the emerging theory. The categories identified during the initial coding stage are examined for patterns and relationships until a core category is arrived at. This core category is developed by seeking links with existing theory in the literature. Finally the Researchers return to the original data set or collect some new data with which to compare and validate the theory (Payne, 2007).
In contrast to the highly specified Straussian GT methodology, Foucauldian DA has no formal set of procedures for analysing data (Arribas-Ayllon and Walkerdine, 2008). Billig has argued that the key to analysing discourse is "scholarship and the development of an analytic mentality rather than adherence to a methodology (Arribas-Ayllon and Walkerdine, 2008). Willig (2003) has provided guidelines for analysing discourse within the Foucauldian DA approach. The main features are an attempt to identify the different ways in which discourses concerning a particular subject are constructed and a questioning of what the functions of certain discourses might be. According to Willig (2003), it is important to think about how discourses might be constructed in order to position the participant in a particular way and how such positioning facilitates or limits the potential for action. Finally Willig (2003) advocates a careful consideration of how particular discourses, and the positioning associated with them influence the subjective experiences of the participant.
Foucauldian DA and Straussian GT engage in selective recruitment of participants who are familiar with and can provide insights into the particular research focus (Starks and Trinidad, 2007). Data tends to be elicited through interviews or focus groups. Straussian GT is concerned with gaining an authentic representation of the experience of the participant. The words of the participant are taken at face value (Lyons, 2007). Foucauldian DA, in contrast is concerned with how language is used to construct particular discourses. The constructive role of language is not common-sense and participants may not be aware of the consequences of their discourse. For this reason, a Foucauldian Discourse analysis may not be compatible with the subjective experiences of participants (Lyons, 2007).Evaluating the Research.
Criteria for evaluating qualitative research are yet to be firmly established. Most agree that the evaluative criteria should depend to some extent on the epistemological position and goals of the approach that has been adopted (Lyons, 2007). In both Foucauldian DA and Straussian GT, analysts strive for reflexivity and transparency by openly stating their perspectives. For Foucauldian discourse analysts, it is particularly important to consider the social implications of the accounts that they have provided. The emphasis is on the reader to evaluate the conclusions that are provided with knowledge of the Analyst's perspective in mind. The method for reporting Foucauldian DA emphasises that the analyst should try to present as much of the relevant text as possible and demonstrate how analytic conclusions were reached so that Readers can judge for themselves whether the interpretations were justified.
To conclude, this essay has argued that neither GT nor DA are unitary approaches and there are significant differences between versions of the same approach as well as between approaches. The main differences between GT and DA lie in conceptualisations of the relationship between language and reality. From these differing conceptualisations emerge different concerns regarding objectivity and contrasting social and political agendas. DA and GT have similar methodologies, though DA is less clearly specified. Both emphasise the need to justify the conclusions reached with evidence from the text and have similar concerns regarding reflexivity and transparency.
- Arribas-Ayllon, M and Walkerdine, V (2008) "Foucauldian Discourse Analysis In Willig, C. and W. Stainton Rogers. The SAGE handbook of qualitative research in psychology. Los Angeles, Calif. ; London, SAGE, pp 91-109
- Camic, P. M., J. E. Rhodes, et al. (2003). Qualitative research in psychology : expanding perspectives in methodology and design. Washington, D.C., American Psychological Association ; London : Eurospan.
- Charmaz, K. and Henwood, K.L. (2008) "Grounded theory, In Willig, C. and W. Stainton Rogers. The SAGE handbook of qualitative research in psychology. Los Angeles, Calif. ; London, SAGE, pp 240-260
- D'Abundo, M. and P. Chally (2004). "Struggling with recovery: participant perspectives on battling an eating disorder." Qual Health Res 14(8): 1094-1106.
- Eisner, E.W. (2003) "On the Art and Science of Qualitative Research in Psychology, In Camic, P. M., J. E. Rhodes, et al. Qualitative research in psychology : expanding perspectives in methodology and design. Washington, D.C., American Psychological Association ; London : Eurospan, pp 17-29
- Lyons, E. (2007) "Doing Qualitative Research, In Lyons, E. and A. Coyle. Analysing qualitative data in psychology. Los Angeles, Calif. ; London, SAGE, pp 3-9
- Lyons, E. (2007) "Analysing qualitative data: comparative reflections, In Lyons, E. and A. Coyle. Analysing qualitative data in psychology. Los Angeles, Calif. ; London, SAGE, pp 158-175
- Malson, H., D. M. Finn, et al. (2004). Constructing 'The Eating Disordered Patient': A discourse analysis of accounts of treatment experiences." Journal of Community & Applied Social Psychology 14(6): 473-489.
- Payne, S (2007) "Grounded Theory, In Lyons, E. and A. Coyle. Analysing qualitative data in psychology. Los Angeles, Calif. ; London, SAGE, pp 65-87
- Starks, H. and S. B. Trinidad (2007). "Choose your method: a comparison of phenomenology, discourse analysis, and grounded theory." Qual Health Res 17(10): 1372-1380.
- Willig, C. (2003) Discourse Analysis, In Smith, J. A. Qualitative psychology : a practical guide to research methods. Los Angeles ; London, SAGE, pp 159-184
- Willig, C. and W. Stainton Rogers (2008) "Introduction, In Willig, C. and W. Stainton Rogers The SAGE handbook of qualitative research in psychology. Los Angeles, Calif. ; London, SAGE, pp 1-13.