This paper discusses grounded theory as one of the qualitative research designs. It describes how grounded theory generates from data. Three phases of grounded theory - open coding, axial coding, and selective coding - are discussed, along with some of the issues which are the source of debate among grounded theorists, especially between its founders, Glaser and Strauss.
Grounded theory was introduced by Glaser and Strauss in their 1967 book, The Discovery of Grounded Theory. The book was based on a justification for using qualitative research to build up theoretical analysis. As Goulding (1999) mentions, it was written in part as an objection against what the authors viewed as a rather passive acceptance that all the 'great' theories had been discovered and that the role of research lay in testing these theories through quantitative 'scientific' procedures.
What Glaser and Strauss suggested as grounded theory is actually a "systematic, qualitative process used to generate a theory that explains, at a broad conceptual level, a process, an action, or interaction about a substantive topic" (Creswell, 2002, p.439). It is a qualitative methodology which obtains its name from the practice of generating theory from research which is 'grounded' in data (Babchuk, 1997). It can be defined as a method for analysing data which is most commonly used on naturalistic field data but has also been used to analyse historical and documentary data (Star, 1998). The grounded theory approach uses a "systematic set of procedures to develop an inductively derived grounded theory about a phenomenon" (Strauss & Corbin, 1990, p.24). This methodology is a general method of comparative analysis to realise theory with four central criteria - work, relevance, fit, and modifiability (Creswell, 2002). It will answer the question of 'What was going on in an area?' by generating either a substantive or formal theory - theory related to a case and developed inductively from empirical data to reach an abstract level (Star, 1998). The strongest cases for the use of grounded theory are in studies of comparatively unexplored areas (Samik-Ibrahim, 2000). Grounded theory is used to generate a theory rather than use one 'off the shelf' to enlighten a procedure, action, or interaction, a step-by-step, systematic process to stay close to the data (Creswell, 2002).
Although Glaser's and Strauss's collaborative work led to the introduction of grounded theory, their later works show epistemological differences between them (Glaser, 1978, 1992; Strauss, 1987; Strauss & Corbin, 1990). Glaser has a rigorous positive perspective towards qualitative analysis, while Strauss has a pragmatic epistemology into empirical inquiry through grounded theory. Whereas Glaser's standpoint tends to be more traditional positivism with emphasis on supposition of an objective and external reality as well as being a neutral observer, Strauss's work is based on the assumption of having an unbiased position in collecting data and applying a certain technical procedures by letting the participants have their own voice (Glaser, 1992; Strauss & Corbin, 1990). This viewpoint, as Charmaz explains, "moves into postpositivism" because it represents participants "as accurately as possible, discovering and acknowledging how respondents' view of reality conflict with their own" (2000, p.510). No matter what their philosophical perspectives are, they have an almost similar standpoint with respect to the main processes, including categorising and constant comparison to produce the theory grounded in data.
The rationale of grounded theory studies is to investigate and recognise how complicated phenomena occur. Strauss and Corbin (1998) suggest the following goals for grounded theory.
Strauss and Corbin (1998) suggest the following assumptions on which grounded theory methodology is predicated.
In order to start applying grounded theory methodology as a research design, grounded theorists need to examine whether or not it suits the area being studied and how it can help the research problem to be clarified. Therefore, identifying an overlooked area is the first thing to do.
Allan (2003) mentions that some people think of the grounded theory method as meaning fieldwork before a literature search. This is a misunderstanding of the original principle put forward by Glaser and Strauss (1967) who persuaded researchers to "use any material bearing in the area" (p.169). This remark can be found in the writings of other authors. Strauss and Corbin (1998) saw the literature as a foundation of professional knowledge and referred to it as literature sensitivity. A review of the pertinent literature reveals current thinking in the area. It should be mentioned though that this literature review should not bring about any hypotheses.
Applying grounded theory to the areas where an extensive, reliable and empirically based literature exists may cause some difficulties. Literature which already exists might prejudice or affect the perceptions of the researcher (Goulding, 1999). Also, there is a risk of entering the field with prior attitudes, whether aware of it or not, of testing such existing work rather than developing original insights about the area of study. To avoid this, it is suggested that the researcher go into the field at a very early stage and collect data. Unlike the case study which gains benefit from the existing development of theoretical propositions to guide data collection and analysis, grounded theory should have no pre-conceived ideas or hypothesis (Glaser & Strauss, 1967). In other words, there are conflicts in terms of collecting data and other procedures of grounded theory. Throughout the course of the research it is common to gather an extensive amount of data in various forms such as interview transcripts, field notes on observations, memos, diagrams and conceptual maps.
Grounded theory coding is a kind of content analysis to find and conceptualise the core issues from within the huge pile of the data. Throughout the analysis of an interview, for example, the researcher will become conscious that the interviewee is using words and phrases that highlight an issue of importance or interest to the research. This is noted and described in a short phrase. This issue may be mentioned again in the same or similar words and is again noted. Allan (2003) describes this process as 'coding' and the short descriptor phrase is a code. According to him, coding should be performed with an open mind, without predetermined ideas. Predetermined ideas should not be forced on the data by looking for confirmation of previously established ideas.
In the course of coding, more than one code may come out from the same text. The data should be reviewed many times, looking and re-looking for emerging codes. Strauss and Corbin (1998) suggested coding by "microanalysis which consists of analysing data word-by-word" and "coding the meaning found in words or groups of words" (pp.65-68). However, the analysis technique of coding by microanalysis of the data, word by word and line by line, has two disadvantages. Firstly, it is very time consuming. If the data comes from interviews, the transcription of each interview contains a mass of data that has to be studied to locate the information relevant to the research topic. Secondly, it may lead to confusion at times. Dividing the data into individual words sometimes causes the analysis to become lost within the details of data. Therefore, it is useful to identify key points (rather than individual words) and let concepts emerge. The selection of points, in order to address research questions, is in line with qualitative coding analysis and is a protection against data overload (Allan, 2003).
Personal thoughts may affect the process of coding and consequently the categories formed. Strauss and Corbin (1989) also believe that the interpretation of events by the researcher influences the naming of categories whereas Glaser (2002a) suggests that personal distance for accuracy is supposed to be an 'attitude' of the qualitative data analysis researcher. The grounded theory researcher, in contrast, does not need this attitude to get a description accurate, which is not his or her goal. The grounded theory method automatically puts him or her on a conceptual level, which goes beyond the descriptive data.
While naming concepts, grounded theory does not attempt to understand the world of the research participants as they construct it (Glaser, 1998). Grounded theory is not an enquiry that makes sense of and is true to the understanding of ordinary actors in the everyday world. According to Glaser (2002a) grounded theory discovers patterns that the participants do not understand or are not aware of. Grounded theory creates conceptual hypotheses that apply to any relevant time, place, and people with emergent fit and then is modified by constant comparison with new data as it explains what behaviour obtains in a substantive area. When concepts emerge they must be categorised in order to make relationship among them form theory. This process begins with open coding.
At the stage of analysing the data and looking for codes, the coding is 'unfocused' and 'open'. During this process the data are analysed and the grounded theory researcher may recognise hundreds of codes which might have potential meaning and relevance (Goulding, 1999). In the course of open coding the grounded theorist engages in breaking down, analysing, comparing, labelling and categorising data. In open coding, incidents or events are labelled and assembled together through constant comparison to form categories and properties (Babchuk, 1997).
Coding might start with a full transcription of an interview, after which the text is analysed in an effort to recognise key words or phrases which connect the participant's description to the experience under study. This procedure, as Spiggle (1994) describes it, is associated with primary concept development which consists of "identifying a chunk or unit of data (a passage of text of any length) as belonging to, representing, or being an example of some more general phenomenon" (p.493). Besides open coding, it is vital to incorporate the use of memos. Memos are notes the researcher writes during the research process or immediately after data collection to elaborate on ideas about the data and the coded categories (Creswell, 2002) as a way of recording the impressions of the researcher and describing the situation. These are fundamental since they provide a bank of ideas which can be reviewed in order to draw the emerging theory. Memos facilitate reorienting the researcher at a later date (Goulding, 1999).
Coding allows for direction before becoming selective. It breaks down the data into analytical portions which can afterward be raised to a conceptual point. Questions that need to be continually addressed include the following.
In order to create new categories for possible inclusion in developing theory, Gerson (1991) suggests 'supplementation' as a complementary way. It can be situated between coding and theoretical sampling. Supplementation starts with an extant category, and systematically elaborates contrasting categories in order to provide the raw material for theoretical sampling, cross-cutting and making the theories richer. Supplementation is equivalent neither to testing hypotheses on the one hand, nor to constructing new categories via coding on the other. Indeed, it can be done without reference to particular data at all, focusing instead on the conceptual organisation and relationships of the developing theory. Keeping the theory in a state of permanent confrontation with data is the work of theoretical sampling, not of supplementation. While theoretical sampling tells us what to worry about, supplementation tells us the terms in which we should worry about it (Gerson, 1991).
In grounded theory research, the inquirer engages in a process of gathering data, sorting it into categories, collecting additional information, and comparing the new information with merging categories. This process of slowly developing categories is called the 'constant comparative procedure' (Creswell, 2002). By comparing each concept in turn with all other concepts, further commonalities are found which form the even broader categories. Glaser and Strauss (1967) described this method of continually comparing concepts with each other as their constant comparative method (pp.115).
Constant comparative method is a fundamental feature of grounded theory. As the name implies, this involves comparing like with like to look for emerging patterns and themes. As Spiggle (1994) describes it, "[c]omparison explores differences and similarities across incidents within the data currently collected and provides guidelines for collecting additional data. Analysis explicitly compares each incident in the data with other incidents appearing to belong to the same category, exploring their similarities and differences" (p.493). To a degree, constant comparative process reveals the researcher's personal predilection, which may bias the data (Glaser, 2002b). Goulding (1999) also believes that as a result of constant comparison of subsequent data, codes are reduced and grouped into meaningful categories.
During this level of coding, theoretical saturation should be reached. This means that no new properties, dimensions, or relationships will emerge during analysis. Saturation is "the state in which the researcher makes the subjective determination that new data will not provide any new information or insights for the developing categories" (Creswell, 2002, p.450). Theoretical saturation is realised when
What has been discussed so far is the first stage of grounded theory. Concepts are a progression from simply describing what is occurring in the data, which is an attribute of open coding, to explaining the relationship between and across incidents (Glaser & Strauss, 1967). This needs a different, more complicated, coding technique which is referred to as 'axial coding' and involves the process of abstraction onto a theoretical level.
Once a concept has been identified, its attributes may be explored in depth, and its characteristics dimensionalised in terms of their strength or weakness. Finally the data are subsumed into a core category which the researcher has to justify as the basis for the emergent theory. The core category is "the central phenomenon around which all the other categories are related" (Strauss & Corbin, 1990, p.116). A core category unites all the strands in order to provide an explanation of the behaviour under study (Goulding, 1999). It represents the description of hypothetical relationships between categories and subcategories (Babchuk, 1997). Axial coding is the appreciation of concepts in terms of their dynamic interrelationships. These should form the foundation for the creation of the theory. The focus of axial coding is to construct a model that details the specific conditions that give rise to a phenomenon's occurrence. Strauss and Corbin (1998) believe that the purpose of axial coding is to reassemble data that were fractured during open coding.
In axial coding, four analytical processes are occurring: (a) continually relating subcategories to a category, (b) comparing categories with the collected data, (c) expanding the density of the categories by detailing their properties and dimensions, and (d) exploring variations in the phenomena (Brown, Stevenson, Troiano & Schneider, 2002).
With regard to the process of developing grounded theory, it may be argued that there are three basic stages that need to be addressed. The first deals with the collection and interpretation of the data and is mainly concerned with demonstrating how, why and from where early concepts and categories were derived. In accordance with the principles common to the method, any theory should be traceable back to the data (Goulding, 1999). Consequently, evidence needs to be provided as does the relationship between concepts, categories and this evidence.
The second stage is to 'abstract' the concepts and try to find theoretical meaning. At this stage the concepts should be adequately developed as to warrant an extensive re-evaluation of compatible literature in order to demonstrate the 'fit', relationship and, where applicable, the extension of that literature through the research findings.
The final stage should present the theory, bringing together the concepts and integrating them into categories which have explanatory power within the context of the research.
It is not necessary to ask the participants to see if the theory covers their situation, as is usual in narrative research designs. According to Glaser (2002a), inviting participants to review the theory for whether or not it is their voice is wrong as a 'check' or 'test' on validity. They may or may not understand the theory, or even like the theory if they do understand it. Grounded theory is generated from much data, of which many participants may be empirically unaware. Indeed, grounded theory is not their voice; it is a generated abstraction from their doings and their meanings that are taken as data for the conceptual generation.
Babchuk, W. A. (1997). Glaser or Strauss: Grounded theory and adult education. Paper presented at the Midwest Research-to-Practice Conference in Adult, Continuing and Community Education, Michigan State University, East Lansing, Michigan.
Brown, S. C., Stevenson, R. A., Troiano, P. F. & Schneider, M. K. (2002). Exploring complex phenomena: Grounded theory in students affairs research. Journal of College Student Development, 43(2), 173-183.
Charmaz, K. (2000). Grounded theory: Objectivist and constructivist methods. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp.509-535). Thousand Oaks, Ca.: Sage.
Creswell, J. W. (2002). Educational research: Planning, conducting, and evaluating quantitative and qualitative research. New Jersey: Merrill Prentice Hall.
Gerson, E. (1991). Supplementing grounded theory. In D. Maines (Ed.), Social organization and social process (pp.285-301). New York: AIDINE de Gruyter.
Glaser, B. G. (1978). Theoretical sensitivity: Advances in the methodology of grounded theory. Mill Valley, Calif.: Sociology Press.
Glaser, B. G. (1992). Basics of grounded theory analysis: Emergence vs forcing. Mill Valley, CA: Sociology Press.
Glaser, B. G. (1998). Doing grounded theory: Issues and discussions. Mill Valley, CA: Sociology Press.
Glaser, B. G. (2002a). Conceptualization: On theory and theorizing using grounded theory. International Journal of Qualitative Methods, 1(2). [viewed 1 )ct 2004] http://www.ualberta.ca/~iiqm/backissues/1_2Final/html/glaser.html
Glaser, B. G. (2002b). Constructivist Grounded Theory? Forum: Qualitative Social Research, 3(3). Retrieved October 01, 2004 from http://www.qualitative-research.net/fqs-texte/3-02/3-02glaser-e.htm
Glaser, B. G. & Strauss, A. L. (1967). The discovery of grounded theory. Chicago: Aldine Publishing Company.
Goede, R. & Villiers, C. D. (2003). The applicability of grounded theory as research methodology in studies on the use of methodologies in IS practices. Proceedings of SAICSIT 2003, 208-217.
Goulding, C. (1999). Grounded theory: Some reflections on paradigm, procedures and misconceptions. Working paper series, WP006/99, Wolverhampton: University of Wolverhampton. [verified 24 May 2006] http://www.wlv.ac.uk/PDF/uwbs_WP006-99%20Goulding.pdf
Morse, J. M. (1994). Emerging from the data: The cognitive process of analysis in qualitative enquiry. In J. M. Morse (Ed.), Critical Issues in Qualitative Research Methods. Thousand Oaks: Sage.
Piantanida, M., Tananis, C. & Crubs, R. (2002). Claiming grounded theory for practice-based dissertation research: A think piece. Paper presented at the Conference on Interdisciplinary Qualitative Studies, Roundtable Discussion, Athens, Georgia. http://www.coe.uga.edu/quig/pdf/claim.pdf
Samik-Ibrahim, R. M. (2000). Grounded theory methodology as the research strategy for a developing country. Forum: Qualitative Social Research, 1(1). [viewed 1 Oct 2004] http://www.qualitative-research.net/fqs-texte/1-00/1-00samik-e.pdf
Spiggle, S. (1994). Analysis and interpretation of qualitative data in consumer research. Journal of Consumer Research, 21(3), 491-503.
Stamp, G. H. (1999). A qualitatively constructed interpersonal communication model: A grounded theory analysis. Human Communication Research, 25(4), 531-547.
Star, S. L. (1998). Grounded classification: Grounded theory and faceted classification. Library Trends, 47(2), 218-232.
Strauss, A. L. (1987). Qualitative analysis for social scientists. Cambridge [Cambridgeshire] ; New York: Cambridge University Press.
Strauss, A. & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage.
Strauss, A. & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd ed.). Thousand Oaks, CA: Sage.
|Authors: Alireza Moghaddam is a doctoral candidate in Educational Studies and a graduate research assistant at the University of Western Ontario (UWO). He has a teaching experience in elementary schools, high schools, universities, and teachers colleges. His research interests include online education, digital divide, collaborative teaching, and manpower productivity in education. Email: firstname.lastname@example.org
Please cite as: Moghaddam, A. (2006). Coding issues in grounded theory. Issues In Educational Research, 16(1), 52-66. http://www.iier.org.au/iier16/moghaddam.html