Navigation: This page 'gt.html' ---> External Thinking ---> Main Page. HELP. Contact.

Grounded Theory - compared with Dooyeweerd

Glaser and Strauss' [1967] 'Grounded Theory' is much used in socially rich areas, from nursing to information systems research and in I.S. design, in order to elicit knowledge of a situation in such a way that the concepts that arise are 'grounded' in the situation rather than being products of the researcher's prior categories and models.

There are three main versions of GT, the original one proposed by Glaser and Strauss, an operationalized version propsoed by Strauss and Corbin [1998], which is seen by many almost as heretical, and a subsequent reaction to by Glaser. Here we look at the first two, seeing them in the light of Dooyeweerd. It is instructive to compare GT with Dooyeweerd's recognition of the multi-aspectual nature of human social situations, and his approach to science and knowing.

Contents:


Glaser and Strauss's Original Version of GT

Glaser and Strauss' [1967] approach to grounded theory emphasises the long-overlooked (in 1967) importance of generating theory from data as opposed to verifying existing theory, and it has become widely used in sociologically-rich ISR. They argued that such theory should be generated by grounding it in data rather than in pre-given categories or a priori theoretical constructs. In this can be seen an echo of Dooyeweerd's belief in the importance of the na´ve everyday attitude of thought, rather than approaching the domain of interest with prior theoretical constructs.

What made GT notable, however, was the method defined for it. Originally Glaser & Strauss described four stages: (1) comparing incidents applicable to each category, (2) integrating categories and their properties, (3) delimiting the theory, (4) writing the theory.

When GT is well done it can disclose a good variety of issues. The account in Glaser & Strauss [1967] exhibits good awareness of the everyday life of both researched and researcher. Of those researched: Their research into nurses caring for dying patients shows keen sensitivity to the nurses' tasks, feelings, relationships with patients, care for bereaved families, professional attitudes, creation of "loss rationales" and of "loss stories" in which there was a "continual balancing out of social loss factors" [p.109]. Of researcher: Their account also shows awareness of the diverse everyday realities of sociological research, including: coding being seen as "unnecessary and burdensome", writing memos, that research is "still dependent on the skills and sensitivities of the analyst", the creativity of theory generation, the need for discipline in face of "vagueness and flexibility", the mood of the analyst, importance of teammates, tendency to become committed to one's theory, and so on [pp.102-111].

So the methods of GT, and the experience of using it accumulated since the 1970s, could perhaps be capitalised upon whenever everyday human activity needs to be researched.

General Critique of Original Version

But is GT sufficient on its own to take an everyday attitude when researching? There are reasons for thinking not. Most obviously, GT makes fails to explicitly recognise a number of characteristics of everyday life, such as normativity of the domain, and the importance of normative critique. For example, the category of "loss rationales" emerged, in which nurses rationalized "why the patient, given his degree of social loss, would, if he lived, now be socially worthless" [p.110]. While it is understandable that nurses would seek such comfort for themselves or the bereaved, in today's more empathetic culture, the judgment that someone might be "socially worthless" causes alarm.

An important weakness of GT is that it assumes, rather than draws attention to, the attitudes of both researcher and researched. Glaser & Strauss' [1967] account of their study of nursing shows GT as very lifeworld-orientated. But two things might have contributed to that success which are not made explicit.

By contrast, Wastell's account [1999] of employing GT exhibit's neither of these characteristics. The studied situation, knowledge management in automotive design, seems to be driven by two sets of prior theories. One set is the 'theories' held by automotive designers of what it is 'ok' to comment upon when questioned: the issues that emerge are all negative (grumbles), with consequent omission of important positive everyday factors. The other set is theories in the literature on knowledge management, which is evidenced by the model of KM success that Wastell builds. The categories that Wastell builds into his model seem to bear no relation to the issues, but rather they "map well into models of the KM process as defined in the knowledge management literature" - which can raise suspicion of prior importing of extant theory - and very little new insight is generated. How the model generated is not explained, except that the model is "deliberately kept simple". Everyday issues that a modicum of common sense would suggest are important in knowledge management - such as trust in, appropriateness of, and coherence of the knowledge - are missing from the model. While the tone of Glaser & Strauss [1967] is one of sensitive understanding, the tone of Wastell's paper is not.

Wastell claims validity for his model on the grounds that its main constructs can be traced back to the open coding process - but this might be undermined by a second weakness in GT discussed below - that it has been endorsed by the participants, that it was triangulated, and that it led to practical changes - but the negativity of the issues suggests that participants might not have been as sensitive to a wide range of issues as Glaser & Strauss' nurses were. In addition, whereas Glaser & Strauss gained their data by careful observation of everyday activity, Wastell gained his from discussion separated from everyday activity (workshops and interviews).

A second weakness might also be found in GT, which is exhibited even in Glaser & Strauss' [1967] account. It is found in the open coding process itself. From the statement "I was afraid of losing my composure when the family started crying over their child" [p.107] Glaser & Strauss abstracted the category 'composure'. But why was not 'afraid' abstracted? Arguably it is more salient that the nurse has a reason for not sharing the emotion of the family; is it not good to empathise? Of the constructed categories, to say that 'age' is the most important might have been misleading; it seems that what really gave the most poignant 'social loss' is when a child dies. Rather than being to do with age per se, might it not have more to do with the utter dependency of the child, and the category of 'self-giving love' that families exercise around children? There might indeed be good reason for Glaser & Strauss' selection of their categories in both cases, rather than this author's alternative suggestions, but it is not presented in their [1967] work. On the other hand, this author might be able to explain their choice of 'age' on the widespread desire for quantitative reasoning at the time rather than a well-founded notion of aspects as spheres of meaning such as Dooyeweerd offers us. It was also awareness of Dooyeweerd's aspects that raised the possibility of 'afraid' (a pistic factor) rather than 'composure' (a social one) being the more important. What this might suggest is that a good suite of spheres of meaning, such as Dooyeweerd's aspects, could be very useful with which to enrich GT.

Dooyeweerd's Aspects Enrich Grounded Theory?

At first sight, it might seem that Dooyeweerd's aspects are prior categories from which GT wishes to escape. But that is not so. The concern that motivated Glaser and Strauss's original Grounded Theory proposal was was "an overemphasis in current sociology on the verification of theory, and a resultant 'de-emphasis' on the prior step of discovering what concepts and hypotheses are relevant for the area that one wishes to research" [p.1], ".. that the emergence of categories will not be contaminated by concepts more suited to different areas" [p.37] and that "The consequence is often a forcing of data, as well as a neglect of relevant concepts and hypotheses that may emerge" [p.34].

Dooyeweerd's aspects are not categories in this sense, but rather spheres of meaning on the basis of which the researcher can formulate useful categories. Much of the practice of GT presupposes these very spheres of meaning. For example, asking 'open' questions like "Tell me about the process you go through" "How does this benefit the company?" presuppose that process and benefits are meaningful - which is simply a restatement of the formative aspect. Glaser & Strauss do recognise that incidents could be "coded in a multitude of ways" and that "Without theoretical criteria, delimiting a universe of collected data, if done at all, can become very arbitrary and less likely to yield an integrated product ..." [p.113]. Dooyeweerd's aspects are just theoretical criteria in the widest possible sense and since, Dooyeweerd argued, they cohere, they might yield an integrated product.

Summary: Grounded theory, in its original 1967 vision, has some important everyday-orientation especially relating to seeking diversity of the researched domain and being sensitive to the everyday lives of both researched and researcher. It lacks some important characteristics and tends to assume, rather than encourage, an everyday attitude in the researcher. It might be enriched by the notion of distinct spheres of meaning as found in Dooyeweerd's aspects.

MAKE Re-emphasises Theory Generation

Winfield's [2000] MAKE (Multi-Aspectual Knowledge ELicitation) is a methodology for eliciting all relevant concepts of a situation in all relevant aspects. It is useful in the first stage of Grounded Theory. MAKE helps to re-emphasise theory generation. MAKE helps to separate out categories so as to reduce 'contamination'. MAKE does not force but, in fact, liberates and stimulates. MAKE helps overcome the neglect of relevant concepts.

MAKE is exactly aimed at theory generation, rather than verification, and thus is aimed at the appropriate task. Of the two main steps in generating (grounded) theory, MAKE can contribute primarily to the first, in bringing to light the relevant information from which theories may be built. Winfield [2000] does also refer to its use in theory generation, but we do not discuss that here.

MAKE fulfils the requirements that Glaser and Strauss claim are necessary for a good theory [1967:3]. First, "The theory should provide clear enough categories and hypotheses so that crucial ones can be verified in present and future research; they must be clear enough to be readily operationalized in quantitative studies when these are appropriate." The clarity of categories MAKE generates is enhanced by virtue of the focus on single aspects, each of which is a clearly distinguished sphere of meaning, and by separating them out clearly the concepts become untangled. Second, "The theory must also be readily understandable to sociologists of any viewpoint, to students and to significant laymen." The concepts that are generated by means of MAKE are inherently likely to be "readily understandable to sociologists of any viewpoint, to students and to significant laymen" because the kernel meaning of each aspect is readily grasped by the intuition (as Dooyeweerd claimed and as both Winfield [2000] and Lombardi [2001] have found in practice), so that it does not take an expert to understand it. Third, "Theory that can meet these requirements must fit the situation being researched, and work when put into use. By 'fit' we mean that the categories must be readily (not forcibly) applicable to and indicated by the data understudy; by 'work' we mean that they must be meaningfully relevant to and be able to explain the behavior under study." The concepts, and hence the theories compiled from them, are exactly those that are "meaningfully relevant" and the process by which MAKE enables the interviewee to generate their own meaningful categories ensures the 'fit'.

MAKE Reduces Contamination

The 'contamination' of categories by concepts "more suited to different areas" that Glaser and Strauss [1967:37] wish to avoid is accounted for by Dooyeweerd as attempts to reduce the meaning of one aspect to another. MAKE helps to reduce the incidence of such contamination by enabling and encouraging the interviewee to be sensitive to any unnaturalness in trying to place a concept within an aspect that has already been identified, and making it easy for them to identify a new aspect in which the concept is meaningful. Far from forcing, this can liberate the interviewee, and we shall now discuss why this is so.

MAKE Liberates and Stimulates rather than Forces

The basic reason why MAKE does not force is that aspects are not categories. Rather, they are spheres of meaning within which we form our own categories. Far from constraining, MAKE frees and stimulates the interviewee to generate their own categories and to uncover those they would usually take for granted and fail to bring into the process of theory generation.

The reason this is so is as follows. The categories should be grounded. In Dooyeweerdian terms, this is equivalent to saying that they should be those that might be found in the rich scenario of everyday living in the situation. Grounded Theory goes a little further, saying the categories should be grounded in the data. But 'data' implies an extra step of analytical distinction (of what is salient from what is not) between the rich scenario and our formation of categories. It is this extra step to which Dooyeweerd would draw our attention.

Dooyeweerd argued that such a step is by no means 'neutral' but that the very act of making distinctions itself is grounded on our selection of a particular sphere of meaning (i.e. aspect) from which to view the situation. Dooyeweerd contended that such spheres of meaning are irreducible to each other, in that the meaning of one sphere (say, the juridical) cannot be cast in terms of that of another (say the physical) without denaturing it. (The possibility of metaphorical inter-aspect translations was also discussed by Dooyeweerd but will be ignored here since it does not materially affect our argument.) By long reflection, Dooyeweerd delineated at least fifteen such spheres of meaning.

Aspects are like the our eyes that enable us to view the situation. Just as when colour blindness results from absence of certain colour cells in the retina, so aspectual blindness results when the observer selects only a subset of aspects from which to view the situation. In much scientific work, it is common to narrow this down to a single aspect (for example, physical sciences do not usually include in their theories issues like justice).

Forcing can occur for two reasons. One is when the analyst provides pre-selected categories within an aspect, which is the situation that often pertained in sociology prior to Glaser and Strauss. A Dooyeweerdian account of why this is detrimental is that the full meaning of the aspect is not allowed to be developed. But taking note of the kernel meaning of each aspect we can overcome this. For example, the business and economic sciences focus on production and consumption and a financial medium of exchange, whereas to Dooyeweerd the kernel meaning of the economic aspect is frugality and the skilled management of limited resources. So the widespread assumption that it is the role of business to maximise something (e.g. owner value [====, ]) is not only revealed as just one assumption that may be questioned but is revealed as deeply flawed. Interviewees may be forced, not only by pre-selected categories but also by the cultural assumptions and values pertaining in the group setting in which the discussion is carried out, to focus on issues of maximization rather than those of frugality. Careful understanding of the kernel meaning of aspects can overcome such forcing.

The other is when the analysis is undertaken from within a narrow range of aspects. Then meaning that is relevant is overlooked and so relevant concepts are neglected. This leads us to ...

Overcoming Neglect

Glaser and Strauss [p.35] say that "multiple formal theories are also necessary, since one theory never handles all relevancies, and because by comparing many theories we can begin to arrive at more inclusive, parsimonious levels."

Since aspects are distinct spheres of meaning and Dooyeweerd's suite of aspects attempts complete coverage of all possible meaning, relevance is always with respect to a particular aspect (or sometimes a few aspects). Since MAKE is adept at helping the interviewee to realise the wide range of relevant aspects, it constitutes a good starting point for generating such inclusive theories, and since it is efficient in this process, the result is likely to be reasonably parsimonious.

Winfield found that, on average, around 13 out of 15 aspects were identified as relevant by the interviewee using MAKE. Our study agrees, showing that on average interviewees identified 9 out of the 10 post-biotic aspects.


Strauss and Corbin's Version of GT

Here is an earlier brief comparison of Strauss and Corbin's [1998] version of GT with Dooyeweerd. Some of the points are also relevant to the G+S version.

That the aspects do not inhibit or distort much has been demonstrated in the use of Winfield's Multi-Aspectual Knowledge Elicitation method.

However, on one point Strauss & Corbin's version of GT is different from a Dooyeweerdian approach. SCGT is rather mechanical, while Dooyeweerd's approach emphasises the human being who is the researcher, with their beliefs and preferences, and urges that we take them fully into account in our research. In this way, Dooyeweerd is more akin to the original version of GT.


References

Glaser and Strauss (1967) The DIscovery of Grounded Theory. New York: Aldine de Gruyter.

Strauss and Corbin (1998) ====.


This page is part of a collection of pages that links to various thinkers, within The Dooyeweerd Pages, which explain, explore and discuss Dooyeweerd's interesting philosophy. Email questions or comments would be welcome.

Copyright (c) 2002 Andrew Basden. But you may use this material subject to conditions.

Number of visitors to these pages: Counter. Written on the Amiga and Protext.

Created: 25 April 2003 Last updated: 12 March 2004 link to MAKE. 5 November 2004 Added major first section on Glaser and Strauss' original version, and new Intro. 4 September 2009 New sections, from paper being written, on original GT, its critique and with Dooyeweerd; some modification of S&CGT.