Copyright Sociological Research Online, 1996

 

Raymond M. Lee and Nigel Fielding (1996) 'Qualitative Data Analysis: Representations of a Technology: A Comment on Coffey, Holbrook and Atkinson'
Sociological Research Online, vol. 1, no. 4, <http://www.socresonline.org.uk/1/4/lf.html>

To cite from articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 18/11/96      Accepted: 16/12/96      Published: 23/12/96

A Comment on Coffey, Holbrook and Atkinson

1.1
In their article 'Qualitative Data Analysis: Technologies and Representations', Coffey et al. (1996) note that the ethnographic enterprise, especially but not exclusively in anthropology, has become fragmented. Traditional methods and styles of data collection, analysis and reportage are no longer seen as unproblematic. In all of this is felt the impact of contemporary cultural perspectives, such as poststructuralism, postmodernism, feminism and postcolonialism, through which 'runs a discursive turn, treating as central but problematic the relations of language, knowledge and power' (1996: ¶1.3). Coffey et al. go on to suggest that one strand of contemporary computing, that involving the use of hypertext software and hypermedia, is well-adapted to the diverse range of representational strategies opened up by contemporary critiques. However, they contrast both the theoretical ferment they describe and the exciting potential of hypertext with what they describe as a 'centripetal tendency'. What is represented by this tendency is:

... a convergence, endorsed by some qualitative researchers and methodologists, towards a single ideal-type of data collection storage and analysis. That model combines computing techniques with methodological perspectives claimed to be associated with 'grounded theory'. One can detect a trend towards a homogenization, and the emergence of a new form of orthodoxy, especially at the level of data management. We note that the use of microcomputing strategies for qualitative data handling has become widespread, and this includes an almost globalizing process within the research community. The presuppositions and procedures that are inscribed in contemporary software for qualitative data analysis are implicitly driving a renewed orthodoxy that is being adopted in a large number of research sites around the world. (1996: ¶1.4)

1.2
We do not disagree with Coffey et al.'s characterization of the current state of ethnography, at least within anthropology. Nor would we dissent from their assessment of the analytic possibilities opened up for qualitative researchers by hypertext and hypermedia. However, neither of these arguments, we believe, depend on a contrast with an assumed orthodoxy. Put bluntly, their characterization of the current state of qualitative computing is a red herring. Moreover, we believe, their assertion of a link between CAQDAS (Computer Assisted Qualitative Data Analysis Software) and grounded theory is overdrawn.

CAQDAS as Orthodoxy

2.1
The notion of an orthodoxy implies that a healthy plurality of approaches is being replaced by some kind of dominant discourse which suppresses variation and homogenises practice. The first point to make about this is that claims to this effect, or indeed counter-claims, must inevitably be conjectural. Traditionally, discussion of analytic procedures did not feature loudly in the literature on qualitative research (Miles, 1983). That literature tended to be dominated by discussions of research ethics, rather little on the practicalities and even less on data analysis. To put the point in another way, the data which would allow us to make comparisons between past and present in relation to qualitative data analysis are simply not there. We can, however, contrast what we believe to be the situation in qualitative research with other fields where computers have increasingly begun to be used.

2.2
Kling and Iacono (1988) describe how the growing use of computers in a number of areas has been driven by what they term 'computerization movements'. Computerization movements can be regarded as social movements in the sense articulated by Blumer (1969); 'collective enterprises to establish a new order of life'. Kling and Iacono suggest that computerization movements share a number of key ideological beliefs. Broadly speaking these are (a) that there is a need to remain at the cutting edge of technological development if a particular area of social life is to develop, (b) that computerization has only beneficial consequences and (c) that it is the resistance of users which provides the primary barrier to further computerization. There is little doubt that many of those involved in CAQDAS are enthusiastic about the possibilities opened up by the application of new technologies to qualitative research. We would contend, however, that it is difficult to discern in the literature on computer methods in qualitative research anything like sentiments of the kind Kling and Iacono identify. Far from propagating an orthodoxy, developers, popularizers and commentators have often stressed the need for epistemological awareness and debate in relation to software use. This is explicit, for example, in the introductions and contributions to two widely-used edited collections of articles on CAQDAS (Fielding and Lee, 1991; Kelle, 1995). Instead of insisting on a particular model of analysis, we ourselves see computer-based methods as permitting, if we can use a somewhat inelegant term, the multitooling of qualitative researchers, making available to them more or less at will a wide range of different analytic strategies.

2.3
Coffey et al. point out that procedures for coding and retrieving segments of text from fieldnotes or interview transcripts are common in qualitative data analysis software packages. For them, the injunction to code data implied by the software is an aspect of the alleged orthodoxy they want to reject, in part because they see it as 'akin to standardized survey or experimental design procedures' (1996, ¶7.6). One can make a number of points here. First, coding, like linking in hypertext, is a form of data reduction, and for many qualitative researchers is an important strategy which they would use irrespective of the availability of software. One of the striking features of the history of CAQDAS is that many of the early programs emerged more-or-less simultaneously from developers unaware that others were doing the same thing. In most cases, developers saw themselves as computerizing their own analytic methods and with little thought that these methods were unusual. A similar point can be made about the recurrent reinvention of word-processing based approaches for amalgamating text passages on similar topics into separate files. Second, Coffey et al. never address themselves to what analysts do once they retrieve the material they have coded. Our research with users (Fielding and Lee, forthcoming) suggests that CAQDAS users take many different approaches, a point we will return to later. Even, at minimum, the ability effectively to manage data may be a considerable improvement over the ad hoc procedures we suspect frequently underpinned manual analysis. Not to acknowledge this, it seems to us, is to elide three things; program features, analytic procedures and methodological approaches. Third, it is this elision that allows Coffey et al. to equate coding in qualitative research with analogous procedures in survey research. Yet, despite the obvious Lazarsfeldian influence on grounded theory, the quite explicit rules which govern the coding of open-ended survey responses (Lazarsfeld and Barton, 1951) are rarely invoked in qualitative practice.

2.4
To say all of this this is not to say that coding is the only, the best, or even the preferred method for the analysis of qualitative data. Indeed, we would suggest that coding as a procedure in qualitative research, as well as the advent of approaches such as narrative analysis, is historically contingent and reflects in part the impact of a new technology, the tape recorder. What we do claim is that software tools for approaches not based on coding are less invisible than Coffey et al. suggest. Coffey and Atkinson (1996) have suggested that text retrieval programs may be more useful for discourse analytic purposes than so-called 'code and retrieve' packages. Weitzman and Miles (1995) which provides detailed comparative reviews of software programs potentially useful for qualitative analysis describe a range of software types, including text retrievers, textbase managers and conceptual network builders, different from code-and-retrieve packages. Moreover, developers of CAQDAS programs have increasingly included facilities for proximity searching, which may be useful for narrative analysis, and for 'autocoding' which could be adapted to some kinds of semiotic analysis. The provision of new features in CAQDAS programs reflects the generally close relationship between users and developers characteristic of the field, and the general willingness of developers to incorporate features desired by users even if these do not always accord with the epistemological preferences of the developer. Since packages increasingly support procedures, routines and features which are new to qualitative analysis or make procedures possible that were not practical without the power of the computer, it is less and less plausible either to argue that the software is merely an aid to code-and-retrieve or to argue that code-and-retrieve is the sine qua non of qualitative analysis. Besides code-based procedures and hypermedia approaches, qualitative researchers have available to them software resources based on production rule systems (Heather and Lee, 1995). The application to case data of techniques such as Boolean minimization also depends on software solutions (Ragin, 1995).

CAQDAS and Grounded Theory

3.1
Citing with approval Marrku Lonkila's (1995) work examining the influence grounded theory has had on software developers, Coffey et al. argue that 'aspects of grounded theory have been over-emphasized in the development and use of qualitative data analysis software, while other approaches have been neglected in comparison'. Before accepting this position too readily, however, it may be useful to explore in a more critical way what we might mean by the term 'grounded theory'. Jennifer Platt (1983, 1995) has argued that the history of qualitative research has been mythologized. Although we cannot do full justice to her argument here, the message we take from Platt's careful debunking of qualitative methodology's 'creation myth' is that we should be careful not to mythologize the status of grounded theory, too. When qualitative researchers are challenged to describe their approach, reference to 'grounded theory' has the highest recognition value. But the very looseness and variety of researchers' schooling in the approach means that the tag may well mean something different to each researcher. A detailed examination of work claiming the label may deviate sharply from what Glaser and Strauss had in mind (and, of course, their own work has developed substantially, and their approaches have diverged). There are, in fact, several logically-possible implications: (i) work that is done in the qualitative vein bears little resemblance to grounded theory procedures but lays claim to do so for purposes of legitimation; (ii) the influence and application of the original grounded theory procedures is limited to a far smaller group than claim the label; (iii) significant methodological developments outside grounded theory have not been applied to, or brought into line with, the original tenets of grounded theory (i.e., the original inspiration has grown stale and out of touch with research practice); (iv) subsequent formulations by the authors of grounded theory following publication of the original work have not been systematically taken to heart by those claiming to work in the grounded theory tradition. We should neither assume that qualitative research only involves grounded theory nor that CAQDAS supports only a grounded theory approach. There are other viable approaches to qualitative analysis, and indeed the practice of qualitative researchers may be rigorous without their explicitly employing any of the several approaches blessed with a memorable 'label' and a school of avowed followers.

3.2
The implication of all this for our understanding of CAQDAS is that we must not allow the easily-declared assertions of affiliation to a grounded theory approach to prefigure our understanding of what CAQDAS can (or should) do to support qualitative analysis. For similar reasons we prefer an empirical approach - however limited - based on researchers' own practice to adjudicate matters such as the influence of grounded theory on computer-assisted qualitative analysis. (As a general point we can observe that much of the literature on research practice is based on assertion. There is actually relatively little empirical study of how researchers behave in practice.) To identify research studies where it was likely a computer had been used to analyse qualitative data, we inspected the Social Science Citation Index for studies citing John Seidel's descriptive writings on the ETHNOGRAPH (Seidel and Clark, 1984; Seidel, 1985; 1988), a package often thought to be avowedly oriented to grounded theory. In all we found 163 studies which reported empirical findings. Of these studies 50, or 30.7%, cited a work associated with the grounded theory tradition; 70% did not. (Works we took as being associated with grounded theory were: Glaser and Strauss, 1967; Glaser, 1978; Strauss, 1987; Strauss and Corbin, 1990.) The table below shows methods texts cited in those articles which did not cite a work in the grounded theory tradition.


Seidel citations: Methods books cited by those NOT citing works associated with grounded theory
TitleNumber
Miles and Huberman, Qualitative Data Analysis14
Patton, Qualitative Evaluation Methods13
Lincoln and Guba, Naturalistic Evaluation12
Spradley, The Ethnographic Interview11
Lofland, Analysing Social Settings7
Bogdan and Biklen, Qualitative Research for Education6
Tesch, Qualitative Research4
Fetterman, Ethnography Step by Step4

3.3
From this we conclude that grounded theory is an important, but by no means ubiquitous, influence on studies where there is a strong likelihood that computer-based analysis has been used. Users of qualitative data analysis software apparently draw on a range of methodological inspiration. Admittedly, almost all of the studies considered here take what Coffey et al. describe as a realist approach to the representation of social reality. However, in our own research with users of qualitative analysis programs (Lee and Fielding, 1995) we have found little hesitancy about abandoning program use in cases where software did not meet the analyst's needs or where it was perceived to be at variance with the researcher's epistemological presuppositions. Our suspicion is that rather than an assumed orthodoxy inhibiting the take-up of a new paradigm, what we are largely seeing here is a cultural lag. It is only comparatively recently that texts have become available that describe in detail the analytic procedures associated with discursive approaches. (We have in mind here Coffey and Atkinson's (1996) own excellent text; see also Alasuutari, 1995.)

Hypertext

4.1
If there is an affinity between CAQDAS and a methodological tradition in qualitative research, we believe it to be with those code-based lineal descendants of analytic induction (Fielding and Lee, forthcoming). The affinity between grounded theory and computer methods may lie elsewhere, as one of the author's of Coffey et al.'s article has demonstrated (Weaver and Atkinson, 1994). In evaluating the hypertext program GUIDE, Weaver and Atkinson show how hypertext methods fit well with grounded theory procedures. We believe their characterization to be rather apt. Whatever else it may be, the grounded theory approach is also a data management strategy. That strategy combines a tough-minded reluctance to collect more data than is theoretically necessary with an expansive concern to seek theoretically relevant data wherever it may be. The tools for doing this, memoing, theoretical saturation and theoretical sampling, depend on links, associations and trails which are difficult to maintain. Hypertext provides a technical means for doing so.

4.2
Of course, Coffey et al.'s intention is not to rescue grounded theory as an analytic approach, rather they seek to transcend it. As they put it:

It is, after all, part of the rationale of ethnographic and similar approaches that the anthropologist, sociologist, historian, psychologist or whoever, recognizes the complexity of social inter-relatedness. We recognize the over-determination of culture, in that there are multiple, densely coded influences among and between different domains and institutions. It is therefore part of the attraction of hypertext solution that a sense of dense interconnectedness is preserved, enhanced even, while linearity is discarded. (Coffey et al., 1996: ¶8.5)

4.3
Connected to this is a blurring of traditional distinctions between 'analysis' and publication'. Using hypertext, a reader rather than simply consuming an ethnography would be able to interact with the analysis interactively, flexibly and in a non-linear way.

4.4
We find the possibilities Coffey et al. sketch intriguing and exciting. As ever, however, the proof of the pudding is in the eating. Weaver and Atkinson (1995) and, earlier, Cordingley (1991) have pointed to some of the technical limitations of hypertext systems as qualitative research tools. We suspect that for the future the logic of inquiry supporting hypertext needs to be further explicated. The use of hypertext, and beyond that multimedia approaches, facilitates 'untutored use'. Although we do not use the term 'untutored' in a pejorative way, nevertheless we suspect that there is an issue about what background one might need to produce meaningful interpretations from a hypertext resource. In other words, hypertext may make it feasible for anyone to create the appearance of an ethnographic text, but a text which its creator finds impossible to explain or defend. If we understand Coffey et al. correctly, it is the originating ethnographer who chooses items for expansion and reference and who specifies links through the hypertext database. However, it is precisely this which may subvert the very polyvocality Coffey et al. wish to facilitate. Faced with an apparently smooth and user-friendly resource offering all manner of subsidiary and supporting information, the naive user may feel that it contains 'all there is to know' about the topic at hand. A resource seen by its architects as encouraging a sophisticated appreciation of the very contingency of social knowledge will instead be received and used as watertight, supremely- and ironically-authoritative. Alternatively, the tendency of hypertext to blur distinctions between 'data', 'analysis', 'interpretation' and so on may for some be simply a recipe for confusion and indecision produced by a maze of links and connections going nowhere in particular.

4.5
Coffey et al. end their paper on a modest note disclaiming prophetic status. Although we are unconvinced by their criticism of existing computer-based approaches to qualitative data analysis, we are actually excited by the possibilities they envisage for hypertext. Qualitative researchers have gone beyond seeing the computer either as a panacea for analytic woes or as a devil-tool of positivism and scientism. Let a hundred flowers bloom!

References

ALASUUTARI, P. (1995) Researching Cuture: Qualitative Method and Cultural Studies. London: Sage.

BLUMER, H. (1969) 'Social movements' in B. McLaughlin (editor) Studies in Social Movements: A Social Psychological Perspective. New York: Free Press.

COFFEY, A. & ATKINSON, P. (1996) Making Sense of Qualitative Data: Complementary Research Strategies. Thousand Oaks, CA: Sage.

COFFEY, A., HOLBROOK, B. & ATKINSON, P. 1996. 'Qualitative data analysis: technologies and representations', Sociological Research Online, vol. 1, no. 1, <http://www.socresonline.org.uk/1/1/4.html& gt;.

CORDINGLEY, E. (1991) 'The Upside and Downside of Hypertext Tools: The KANT Example' in N. G. Fielding & R. M. Lee (editors) Using Computers in Qualitative Research. London: Sage.

FIELDING, N. G. & LEE, R. M. (1991) Using Computers in Qualitative Research. London: Sage.

FIELDING, N. G. & LEE, R. M. (forthcoming) Computer Analysis and Qualitative Research. London: Sage.

GLASER, B. G. (1978) Theoretical Sensitivity. Mill Valley, CA: Sociology Press.

GLASER, B. G. & STRAUSS, A. L. (1967) The Discovery of Grounded Theory. Chicago: Aldine.

HEATHER, N. & LEE, R.M. (1995) 'Expert Systems for the Social Scientist' in R.M. Lee (editor) Information Technology for the Social Scientist. London: UCL Press.

KELLE, U. (1995) Computer-Aided Qualitative Data Analysis: Theory: Methods and Practice. London: Sage.

KLING, R. & IACONO, S. (1988) 'The Mobilization of Support for Computerization: The Role of Computerization Movements', Social Problems, vol. 35, pp. 226 - 43.

LAZARSFELD, P. F. & BARTON, A. H. (1951) 'Qualitative Measurement in the Social Sciences: Classification, Typologies and Indices' in D. Lerner & H.D. Lasswell, (editors) The Policy Sciences: Recent Developments in Scope and Method. Stanford, CA: Stanford University Press.

LEE, R. M. & FIELDING, N. G. (1995) 'Users' Experiences of Qualitative Data Analysis Software' in U. Kelle (editor) Computer-Aided Qualitative Data Analysis: Theory, Methods and Practice. London: Sage.

LONKILA, M. (1995) 'Grounded Theory as an Emerging Paradigm for Computer-Assisted Qualitative Data Analysis' in U. Kelle (editor) Computer-Aided Qualitative Data Analysis: Theory, Methods and Practice. London: Sage.

MILES, M. B. (1983) 'Qualitative Data as an Attractive Nuisance: The Problem of Analysis' in J.V. Maanen (editor) Qualitative Methodology. Beverly Hills: Sage.

PLATT, J. (1983) 'The Development of the "Participant Observation" Method in Sociology: Origin Myth and History', Journal of the History of the Behavioral Sciences, vol. 19, pp. 379 - 93.

PLATT, J. (1995) 'Research Methods and the Second Chicago School' in G. Fine (editor) A Second Chicago School: the Development of a Postwar American Sociology, London: University of Chicago Press.

RAGIN, C. C. (1995) 'Using Qualitative Comparative Analysis to Study Configurations' in U. Kelle (editor) Computer-Aided Qualitative Data Analysis: Theory, Methods and Practice. London: Sage.

SEIDEL, J. V. (1985) The Ethnograph Verson 2.0 User's Manual. Littleton, CO: Qualis Research Associates.

SEIDEL, J. V. (1988) The Ethnograph Verson 3.0 User's Manual. Littleton, CO: Qualis Research Associates.

SEIDEL, J. V. & CLARK, J. A. (1984) 'THE ETHNOGRAPH: A Computer Program for the Analysis of Qualitative Data', Qualitative Sociology, vol. 7, pp. 110 - 25.

STRAUSS, A. L. (1987) Qualitative Analysis for Social Scientists. Cambridge: Cambridge University Press.

STRAUSS, A. L. & CORBIN, J. (1990) Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Newbury Park, CA: Sage.

WEAVER, A. & ATKINSON, P. (1995) Microcomputing and Qualitative Data Analysis. Aldershot: Avebury.

WEITZMAN, E. A. & MILES, M. B. (1995) Computer Programs for Qualitative Data Analysis. Thousand Oaks, CA: Sage.

Copyright Sociological Research Online, 1996