Volume 7, No. 2, Art. 24 – March 2006

Review:

Robert L. Miller

Lyn Richards (2005). Handling Qualitative Data: A Practical Guide. London: Sage Publications, xii + 207 pages, ISBN: 0 7619 4258 0 (hardcover) £70.00, 0 7619 4259 9 (pbk) £21.99

Abstract: Handling Qualitative Data: A Practical Guide is an introductory textbook covering all stages of qualitative research from the initial conceptualisation of a project, through data collection and analysis, to writing up. The author, Lyn RICHARDS, is a well-known developer of two key qualitative software analysis packages, NUD*IST and NVivo. While RICHARDS clearly advocates the use of qualitative analysis software, the text is "generic" and could be used in tandem with any qualitative software package. The book concentrates on practical advice about the use of software to manage and analyse qualitative data, and provides insights in these areas. The consideration of issues around team-based qualitative research is another strong point. However, due in part to its short length, the overall coverage of topics tends to be superficial. In itself, the book does not provide sufficient detailed support for a student who would like to use it as her/his main source of guidance for carrying out a qualitative research project.

Key words: software, introduction, CAQDAS, textbook

Table of Contents

1. Context

2. Structure of the Text

3. Evaluation

Notes

References

Author

Citation

 

1. Context

When a developer of two of the most widely-used and influential software packages for analysing qualitative data, NUD*IST and NVivo (see e.g. BAZELEY & RICHARDS, 20001)), produces a textbook with the title Handling Qualitative Data: A practical guide, even that innocuous-sounding title has an "edge". NUD*IST and NVivo are successful entities in a limited market noted for its competition and for the lack of cooperation between protagonists. The failure of the major qualitative analysis software programs to evolve means of transferring material from one package to another is a classic example of the "prisoner's dilemma" paradox where, unfortunately, minor short-term gain has prevailed over major long-term mutual benefit. The inability to move data between packages has slowed the uptake of qualitative software generally and, of more significance, constitutes a barrier to qualitative research generally—particularly that which involves cooperation or comparison. Could this textbook be a commercially-motivated gambit, an attempt to emulate the success of SPSS among quantitative analysis packages by locking large numbers of students into NUD*IST/NVivo from the outset of their careers, thereby laying the groundwork for maintaining a market dominance? [1]

Fortunately, this is not the case. While the textbook assumes that the reader will be working with a qualitative data analysis package, RICHARDS does not use the book as a platform to promote her own software (see also MORSE & RICHARDS 20022)). She is upfront in informing the reader that she is the creator of several well-known packages and has tried to ensure that the text is written in as "generic" a style as possible. Though each chapter is paired with a set of online tutorials on the NVivo package located at the Sage Publications website, the text could be used with some other qualitative package. The book of course fits neatly with NUD*IST/NVivo (it would be odd if it did not) and it may be that some of the jargon used and features mentioned are related to those packages but this is understandable from someone who must be steeped in her own creations. [2]

A broader question is whether someone who has developed several of the premier packages for the computerised analysis of qualitative material can have the balance necessary in order to produce a general qualitative textbook. For better or worse, misinformed or not, many qualitative researchers have misgivings or a positive aversion to the computerisation of qualitative analysis. Is the textbook bound to fall into a "realist" camp? Here, again, I would give a (albeit qualified) "Not guilty". A basic premise of the text is the dismissal of those who reject the use of qualitative software. Anyone devoted to a completely holistic approach that does not employ any means of organising material other than absorption will not find this text congenial. [3]

RICHARDS' strategy for making her textbook open to as wide a constituency as possible is to emphasise the data management and record-keeping advantages of qualitative software. Features of qualitative packages such as the ability to cross-index without destroying the structure of data, to carry out retrievals of information employing both simple and complex searches, to maintain a reliable record of the steps and blind alleys carried out over time in an analysis, etc., are stressed. This strategy, while inclusive, paradoxically does weaken the text. It is the more controversial aspects of qualitative software, when it is used actively for analysis rather than data management, that are the least understood. A more partisan approach that addressed more forcefully how computers are employed to analyse qualitative material and, arguably, to generate concepts, would have moved the text away from a neutral ground, but would have made it more informative for the reader who is seeking to explore the widest potential of qualitative software. [4]

2. Structure of the Text

The text's applied focus and its avowed intention of taking the student through all stages of a qualitative project from initial conception to writing up (termed by RICHARDS, "telling") sets its structure. The book is divided into three sections: "Setting up", "Working with the data", and "Making sense of your data", though in fact the central section on "data work" dominates. Chapter 1, "Setting up your project", begins with admonitions to consider the purposes and goals of a research project and to think through the types of data that will be required. Several of what turn out to be recurring themes of RICHARDS' approach are present in this initial chapter. RICHARDS advocates "starting a log trail"—a precursor to later chapters in which the keeping of research memos and retaining a record of avenues of thought and enquiry, including dead-ends, is central. ("Memos" indexed by date and the ability to keep a record of the temporal sequence of lines of enquiry are of course features that are facilitated by a software analysis package.) The choice of software receives a prominent mention from the outset (though RICHARDS does not state the prime reason that a software choice must be made carefully—that the lack of compatibility between packages will very quickly lock the researcher into their initial choice). [5]

Chapter 2, "Qualitative data", begins by attempting to explain the nature of qualitative data. RICHARDS deliberately avoids drawing a distinction between qualitative and numeric/quantitative data. However laudable this stance is, there is a cost. The contextual, unique nature of true qualitative information is not presented with clarity and a student new to social research could be left confused, believing that qualitative data is somewhat akin to a wide range of responses to an open-ended survey question. The author scans sources of qualitative information, particularly interviewing. This scan is extremely brief and would need considerable expansion before it became a guide to methods of qualitative data collection. The overview of data collection methods is confused by the inclusion of the intellectual work around fieldwork, such as discussion with co-researchers, as a mode of data collection itself. Again, keeping a "log trail" is emphasised, here as a means of reflexively recording one's own thoughts about the project. [6]

Chapter 3, "Data records", is about the process of recording data. Its stance is that decisions taken at the point of data recording do imply taking a stance on what will be considered relevant for an analysis and so this chapter could be considered part of an expanded Section II, "Working with data". Interestingly, RICHARDS assumes that the researcher probably will not be attempting to record everything and that "data reduction" of some sort will be taking place. This realistic approach is buttressed by suggestions about how to ensure that significant information is not being lost. Much of the latter part of the chapter is devoted to discussion in a generic manner of the steps involved in contextualising and importing data into a software package. [7]

Section II formally begins with Chapter 4, "Up from the data", which centres upon the initial abstraction and generalisation from a batch of qualitative data. RICHARDS advocates focusing on "interesting" sections of data and looking at its context, consequences and the hypothesised strategies of the actor. Her most strong statement about the process of memo-writing/annotation is given here, as a means of recording the ideas that arose from the initial interaction with the data. [8]

Chapter 5, "Coding", considers the process that most would associate most strongly with qualitative software packages. RICHARDS distinguishes between three types of coding: descriptive coding, recording substantive characteristics in a way akin to a quasi-quantitative analyst; topic coding, coding material into a subject-based structure; analytical coding, coding data into an evolving structure based upon the analyst's ongoing interpretation of the action. [9]

The latter part of this chapter reflects RICHARDS' practical experience, advocating coding consistency tests (between multiple coders and of yourself over time) and avoiding "the coding trap" (getting carried away with ever more elaborate coding schemas that will never feed into a subsequent analysis). [10]

The core of Richard's approach to qualitative data analysis probably is located in Chapter 6, "Handling ideas". Continuing her emphasis upon recording thoughts through memo-writing, this chapter shows a closer link than most to NUD*IST/NVivo in its use of the ideas of "nodes" (sets of categories) and "trees" (linked sets of categories/nodes). It contains a core statement of her approach to analysis. In RICHARDS' view, "organizing is necessary for creativity":

"As you form a category system, the project changes radically. Your uncoordinated collection of ideas is transformed into a data management system, a filing system, which, like the library catalogue, will ensure that you know where a node is, and where a new node should go. In the process you will have pruned or merged a lot of the categories you made earlier, and the result is a smaller, tighter and organized system of ideas" (p.112). [11]

RICHARDS' own experience of working with tree/node systems shows here with practical advice on how to work with category systems and how to avoid problems with them. This is a brave attempt to explain the thought processes an analyst would go through in manipulating node connections when developing a conceptual structure in an analysis. She also advises about how to avoid the pitfalls of having a team of researchers working on a common category system. [12]

Chapters 7 and 8 are located in Section III, "Making sense of your data", but in fact again could be considered part of a greater "data analysis" section. Chapter 7, "What are you aiming for?" centres on the general problem of how one knows when the outcome of an analysis is beginning to take shape. RICHARDS' basic answer is: when the analysis has generated at least a new "local theory or explanation", a model. However, this is qualified by a variety of other, more specific, criteria. For example, the analysis should meet the goals of the project, it should be more than description, and the results should be usable in the sense of having a practical utility. Attaining "saturation", having a model or explanation that works at different levels, "simplicity", "elegance and balance", "robustness", "completeness", and face validity are all considered positive signs. There is a discussion of validity, but it is qualified: "You will find a recent literature offering new and interesting ways of gauging validity in qualitative research, and new words for them (credibility, transferability, dependability, confirmability)" (p.139). [13]

Perhaps the most explicitly software-related chapter is Chapter 8, "Searching the data". This chapter attempts to give the reader a feel for the process of computerised coding and text searches. It concentrates particularly upon specific code searches and would have been better placed earlier, between Chapters 5 and 6. [14]

Chapter 9 does fall within a closing section, being concerned with ways of arriving at core themes. Under this general rubric, RICHARDS surveys a variety of summarising techniques: coding and category handling; modelling; typologies and matrices; case studies; writing (not given here in the order that she uses in the chapter). "Coding and category handling" and "modelling" pick up the themes of Chapters 5 and 6, albeit at a more generalised level—the use of coding to work towards synthesis and working with the interrelationships between category systems to arrive at "models". "Typologies" are sets of common categories and "matrices" are the possible combinations of typologies. "Case studies" are an "ideal typical" way of depicting results and "writing" leads into the concerns of the final chapter. [15]

The title of the final chapter, "Telling it", is used deliberately by RICHARDS to signify that reporting the results of a research project is more than just writing it up and involves communicating one's findings effectively to an audience. There is good advice about grappling with the overwhelming bulk of qualitative data, overcoming "writer's block", and the much-neglected subject of what a report of a qualitative project should actually contain. The set of cautionary tales about the ways that researchers can stumble at the last post and fail to produce an effective final report ("the patchwork quilt", "the illuminated description", "the leap of faith", "yourself as hero") is exemplary. [16]

3. Evaluation

While the book has some interesting features, the coverage of material is hurried and at a surface level. In itself, the book does not provide sufficient detailed support for a student who would like to use it as their sole source of guidance for carrying out a qualitative project. [17]

Some good points stand out. The dynamics of working in teams are emphasised throughout. One gets the feeling that the author is speaking from experience, with a practical bent to her advice. A positive feature is that a lot of the discussion is oriented towards team research—much of what she says will resonate with members of group projects. The influence of NUD*IST and NVivo is apparent in RICHARDS' advocacy of keeping a trail of decision-making through the constant writing of "memos" and saving earlier drafts of ways of structuring data, including false starts and unhinged codings. A number of figurative devices are employed to help orient the reader (e.g., an elephant icon to signal that a section is about large-scale projects) but I found that these tend to be more irritating than informative. Each chapter begins with an elaborate flowchart that depicts how it fits into the overall chapter structure of the book. Again, I personally found these charts more confusing than informative. There is an attempt to anchor abstract discussion in hypothetical examples, but these examples are minimally introduced and appear embedded, without warning, within discussions of theoretical matters. It would have been preferable if RICHARDS had used a couple of thoroughly introduced examples from the outset and, better still, if she had stuck consistently to the same examples throughout the text, allowing the reader to see an idea come into being and then move from data collection through to analysis, data reduction and "telling" the findings. As presented, the examples do not serve the purposes of clarification or illustration. [18]

There is a good range of "Suggestions for further reading" at the end of each chapter that are designed to help the student who needs more support. However, this provision could have been more "user friendly": instead of giving the full names of authors and the titles or relevant chapters of the suggested books, RICHARDS only provides a basic reference (e.g., STRAUSS & CORBIN, 1998), leaving it up to the student to locate the full title in the References and then to find the relevant sections in the books themselves. [19]

I wondered whether the book is in effect one half of a coherent multi-media teaching tool and checked to see whether the links to the Sage website NVivo tutorials would add depth to the textbook's discussions. The tutorials, however, are a straightforward introduction to the mechanics of NVivo and will take the student little further with the conceptual issues of qualitative research. [20]

In sum, the text does not have sufficient depth to provide a real basis for a student needing a definitive resource about the conduct of qualitative research. [21]

References

Bazeley, Patricia & Richards, Lyn (2000). The NVivo Qualitative Project Book. London: Sage.

Lang, Iain (2004, January). Review: Janice M. Morse & Lyn Richards (2002). Readme First for a User's Guide to Qualitative Methods [13 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research [On-line Journal], 5(1), Art. 28. Available at: http://www.qualitative-research.net/fqs-texte/1-04/1-04review-lang-e.htm [Date of access: January, 9, 2006].

Lichtman, Marilyn (2001, September). Review: Patricia Bazeley & Lyn Richards (2000). The NVivo Qualitative Project Book [25 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research [On-line Journal], 2(3), Art. 26. Available at: http://www.qualitative-research.net/fqs-texte/3-01/3-01review-lichtman-e.htm [Date of access: January, 9, 2006].

Morse, Janice M. & Richards, Lyn (2002). Readme First for a User's Guide to Qualitative Methods. Thousand Oaks, London, New Delhi: Sage.

Strauss, Anselm L. & Corbin, Juliet (1998). Basics of Qualitative Research: Grounded Theory Procedures and Techniques (2nd edition). London: Sage.

Notes

1) See also the review by LICHTMAN (2001). <back>

2) Reviewed by LANG (2004). <back>

Author

Robert L. MILLER is Senior Lecturer in Sociology in the School of Sociology & Social Policy, Queen’s University, Belfast.

Contact:

Robert L. Miller

School of Sociology & Social Policy
Queen’s University
Belfast BT7 1NN
Northern Ireland
United Kingdom

E-mail: r.miller@qub.ac.uk

Citation

Miller, Robert L. (2006). Review: Lyn Richards (2005). Handling Qualitative Data: A Practical Guide [21 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(2), Art. 24, http://nbn-resolving.de/urn:nbn:de:0114-fqs0602244.