Family Practice

This Article Abstract FREE Full Text (PDF) E-letters: Submit a response Alert me when this article is cited Alert me when E-letters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (8) Request Permissions Google Scholar Articles by Barbour, R. S Articles by WoReN, M. o. Search for Related Content PubMed PubMed Citation Articles by Barbour, R. S Articles by WoReN, M. o. Social Bookmarking          What's this?

Family Practice Vol. 17, No. 1, 83-89
© Oxford University Press 2000

Methodologies

Acquiring qualitative skills for primary care research. Review and reflections on a three-stage workshop. Part 2: analysing interview data

Rosaline S Barbour, Members of WoReN

Department of General Practice, University of Glasgow, Glasgow, UK.
Members of WoReN: Maggie Booth (Researcher); Sandra Burley (Lecturer in Practice Nursing); Brian Cook (General Practitioner); Margaret Crawley (Audit Co-ordinator); Chris Elwood (Researcher); Valerie Featherstone (WoReN Research Facilitator); Hayley Gibson (Audit Information Assistant, Cancer Services); Chris Jary (General Practitioner); Heather Middleton (Practice Nurse); Caroline Plews (Research Associate); Gill Rowland (Community Child Health Doctor); Bryony Simpson (Speech and Language Therapist); Shirley Stephenson (Research Health Visitor); and Jenny Wright (Assistant Project Manager, Cancer Services).

Rosaline S Barbour, Department of General Practice, University of Glasgow, 4 Lancaster Crescent, Glasgow G12 0RR, UK.

Barbour RS and Members of WoReN. Acquiring qualitative skills for primary care research. Review and reflections on a three-stage workshop. Part 2: analysing interview data. Family Practice 2000; 17: 83–89.

Received 27 April 1999; Revised 9 August 1999; Accepted 6 September 1999.

Abstract

This paper reflects on one Primary Care Research Network's (WoReN's) experience of running a workshop on analysing qualitative interview data, provided as the second of a three-part workshop concerned with acquiring qualitative interviewing skills. It discusses the aims and limitations of the short workshop format in meeting the needs of practitioners embarking on the process of analysing qualitative data, drawing upon and reviewing the relevant research methods literature. Particular attention is paid to the role of qualitative data analysis computer packages and the debate on ‘grounded theory’. We conclude by making suggestions with regard to designing and running data analysis workshops within primary care.

Keywords. Methodology, qualitative research, research training..

Introduction

This second paper follows on from our discussion about a workshop designed to allow participants to develop skills in generating data through semi-structured interviewing. It describes the second half-day workshop in our series, which attempted to address the practicalities of analysing interview transcripts. The discussion also draws on our third workshop where we began the process of collaboratively writing up our experience of the workshops and considered the challenges involved in presenting qualitative data.

Aims and scope of the second workshop

This workshop sought to give the 12 participants ‘hands-on’ experience of analysing transcripts of interviews, which they had carried out themselves. Rather than seeking to provide a definitive template for this task, it attempted to draw to their attention the complexities and problems which they were likely to encounter, whilst suggesting some approaches which might render the task more comprehensible and thus, amenable to explanation. Ritchie and Spencer¹ charged qualitative researchers with the task of making our analyses more transparent and accessible by providing documentation of our methods and techniques to a wider audience, including commissioners and potential users of research; it was our hope that the workshops would better equip these new researchers to meet and rise to this challenge. Hopefully, the experience of workshop participants in reviewing successive drafts of this paper will itself also have contributed to their awareness of the processes involved in analysis and writing up.

In the real world, the process of analysing qualitative data is a long and complex one. Malterud commented "Certain of the skills required in the qualitative research process—especially the ones related to the analysis of data—may be underestimated by medical researchers entering this field, accustomed to standardized procedures from statistical analysis"² (p. 201). In contrast to this, the ‘virtual reality’ of the workshop situation ensured that we were working with relatively short transcripts of 20-minute interviews. Even allowing for this, however, we hoped to be able to convey to workshop participants the unfolding and iterative³ nature of coding, interpretation, re-coding and re-formulation of emergent themes/theoretical frameworks for understanding the data generated.

Content of the second workshop

In this second workshop, transcripts of interviews carried out at the previous workshop were distributed to participants who had acted as ‘interviewers'. Although, presumably, everyone remembered whom they had interviewed the previous week, these transcripts identified only the ‘interviewer', with the interviewee given a number and described with reference to the details provided on the short pro-forma. Such ‘anonymity’ as this afforded was considered important, given the possibility of individual comments being singled out as part of the collaborative exercises on analysis planned for a later stage in the workshop.

In the intervening week between the first and second workshops, we had engaged the services of two experienced audio typists to produce the transcripts. Whilst this may be something of a luxury often reserved for more experienced researchers, with novices frequently having to do their own transcribing, we decided to delegate this task to ensure both that transcription was carried out soon after completion of the ‘interviews' and that all participants were working with transcripts of the same quality. However, transcribing one's own tapes can be a salutary experience, which, amongst other things, ensures that researchers become more attentive interviewers in the future, asking interviewees to repeat indistinct speech, clarifying concepts or ideas expressed and keeping interviewees ‘on track', etc. A further ‘add on' is that researchers very quickly become very familiar with their own data and subsequently can become very adept at extracting and retrieving significant excerpts when the time comes, rendering their analysis perhaps that bit more sophisticated. Encountering their own earlier comments transformed into ‘data' also emphasized for participants the ways in which text can become disembodied from the context in which it was generated.

We worked on the basis that any 1-hour interview equates to 4 hours worth of transcribing⁴ and a further 20 hours of analysis.⁵ Although the decision had, on this occasion, been taken out of participants' hands, we discussed whether transcribing in full is a luxury, and looked briefly at possible alternatives, such as storing data on audio tapes and making only partial transcripts, taking brief notes or devising a pro-forma to summarize information.

Participants were then given the opportunity to carry out an initial analysis of the transcripts after introductory guidelines had been given in the form of a presentation at the beginning of the afternoon. Participants worked firstly in pairs to carry out initial coding, and then in a group to identify common themes and begin to aggregate findings. Between the first and second workshops (and again between the second and third workshops), the two facilitators, R Barbour and V Featherstone, carried out a more detailed analysis of the data generated. This was used to guide identification of common themes in the second workshop and formed the basis for writing up a section of the discussion document circulated to participants prior to the third workshop. This third workshop was convened, in addition to the two originally planned, in order to explore further some of the issues identified during the first two workshops.

To compute or not to compute?

Writing in 1995, Richards commented, "The debate about whether to compute in qualitative research seems to be over"⁶ and Lee and Fielding argue, "even, at a minimum, the ability effectively to manage data (provided by all of the available packages) may be a considerable improvement over the ad hoc procedures we suspect frequently underpinned manual analysis"⁷ (para. 2.3). Given the time constraints imposed by the workshop format, we opted to involve participants in manual coding/indexing of transcripts, whilst acknowledging that coding and data retrieval work could equally have been performed using any one of a number of available computer packages. (For an overview of a range of such packages, see Fielding,⁸ and advice is also available with regard to matching the software package with characteristics of the data set involved.^9,10 For a comparison of ATLAS/ti and NUD*IST, see Barry,¹¹ and for a detailed account of the application of NUD*IST, see Buston.¹²)

Manual coding certainly impressed upon workshop participants the practicalities and complexities of the task. We reminded them of the advice provided by an eminent anthropologist, "You'll need a bigger table than you think."¹³ To this, we would add "and a stock of coloured pens" (a strategy also alluded to by Coffey et al.¹⁴). Many experienced researchers in any case choose to combine computer and manual analysis—as acknowledged by Kelle,¹⁵ who talks about the possibility of "integrating paper and pencil work" (para. 2.9). This can be particularly useful for more complex analytical tasks.

Open coding/identifying broad themes

This exercise involved participants in carrying out what was essentially a content analysis using broad coding/ indexing categories. "Since all analysis is essentially comparative analysis, the purpose of indexing is simply to facilitate comparative analysis by gathering all data on a particular topic under one heading, in order to make the study of material manageable for analysis purposes."¹⁶ (p. 146). ‘Coding’ or ‘indexing’ thus anticipated the next stage of the process (see Aggregating data below).

Ritchie and Spencer¹ acknowledge that the first version of an index is often largely descriptive, drawing upon what they term "a priori issues": "those informed by the original research aims and introduced into interviews via the topic guide." (p. 180).

In contrast to pre-coded questionnaires or other more structured approaches, however, later versions of qualitative coding (or indexing) categories do not rely exclusively on "a priori issues". Coding categories developed as analysis unfolds will also reflect "emergent issues, raised by the respondents themselves, and analytical themes arising from the recurrence or patterning of particular views or experiences" ¹ (p. 180). Indeed, it is often only once analysis is well under way that the researcher identifies important themes which will guide the direction of the final report and/or published papers.

Kelle¹⁵ describes ‘open coding’^17,18 as the application of coding categories drawn either from researchers' own common sense knowledge or ‘in vivo’ codes, which focus on concepts or words used by interviewees and use these to order the data. Workshop participants had no difficulty with developing ‘common sense’ codes, applying the categories which one might well have anticipated in advance, given the topic being explored in the interviews, i.e. Ritchie and Spencer's "a priori issues". However, they found it much more difficult to create ‘in vivo’ codes, suggesting that there may be an important distinction to be made between the two procedures. In contrast to "a priori" codes, ‘in vivo’ codes involve the researcher in identifying and capitalizing upon the unexpected in his/her data in order to illuminate the phenomenon under study.

An attempt was made to illustrate the development of the two types of coding by carrying out a preliminary analysis of a transcript of part of the video interview produced for the first workshop. This was shown on an overhead projector and overlaid with coding symbols, which related to specific blocks of text. This allowed the analysis to be built up cumulatively, starting off with broad categories and moving on to identifying more specific ideas or use of particular words or phrases (Kelle's¹⁵ ‘in vivo’ codes).

Re-encountering the video interview rendered as a transcript, together with the experience of reading transcripts of their own ‘interviews' (both as ‘interviewer' and ‘interviewee'), impressed on participants the potential for important nuances to become lost in the process of transcribing. Particularly helpful here is Jenny Kitzinger's⁴ suggestion of reading over transcripts whilst listening to the original tape-recorded interview and making notes of these as you go, picking up voice intonations for joking or irony, noting laughter, and recalling relevant facial expressions or gestures, etc. in order to put remarks into perspective. Again, transcribing one's own tapes (or, indeed, frantically trying to recall interview interactions in the face of a failed tape-recorder) may provide surprises for researchers in terms of their ability to remember hitherto forgotten aspects of the interview.

After participants had carried out an initial coding exercise (in pairs working on the two transcripts provided) to identify ‘common sense’ categories, the following advice was given to them on further ‘open coding’ of transcripts.

• Focus on transcripts one by one at first, but keep an eye open throughout for issues which might have a bearing on the comparison across the 12 interviews
  
• Look out for distinctions made by the respondent
  
• Look out for qualifications/disclaimers made by the respondent
  
• Look for the respondent contradicting him/ herself
  
• Does the respondent give any indication of how she/he thinks her/his views compare with other people's views?
  
• Does the respondent offer any explanation/ analysis of the features of his/her own situation which colour his/her views?
  

As workshop organizers, we were mindful of the danger (identified by Coffey et al.¹⁴ in relation to using computer packages) of resorting to what they term a ‘coding fetish’ and forgetting that this procedure does not constitute ‘analysis’. This danger is also present—albeit to a more limited extent—in relation to manual coding, and we were, therefore, keen to pursue the next stage in the process.

Aggregating data

In addition to analysing individual interview transcripts, the workshop aimed to provide some guidance on aggregating analysis of the set of interviews conducted in the course of our notional research project. The next exercise to be tackled involved establishing the prevalence and patterns of occurrence of common themes, as well as identifying exceptions. With only 12 interviews, this was a relatively simple exercise, accomplished with the aid of a flipchart, but, where a large number of interviews is carried out, the researcher occasionally may opt to store some data in a form amenable to limited quantitative analysis¹⁹ and it may be useful to know that the computer package NUD*IST now has an interface to the quantitative data management programme SPSS.⁸

The two workshop facilitators had carried out some analysis on the set of 12 interviews prior to the second workshop, and oversaw the ‘aggregation of findings' exercise by requesting a show of hands to indicate the prevalence of particular themes, also encouraging dialogue between participants in order to tease out differences in content or emphasis. Workshop participants were urged to interrogate their own transcripts further, within the context of the composite picture being built up from discussion of all 12 transcripts, using the following set of questions.

• What are the most common themes?
  
• Are there any notable exceptions, i.e. a few individuals who do not discuss particular themes or who say very different things about particular topics?
  
• What are the range of views expressed with regard to a topic? Are views black and white or is there a spectrum?
  
• Can you identify any sub-categories, i.e. variations on your themes, further distinctions/qualifications?
  
• What concepts are appealed to?
  
• What language is used?
  
• What are the respondent characteristics associated with particular views?
  
• What views go together?
  
• What patterns emerge?
  

Even within this small sample of 12 health professionals' personal views, this preliminary exercise on aggregating data highlighted the ways in which differing levels of experience, areas of professional expertise and academic backgrounds brought out different issues in interviews. Whilst all of the doctors mentioned practical issues, such as the time involved in dealing with drug users, seven of the other nine ‘interviewees' stressed the importance of getting to the "root cause" (a term used by two people) of the problem, and two emphasized that it was not sufficient simply to treat the drug use or substitute a legal drug for an illegal one. Different use of language was often the key to subtle differences in perspectives. For example, only the two GPs and the health services manager spoke of "drug addicts". The one person who talked of the possibility of drug users being "cured" had no direct experience of working with this client group. (This is discussed further in the section on ‘Theorizing'.)

Given the time constraints of the workshops, the facilitators had resorted to the format beloved of children's TV presenters of producing a list of themes ‘we had identified earlier’. We explained that the coding on the interview transcripts had been done using coloured pens, which aids recognition of categories on subsequent trawls of the transcripts and allows for individual words or phrases to be highlighted with reference to particular codes. This mirrored the approach taken in analysing the section of transcript used in the earlier part of the workshop as an OHP transparency to illustrate the process.

Whilst there was considerable congruence between the coding categories developed by workshop participants and those listed by the facilitators, we had inevitably—given the longer time allocated to the task—taken this process further than was possible within the workshop session, and participants commented afterwards that they had found it difficult to link what was done in the workshop session with the issues and themes that the organizers pulled out at the end. The participants felt as if they had missed out on the coding and needed more practice at this.

Further analysis of transcripts by workshop organizers

Following the second workshop, R Barbour and V Featherstone carried out a more detailed analysis of the interview transcripts. We kept a tally of the time spent in this activity, to demonstrate to workshop participants the time-consuming nature of detailed qualitative analysis. Further analysis of the interviews, a summary of which is presented below, took up 3 hours individual work on the part of each of the two facilitators, and a further two 3-hour joint sessions. This allowed us to emphasize the importance of allowing an appropriate block of time for analysis when designing and costing research projects. Our further analysis produced 29 broad and subsidiary themes (listed below). All coding was done by hand on the transcripts, details were extracted and summarized on A5 cards, and one card was used for each theme or sub-theme.

The six broad categories and 23 subcategories identified were as follows:

1. Appropriateness of primary care provision
   • First port of call
     
   • No-one else to provide service
     
   • ‘Normalizing’
     
   • Importance of continuity
     
   
   
2. Is specialist or primary care provision most appropriate?
   • Inadequacy of secondary services
     
   
   
3. Organization of general practice
   • Special clinics (i.e. like family planning clinics)/ times/other doors/particular surgeries
     
   • GPs opting in or out; tensions caused by different responses to working with drug users
     
   • Strategies, e.g. control over prescribing, contracts, sanctions
     
   • Implications for workload/length/frequency of consultations
     
   • Need for training/knowledge about drug users
     
   • Need for staff support
     
   
   
4. Prioritizing
   • Weighing up needs of other patients, comparing with other patients
     
   • The impact on other patients/their views
     
   • Rights of drug users—self-induced problems
     
   • Potential dangers to other patients
     
   • Dangers to/safety of staff
     
   
   
5. Rewards/definitions of success/what can be achieved
   • Stabilizing
     
   • Working with other members of the family
     
   • Reduction in criminal activities
     
   • Relapse
     
   
   
6. Causes of drug addiction/wider social issues
   • Cycle
     
   • Identifying factors
     
   
   

On our A5 summary cards, we also reproduced a shortened account of views (together with the occasional useful verbatim quote). Bearing in mind the importance of not removing data from the context in which it was generated, we added a note to indicate whether comments were:

• made in response to a specific question on this particular topic (Q);
  
• put by the interviewer; provided in response to another related question (Q—with the question reproduced);
  
• provided by way of a delayed response to a direct question (earlier—Q);
  
• a delayed response to a related question put earlier (earlier Q and details of question); or
  
• volunteered comments, made spontaneously by the interviewee, either before the relevant question could be put or other unsolicited comments (Vol.).²⁰
  

Such information would be used in order to determine the weight attached to comments, with the greatest importance being accorded to volunteered statements. Prior to the third workshop, all 12 transcripts with analysis categories added were circulated to all participants, together with examples of two summary cards to illustrate the process, and an explanatory letter.

The ‘grounded theory' debate/rendering accessible the process of interpretation or theorizing

"Standardized, often mechanized procedures are no substitute for genuinely ‘grounded’ engagement with the data throughout the whole of the research process."¹⁴ (para. 7.6). The importance of coding may be over-emphasized "given that a large part of the qualitative researcher's work consists of interpretation and a fine-grained hermeneutic analysis."²¹ (pp. 48–49). Even the use of a sophisticated computer package does not "absolve the (researcher) from writing his or her theorized account."¹⁴ (para. 8.10).

Despite the prominence accorded it in many of the computer packages available²¹ and in textbooks,²² "rarely (in practice) is there a genuine interweaving of data collection and theorizing of the kind advocated by Glaser and Strauss²³ (1967)"²⁴ (who coined the term ‘grounded theory’) (p. 6) (the text in parentheses has been added). ‘Grounded theory’ is invoked with greater frequency than it is practised.^15,24

Rather than viewing qualitative researchers as embarking upon their studies in the "theoretical vacuum"²⁵ implied by a slavish attention to the model proposed by Glaser and Strauss,²³ we subscribed to the now widespread acceptance amongst qualitative researchers that our work should be capable both of being informed by theory and of generating theory.

The approach to interpreting data and theory-building which we advocated throughout the workshop reflected an eclectic but pragmatic approach, combining elements of ‘grounded theory’ and ‘analytic induction' with the occasional borrowing from ‘symbolic interactionism', ‘discourse analysis' or ‘ethnomethodology’ (see Melia,²² for a discussion of the influence of these related but distinct approaches). Frankland and Bloor¹⁶ describe the purpose of ‘analytic induction’ as being

"to derive propositions which apply generally across all the data to the entire universe of relevant cases or transcript items. Generalizability of the analytical propositions (‘theorizing’) is achieved by focusing on the ‘deviant cases’, those indexed items which appear to contradict the analytic proposition (i.e. the exceptions which we alluded to in our workshop checklist) . . . The analyst attempts to modify the analytic proposition to embrace the deviant case . . . Alternatively, the deviant cases may be eliminated by modifying the definition of the population to which the analytic proposition applies." (p. 150 text in parentheses added)

In talking about ‘theorizing’, we were appealing to the three rather different concepts, and, indeed, interpretive activities, outlined by Kelle:¹⁵

• "Grand theory" (accessed via literature immediately relevant to the topic in hand, and also via parallel literatures);
  
• "Theories of members of the investigated culture"¹⁵ (in our case this was co-terminous with the ‘researchers' doing the analysing); and
  
• "Theories developed by a sociological expert about a certain field of social action that have enough empirical content to be tested"¹⁵ (para. 4.8) and which go beyond accepting at face value the propositions advanced by the group under study.²⁶
  

Admittedly not all research concerns itself with all three types of theorizing. Much of the output of health services research, for example, has been criticized for its lack of attention to "grand theory", and many health services researchers would probably not view the third activity as falling within their scope. In the workshop, we briefly acknowledged the importance of reading around the topic of drug use, theoretical discussions of the different models of intervention (including looking at what texts used by the various professionals have to say on the topic) and explanatory frameworks regarding deviant behaviour more generally. We attempted to provide some insight into the processes involved in identifying ‘members' theories' and in encouraging participants to step back from these, emphasizing a more ‘relative’ stance through engaging in comparative analysis. Had this been a ‘real’ rather than a ‘virtual’ research project, this latter exercise could have led to the development of theories which could have been tested and/or explored further.

An analysis of the data generated with regard to what could be achieved in working with drug users suggested that ‘interviewees' were appealing to rather different models of intervention: whereas the two GPs who were involved in providing services for drug users talked about stabilizing drug use, and one of the researchers referred to stabilizing drug users' lives, several others saw the goal as being to get people off drugs, and, as mentioned earlier, one of the ‘interviewees' (who had no direct involvement with this client group) referred to the possibility of drug users being "cured". Interestingly, the comments from one of the GPs shows him ‘having things both ways', by alluding to the far off possibility of coming off drugs, whilst acknowledging the reality of drug users' situations and motivations. He said:

"Every one of them that comes to us says that they want to come off the drug. They're all lying. Most of them don't want to come off the drug at all—they want to stabilize themselves, and it may be that—once they're stable—they can start to think about coming off."

Although GPs—and perhaps others who deal directly with drug users—may abandon the goal of abstinence for that of harm minimization, they may not entirely give up hope; indeed, it is possible that they constantly re-negotiate this tension in their daily dealings with drug users. Such a ‘hypothesis' could be explored further within the context of a ‘real' study, which might also afford the opportunity to compare and contrast the views of service providers with varying lengths and degrees of exposure to drug users.

Reflections on our experience and implications for future workshops

On reflection, we might have benefited from organizing the analysis workshop somewhat differently. For example, having furnished them with our checklist of suggestions for a workable coding strategy, participants could have adjourned with their own transcripts of ‘interviews' to do some homework on coding prior to a second workshop on analysis. This might also have allowed them to explore the possibility of working in pairs in order to practise their new skills at leisure. Alternatively, we could have retained the single analysis workshop format, but could have sent out ‘interview' transcripts for participants to read in preparation for the workshop, focusing on coding. Another strategy might have been to have provided participants with a transcript of part of the video interview, giving examples of broad or ‘common sense' categories, ‘in vivo’ codes and how these might relate to ‘grand theory’, ‘members' theories' or ‘sociological expert theories'.¹⁵ This latter approach, however, might have bombarded participants with too much detail and could have made the whole process of analysis appear much more daunting.

Whilst the twelve 20-minute ‘interviews' illustrated just how much material can be collected in a relatively short space of time, they did not really provide enough ‘meat’ for ‘theorizing'. Perhaps the task of rendering accessible the processes involved in interpretation or ‘theorizing' would have been easier had we contrived—instead of formulating a fresh topic for our notional research project—to use ‘interview’ exercises which supplemented an actual project; where we had already collected data, coded and re-coded it, interpreted/ theorized and could, therefore, have placed the results of the workshop exercises in a wider context. We currently are pondering the possibility of holding further workshops on this same theme (or staging a larger workshop) which would also allow us to accumulate more data.

A perennial question in qualitative research is whether good researchers are ‘born’ or ‘made’ and whether skills are ‘caught’ or ‘taught’. This is particularly pertinent to the interpretation of data. We do not even have access to a clear vocabulary for talking about the processes involved in theorizing around data, and we found ourselves resorting to terms such as ‘lateral thinking’ "intuition-rich immersion"²⁷ and "crystallization"²⁸ —each of which can obscure as much as it can explain.

Melia²² has talked about the "near mystical passages" in Glaser's¹⁷ original book on ‘grounded theory', and there is an ever-present danger of falling into a similar trap in attempting to convey the steps involved in what is essentially a highly individualized and cerebral activity. In the process of working on these papers, we discovered that both facilitators had started out by studying English at their respective universities and pondered whether some of the skills involved in ‘de-constructing’ an interview transcript in the ways in which we were recommending were more akin in some respects to the requirements and tenets of literary criticism than to conventional health services research.

Atkinson²⁹ has been critical of the proliferation of methods textbooks and the resultant taxonomies produced by authors, warning of the dangers of "turn(ing) the pedagogical half truths of textbook knowledge into prescriptions for research practice" (p. 123). Whilst we recognize the potential for workshops such as these described here to add to this unfortunate trend, it is clearly unrealistic merely to direct novice researchers seeking to combine research with their roles as practitioners to the vast body of literature available. Again, we considered that the advice offered in the workshops, supplemented with appropriate cautions and selective reading lists, offered a compromise.

Acknowledgments

We gratefully acknowledge the assistance of Michelle Middleton (Adminstrator) and secretarial support from Liz Bradley, Maria (Winnie) Wilson and Abby Chambers.

References

¹ Ritchie J, Spencer L. Qualitative data analysis for applied policy research. In Bryman A, Burgess RG (eds). Analyzing Qualitative Research. London: Routledge, 1994: 173–194.

² Malterud K. Shared understanding of the qualitative research process: guidelines for the medical researcher. Fam Pract 1993; 10: 201–216.[Abstract/Free Full Text]

³ Marshall C, Rossman GB. Designing Qualitative Research. 2nd edn. Thousand Oaks/London: Sage Publications, 1995.

⁴ Kitzinger J, Barbour RS. Introduction: the challenge and promise of focus groups. In Barbour RS, Kitzinger J (eds). Developing Focus Group Research: Politics, Theory and Practice. London: Sage Publications, 1999: 1–20.

⁵ Murphy E. The potential of qualitative research in primary care. In Jones R, Kinmonth A-L (eds). Critical Reading for Primary Care. Oxford: Oxford University Press, 1992: 19–41.

⁶ Richards L. Transition work! Reflections on a three-year NUD*IST project. In Burgess RG (ed.). Studies in Qualitative Methodology: Computing and Qualitative Research. Volume 5. London: JAI Press, 1995: 105–139.

⁷ Lee RM, Fielding N. Qualitative data analysis: representations of a technology: a comment on Coffey, Holbrook and Atkinson. Sociol Res Online 1996; 1: <http://socresonline.org.uk/socresonline/1/4/1f.html>

⁸ Fielding N. Qualitative data analysis with a computer: recent developments. Social Res Update 1996; 1: <http://www.soc.surrey.ac.uk/sru/SRU1.html>

⁹ Stanley L,Temple B. Doing the business: using qualitative software packages in the analysis of qualitative datasets. In Burgess RG (ed.). Studies in Qualitative Methodology: Computing and Qualitative Research. Greenwich, CT: JAI Press, 1995: 169–193.

¹⁰ Weitzman EA, Miles,MB. Computer Programs for Qualitative Data Analysis. Thousand Oaks, CA: Sage Publications, 1995.

¹¹ Barry CA. Choosing qualitative data analysis software: ATLAS/ti and NUD*IST compared. Sociol Res Online 1998; 3: <http://www.socresonline.org.uk/socresonline/3/3/4.html>

¹² Buston K. NUD*IST in action: its use and its usefulness in a study of chronic illness in young people. Sociol Res Online 1997; 2: <http://www.socresonline.org.uk/socresonline/2/3/6.html>

¹³ Bowen ES. (pseudonym) Return to Laughter. Garden City, New York/London: Doubleday, 1964.

¹⁴ Coffey A, Holbrook B, Atkinson P. Qualitative data analysis: technologies and representations. Sociol Res Online 1996; 1: <http://www.socresonline.org.uk/socresonline/1/1/4.html>

¹⁵ Kelle U. Theory building in qualitative research and computer programs for the management of textual data. Sociol Res Online 1997; 2: <http://socresonline.org.uk/socresonline/2/2/1.html>

¹⁶ Frankland J, Bloor M. Some issues arising in the systematic analysis of focus group materials. In Barbour RS, Kitzinger J (eds). Developing Focus Group Research: Politics, Theory and Practice. London: Sage Publications, 1999; 144–155.

¹⁷ Glaser B. Theoretical Sensitivity: Advances in the Methodology of Grounded Theory Analysis. Mill Valley, CA: The Sociology Press, 1978.

¹⁸ Strauss A, Corbin J. Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Newbury Park, CA: Sage Publications, 1990.

¹⁹ Barbour RS. The case for combining qualitative and quantitative approaches in health services research. J Health Serv Res Policy 1999; 4: 39–43.[Medline]

²⁰ Becker H et al. Boys in White. Chicago: University of Chicago Press, 1961.

²¹ Lonkila M. Grounded theory as an emerging paradigm for computer-assisted qualitative data analysis. In Kelle U (ed.). Computer-Aided Qualitative Data Analysis: Theory, Methods and Practice. London: Sage Publications, 1995.

²² Melia KM. Producing ‘plausible stories': interviewing student nurses. In Miller G, Dingwall R (eds). Context and Method in Qualitative Research. London: Sage Publications, 1998: 26–36.

²³ Glaser B, Strauss A. The Discovery of Grounded Theory. Chicago: Aldine, 1967.

²⁴ Bryman A, Burgess, RG (eds). Analyzing Qualitative Data. London: Routledge, 1994.

²⁵ Mason J. Qualitative Researching. London: Sage Publications, 1996.

²⁶ Atkinson P. Narrative turn or blind alley? Qual Health Res 1997; 7: 325–344.[Abstract/Free Full Text]

²⁷ Miller WL, Crabtree BF. Clinical research. In Denzin NK, Lincoln YS (eds). Handbook of Qualitative Research. Thousand Oaks, CA: Sage Publications, 1998: 340–352.

²⁸ Richardson L. Writing: a method of inquiry. In Denzin NK, Lincoln YS (eds). Handbook of Qualitative Research. Thousand Oaks, CA: Sage Publications, 1998: 516–529.

²⁹ Atkinson P. Some perils of paradigms. Qual Health Res 1995; 5: 117–124.[Free Full Text]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				T. Thompson, R. Barbour, and L. Schwartz Adherence to advance directives in critical care decision making: vignette study BMJ, November 1, 2003; 327(7422): 1011. [Abstract] [Full Text] [PDF]

				T. D. Thompson, R. S Barbour, and L. Schwartz Health professionals' views on advance directives: a qualitative interdisciplinary study Palliative Medicine, July 1, 2003; 17(5): 403 - 409. [Abstract] [PDF]

				C. Jaye Doing qualitative research in general practice: methodological utility and engagement Fam. Pract., October 1, 2002; 19(5): 557 - 562. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) E-letters: Submit a response Alert me when this article is cited Alert me when E-letters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (8) Request Permissions Google Scholar Articles by Barbour, R. S Articles by WoReN, M. o. Search for Related Content PubMed PubMed Citation Articles by Barbour, R. S Articles by WoReN, M. o. Social Bookmarking          What's this?

Online ISSN 1460-2229 - Print ISSN 0263-2136

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics