Family Practice Vol. 19, No. 3, 285-289
© Oxford University Press 2002
Qualitative research and the problem of judgement: lessons from interviewing fellow professionals
School of Primary Care, University of Manchester, Rusholme Health Centre, Walmer Street, Manchester M14 5NP and
a Centre for Health Services Research, University of Newcastle upon Tyne, UK.
Chew-Graham CA, May CR and Perry MS. Qualitative research and the problem of judgement: lessons from interviewing fellow professionals. Family Practice 2002; 19: 285–289.
Received 25 June 2001; Revised 2 October 2001; Accepted 2 November 2001.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. Qualitative research methods are recognized increasingly as valuable tools for primary care research, and add an extra dimension to quantitative work.
Objective. The aim of this study was to illustrate the benefits and problems attending the dual role of clinician/qualitative researcher.
Methods. As part of two studies employing semi-structured interviews of GPs in a North-West conurbation, about the topics of consultations on chronic low back pain and drug misuse in primary care, respondents' views on their interaction with a GP researcher were explored.
Results. Access to the GP by the interviewing GP was easier when the GP researcher was known to the respondent. Such prior knowledge, however, may then influence the content of the data and the manner in which the GP researcher is perceived. During the interview itself, where respondents recognized the researcher as a clinician, interviews were broader in scope and provided richer and more personal accounts of attitudes and behaviour in clinical practice. The GP was also identified as an expert and judge, not just of clinical decision making but also about moral judgements made by GPs in their work. This will impact on the data obtained at interview and must be taken into consideration when the data are interpreted and analysed.
Conclusion. Qualitative research techniques increasingly are advocated as appropriate for research on and in general practice. The professional identity of the researcher plays an important part in constructing the kind of data obtained in such studies, and this must be made apparent in reporting and discussions of such qualitative work.
Keywords. General practice research, interviewing, qualitative research.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Qualitative research techniques can make a significant contribution to studies that seek to examine the experience of illness and the organization of primary care.1 Such methods increasingly are advocated for use in general practice research.2 The principal claim of qualitative researchers is that such techniques offer a means of understanding the authentic perceptions, sentiments and understandings of subjects in such studies.3 Such a view has immediate resonance with general practice, which devotes great efforts to understanding the patient in a holistic way. Most discussion about the employment of qualitative techniques in general practice research has operated from the position that they are to be used primarily in studies of patients or other service users.4,5 Little work has been done on those issues that relate to GPs interviewing their peers.6 To our knowledge, no previous study has addressed this question by asking respondents themselves about the research process.
The present paper explores the ways in which respondents in two studies, one of GPs' views on the consultation in cases of chronic low back pain, the second on the issue of consent in testing for hepatitis C, themselves conceptualized the GP as qualitative researcher.
Background
Qualitative research techniques of all kinds are widely used to understand processes and relationships in primary care. GPs themselves are involved increasingly as researchers in such studies, primarily using interview-based techniques. Such techniques are useful to practitioners in that they can address questions of immediate importance—such as those relating to the doctor–patient relationship—which are otherwise difficult to investigate. That interviewing involves "listening to people and becoming involved with their world"7 makes a strong case for the link with a speciality whose practitioners have this as a motivating force in their day-to-day work. Indeed, the principles of qualitative research reflect many of the concerns of family medicine.
It thus can be tempting to assume that interviewing skills are similar to those needed for clinical history taking. However, such skills are not necessarily transferable,8 and the goal of the general practice consultation is defined in the last instance from the clinician's perspective, while the semi-structured interview is intended to elicit the meanings attributed by the subject to a particular research question. In addition, the power relationships between doctor and patient are not the same as those of doctor–researcher and doctor–doctor.
As we have noted above, investigating the effect of the role of interviewer has tended to focus on patients,4,5 rather than other doctors. However, some authors have addressed specifically the issues involved in such research.6 The results of such debates are, inevitably, ambiguous. It has thus been argued that the results of such interviews by GPs may be more restricted and selective,9 or that they will be more fruitful and expansive.10 In the latter case, respondents might assume a greater degree of sympathy on the part of a medically qualified interviewer, and thus be more responsive and co-operative.11 Certainly interactions which seem to be specific to the researcher have been described.5 In a study of GPs' attitudes to homeopathic prescribing,12 where the interviewer was a final year medical student, for example, respondents adopted the role of either ‘colleague’ or ‘tutor’ but were highly responsive to detailed questions that could be construed as critical of their practice. The present studies offered an opportunity to explore the effects of ‘peer’ interviewing in qualitative research. We were particularly keen to investigate interviewees’ reactions to being interviewed, and we speculate on how this might affect the data obtained.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This paper presents results from two wider studies exploring the management of chronic low back pain (CLBP) in inner city general practice and drug misuse and consent to hepatitis C testing in primary care. Each was a qualitative study employing semi-structured interviews carried out by academic GPs (CC-G and MP). The first study explored the ways in which a group of inner city GPs (n = 22, interviewed by CC-G) understood the causes and presentations of CLBP, and responded to it within the consultation. The second study explored issues of consent around hepatitis C testing of drug misusers in primary care. In this study, MP interviewed 20 GPs.
The two study groups of GPs were obtained by purposive sampling from inner city practices in a North Western conurbation. In the first study, of 50 GPs who were invited to take part 22 agreed to do so.13 None of the respondents was known to the interviewer (CC-G). In the second study, the GP researcher (MP) approached GPs who were principals in practices who managed drug users in collaboration with three Greater Manchester drug services. In this second study, the response rate reached 50%. Here, particularly, many of the respondents knew of, or were acquainted with, the researcher (MP).
The interviews were audio-taped (with consent) and transcribed. The constant comparative method of analysis was employed,14 with emerging themes allowing modification of the interview schedules throughout the two studies. Transcripts formed raw data which were then subject to thematic analysis by CC-G and CM (study 1)13 and CC-G and MP (study 2). All respondents were invited to comment at the end of the interview on their experience of being interviewed by a peer, rather than a researcher from another discipline. The interviewer attempted to explore the problems that such an interview might have caused for the respondent, as well as the benefits that the respondent perceived by the interviewer being a fellow GP.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The GP as researcher
Although the interviewer (CC-G) in study 1 had gone to great lengths to explain that she was also a GP, a small number of respondents seemed to have forgotten this by the time the interview took place. One expressed surprise, saying, "are you? Oh, you don’t look like a GP" (GP10). In such interviews, the atmosphere tended towards formality, and it was clear that one respondent felt that he was being examined, saying, "I would like to ask you, have I answered correctly, are you happy with me?" (GP8). In these interviews, responses tended to be more explanatory, as if directed to a lay person, and these respondents were somewhat resistant to questions about attitudes. For example:
CC-G: "Another thing that patients say is that doctors don’t like treating patients with back pain."
GP11 (1): "I don’t mind. I don’t dislike it."
Here the GP respondents appeared to give little indication of their feelings about the consultations they were describing. In those interviews where the interviewing GP (MP) was known to the respondent, respondents did not treat him as a researcher/interviewer:
GP16 (2): "I appreciate it because it's challenging, and it raises some things that I haven't thought of before, I have a registrar and it's the same sort of thing. We have this kind of discussion."
The GP researcher as peer and confidant
For the most part, however, respondents asserted the benefits of being interviewed by a peer:
GP14 (1): "I think a research assistant may not understand, empathize or feel as GPs do. Knowing you were a GP gave us a shared understanding."
Similarly:
GP20 (1): "I think I’m up front to most people. I might have needed to elaborate a little more if you weren’t a GP. We have a shared understanding, I think, a shared language."
This kind of sentiment ran through many of the interviews in both studies. The idea of a ‘shared understanding’ is problematic for qualitative research, of course, precisely because such methods are intended to interrogate the tacit, taken for granted understandings that underpin everyday life. The GP as interviewer may thus not be able to interrogate the respondents effectively because of such a shared understanding.
The idea that the interviewer and respondent had a ‘shared understanding’ is important, but there is more than a shared understanding of technical questions about medicine at play here. In this context, both researcher and researched are bound together by a powerful set of common experiences and attributions of professional identity.15 The role of the researcher is to some extent ambiguous, and we have already noted that some respondents thought that they were being ‘tested’ on their technical knowledge. However, for the most part, respondents construed themselves, quite naturally as colleagues, engaged in a symbolic exchange with the researcher.
GP14 (1): "Having an academic interest, I realize co-operation is important. I think I am sympathetic to your needs as well—I mean, we do get a lot of requests as GPs for our time and we have to be selective . . . sensible things, I don’t mind doing."
The notion of ‘testing’ runs through accounts of ‘peer’ relationships too. In part, this is a matter of self-presentation, but it reflects the hierarchical relationships that run through the medical world. Many of the respondents, whilst at some stage acknowledging that the interviewer was a GP, also moved on to treat her as a confidante as the interview progressed. This involved sharing privileged information that reflected their perceived vulnerability to external adjudication.
GP3 (1): "I think if it were a specialist it would be more intimidating. I haven’t got enough confidence or experience to overcome the pupil–teacher relationship that I still feel with consultants—and feeling that (. . .) in those situations I would feel that my knowledge was being judged more than my attitudes. (. . .) Perhaps I would feel less intimidated by a non-medic, em, but I suspect in those situations I would probably fall back on making more knowledge-based statements, trying to impress the person!"
In such accounts, the generalist is thus perceived to be vulnerable to the assumed greater technical knowledge possessed by the specialist. However, social judgements about other kinds of vulnerability are also important here:
GP7 (1): "I’ve done two interviews before. One was about anorexia (. . .) and one was about stress in GPs. The first was a researcher who had had anorexia and she was quite antagonistic to me (. . .) There was some sort of transference from her . . . and the other was a researcher into the psychology of it. I think there is a kind of judgmental thing entering into it. I feel less threatened by someone who knows the problem, does the same job as me, understands the difficulties, doesn’t seem to judge and doesn’t have an axe to grind or make a point against GPs."
As we can see, one kind of vulnerability that is entered into here is about the construction of the researcher as an external threat. Researchers may be construed as being antagonistic or politically opposed to particular kinds of professional behaviour. The second kind of vulnerability that was expressed by respondents was more personal. It was about how they felt about their work. The same respondent offers a representative example of such an account.
GP7 (1): "I find it very difficult. Em, I think it's important for them to feel that they have continuing support and appreciation that they are in pain, and that serious causes—life threatening causes—have been excluded. [There are] stories of people, where they went to the doctor with some back pain and it turned out to be secondaries . . . really horrible stories . . . it's very difficult, I find the juggling act very difficult to do."
In another case:
GP9 (1): "It's getting me down. At home, yes, I take it home—deal with it with difficulty—look forward to retirement—can't wait to get out."Some accounts, like the one above, appeared to be to some extent cathartic in content. However, these occurred only where the respondent was aware that the interviewer was a GP. Outbursts of emotionally charged words were, in such cases, not hindered by political correctness or anticipation of what the interviewer was perceived to want to hear. Even so, respondents were aware that in giving an account on tape they were also exposing themselves to some degree of risk. One put it thus, "I just hope to God you’re not from the Daily Mail!"
The GP who was known as a clinician in the Region with expert knowledge in the area of the interview by the interviewers was rarely seen as a confidante:
GP20 (2): "It's a bit like taking a back seat at medical school, speaking to somebody who's obviously more knowledgeable."
The GP researcher as expert
All GPs in study 2 treated the interviewer as an expert in his field:
GP6 (2): "Do you know, are people tested in pregnancy?"
GP19 (2): "I don’t know if you’ve addressed that particularly?"
These GPs were typical in that they asked a number of questions of the GP researcher, in order to glean information from the ‘expert’ who was interviewing them.
Typical of the responses of the respondents when asked to comment on the process of the interview, and, in particular, being interviewed by a fellow GP:
GP5 (2): "No problem. It is interesting for me, always learning something."
The construction of data in an interview where the GP is perceived to be an expert may well differ from an interview where the interview is perceived to have no more technical knowledge in this area than the respondent.
The GP as judge
In both studies, some respondents seemed to be exploring whether the GP interviewer was making moral judgements on their consultations with the groups of patients under discussion:
GP20 (1): "We have a shared understanding . . ."
It is as if the respondent is saying "you know how it is, you’re surely not going to judge my actions."
GP13 (2): "I haven't actually approached it, I must say, it's dreadful, but there are so many other things that crowd in on you. Days just go by and you are worrying about your coronary heart disease audits, and your statin audits, and your . . ."
MP: "I’m not criticizing."
GP13 (2): "I must say, I must take more interest in the drugs people."
GP18 (1): "It is interesting to know if I’m doing the same as other GPs."
Respondents who feel they are being judged will be likely to be cautious in the conversation they have with any interviewer, but particularly a fellow professional. The data obtained may also need to be treated with caution.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This paper reports on a series of semi-structured interviews carried out by two GPs (CC-G and MP) in collaboration with a social scientist experienced in qualitative research methods (CM). The main focus of study 1 was on chronic low back pain as it was experienced by sufferers and encountered by clinicians. Throughout the study, it became increasing apparent that the identity that respondents attributed to the interviewer played an important part in forming the ‘data’ that were collected. Where the interviewer appeared to be identified as a non-clinical researcher, the interview was narrower in focus, with less discussion and diversion, and much less emotionally charged. Where respondents recognized the interviewer as a clinician, they shifted between treating her as a professional peer and a private confidante. In the former state, accounts were formulated as ‘public’ representations of attitudes and activities, intended to be open to scrutiny. In the latter, respondents permitted themselves a degree of vulnerability that itself emphasized their perception that in ‘professional’ company they were safe to do so. Such encounters, which may only be of a few minutes duration in the midst of what is otherwise constituted as a ‘communication between equals’, led to rich and intuitive responses, in contrast to the brief and more calculating responses of those respondents who defined the interviewer as a non-clinician. The risk in such a relationship, however, is that it may create a case of shared conceptual blindness, allowing the interviewer's own feelings and opinions about the field to govern the dialogue and interpretation.16 In study 2, the GP interviewer, because of his widely known interest and expertise in the area under study, was perceived by all respondents as an expert and questioned closely by the respondent GPs—a role which he found difficult to disentangle from the interview process. There was some indication of the sharing of sentiment and feeling, but only with caution, since the GP interviewer was also perceived to be acting as a judge of his or her work. This may be because the topic under scrutiny in study 2 was narrow compared with CLBP in study 1. The GP interviewer in study 1, however, was also seen by some respondents to be a judge of their work and their attitudes; thus the perception that the known interviewer (MP) was more of an expert in his field than CC-G must be important.
It must, however, be remembered that the style of the interviewer will also be important17 and may set the tone of the interview. MP and CC-G spent some time, in supervision, discussing their interviewing styles, analysing the transcripts and assessing how their individual style may have influenced the data collected.18 CC-G, whilst more experienced as a qualitative researcher, did consider herself to be less ‘expert’ than MP around the clinical topic. How this impacted on the construction of the interview and the data obtained is not known, but we can speculate that the data obtained would be influenced by this self-perception of the researcher.
Understanding that the interviewer may play one or more roles in an interview is important in analysing as well as collecting qualitative data. Because qualitative research is intended to give priority to the meanings and attributions that respondents bring to bear on a question, it is important to see the interviewer as someone who is actively involved in constructing those meanings, rather than as someone who is present to ‘collect’ them passively. It is also important to understand that research conducted in such settings is, increasingly, not professionally or politically neutral. Respondents in such studies implicitly understand that there is an element of assessment or adjudication of quality in studies that seek to explore their clinical practice. In this context, resistance to research—either through presenting partial accounts, or through being selectively silent about particular topics—is not uncommon in such work. This resistance should not be understood in terms of non-compliance, but realistically. Respondents rightly seek to protect themselves from external adjudication, especially in an environment where one product of research is understood to be the production of ever greater demands.
In any qualitative study, the researcher's theoretical outlook, interests and expectations inevitably influence all aspects of the research process. Some authors have argued that this is a particular issue for GPs undertaking qualitative research,19 although we would argue that this is not unique to qualitative research and applies equally to other modes of practice. Preconceptions do not inherently imply ‘bias’, however, as long as researchers clarify them to themselves and to readers.19 The challenge for the GP researcher is to bracket their preconceptions sufficiently to accommodate the voice and experience of their informants. Here, much of the existing literature on this topic reflects the notion that researcher and researched normally belong to distinct groups. This is especially the case where the problem of the practitioner-researcher is set up in relation to work with patients. Our argument is that when researcher and researched ‘belong’ to the same kind of professional group, i.e. general practice, we can see the way in which data are formed to meet quite specific ends: in this context, obtaining a purchase on the preconceptions and ‘bias’ held by the respondent is itself crucial to understanding what ‘data’ are, and what they can be held to mean.
|
Acknowledgments |
|---|
We are grateful to the GPs who took part in these studies for their time and candour. We are also grateful to two anonymous readers who gave us valuable comments on an earlier report.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Kinmouth AL. Editorial: Understanding meaning in research and practice. Fam Pract 1995; 12: 1–2.
2
Murphy E, Mattson B. Qualitative research and family practice: a marriage made in heaven. Fam Pract 1992; 9: 85–90.
3 May C. The preparation and analysis of qualitative interview data. In Roe B, Webb C (eds). Research and Development in Nursing. London: Whurr Scientific Press, 1998: 59–83.
4
Hoddinott P, Pill R. Qualitative research interviewing by general practitioners: a personal view of the opportunities and pitfalls. Fam Pract 1997; 14: 307–312.
5
Richards H, Emslie C. The ‘doctor’ or the ‘girl from the University’? Considering the influence of professional roles on qualitative interviewing. Fam Pract 2000; 17: 71–75.
6
Andersson SJ, Troein M, Lindberg G. Conceptions of depressive disorder and its treatment among 17 Swedish GPs. A qualitative interview study. Fam Pract 2001; 18: 64–70.
7
Britten N. Qualitative interviews in medical research. Br Med J 1995; 311: 251–253.
8 Britten N, Jones R, Murphy E, Stacy R. Qualitative research methods in general practice and primary care. Fam Pract 1995; 12: 42–45.
9 Bradley CP. Factors which influence the decision whether or not to prescribe: the dilemma facing general practitioners. Br J Gen Pract 1992; 42: 454–458.[ISI][Medline]
10
Touw-Otten F, Spreeuwenberg C. Multi-disciplinary research in natural and social sciences in general practice. Fam Pract 1985; 2: 42–45.
11
Myerson S. Investigating stresses in general practice using an open ended approach in interviews. Fam Pract 1990; 7: 91–95.
12 May C, Sirur D. Art, science and placebo: incorporating homeopathy in general practice. Sociol Health Illness 1998; 20: 168–190.
13
Chew-Graham CA, May CR. ‘Partners in pain’—the game of painmanship revisited. Fam Pract 2000; 17: 285–287.
14 Strauss A. Qualitative Analysis for Social Scientists. Cambridge: Cambridge University Press, 1987.
15 Good BJ. Medicine, Rationality and Experience. An Anthropological Perspective. Cambridge: Cambridge University Press, 1994.
16
Hamberg K et al. Scientific rigour in qualitative research—examples from a study of women's heath in family practice. Fam Pract 1994; 11: 176–181.
17 Holstein JA, Gubrium JF. The Active Interview. Qualitative Research Methods, Vol. 37. Thousands Oaks: Sage Publications, 1995.
18
Barry CA, Britten N, Barber N, Bradley C, Stevenson F. Using reflexivity to optimise teamwork in qualitative research. Qual Health Res 1999; 9: 26–44.
19
Malterud K. Shared understandings of the qualitative research process. Guidelines for the medical researcher. Fam Pract 1993; 10: 201–206.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. M. Jensen, K. Trollope-Kumar, H. Waters, and J. Everson Building physician resilience Can Fam Physician, May 1, 2008; 54(5): 722 - 729. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Kjeldmand and I. Holmstrom Balint Groups as a Means to Increase Job Satisfaction and Prevent Burnout Among General Practitioners Ann. Fam. Med, March 1, 2008; 6(2): 138 - 145. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Hewitt Ethical Components of Researcher Researched Relationships in Qualitative Interviewing Qual Health Res, October 1, 2007; 17(8): 1149 - 1159. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Ridd, A. Shaw, and C. Salisbury 'Two sides of the coin'--the value of personal continuity to GPs: a qualitative interview study Fam. Pract., August 1, 2006; 23(4): 461 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D Jeavons, A P S Hungin, and C S Cornford Patients with poorly controlled diabetes in primary care: healthcare clinicians' beliefs and attitudes. Postgrad. Med. J., May 1, 2006; 82(967): 347 - 350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V Arora, J Johnson, D Lovinger, H J Humphrey, and D O Meltzer Communication failures in patient sign-out and suggestions for improvement: a critical incident analysis Qual. Saf. Health Care, December 1, 2005; 14(6): 401 - 407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kharicha, S. Iliffe, E. Levin, B. Davey, and C. Fleming Tearing down the Berlin wall: social workers' perspectives on joint working with general practice Fam. Pract., August 1, 2005; 22(4): 399 - 405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Hussey, P. Hoddinott, P. Wilson, J. Dowell, and R. Barbour Sickness certification system in the United Kingdom: qualitative study of views of general practitioners in Scotland BMJ, January 10, 2004; 328(7431): 88. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Stokes, M. Dixon-Woods, and R. K McKinley Breaking up is never easy: GPs' accounts of removing patients from their lists Fam. Pract., December 1, 2003; 20(6): 628 - 634. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Perry and C. Chew-Graham Finding a vein or obtaining consent: a qualitative study of hepatitis C testing in GP methadone clinics Fam. Pract., October 1, 2003; 20(5): 538 - 544. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||