Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Correspondence: WEBB C. & KEVERN J. (2001) Journal of Advanced Nursing 33(6), 798805
Christine Webb, Focus groups as a research method: a critique of some aspects of their use in nursing
44 Chandlers Walk, research
Exeter EX2 8BA,
Objective. To evaluate and critique reports in the nursing literature in the period
UK.
19901999 of the use of focus groups as a research method.
E-mail: cwebb@webbc.u-net.com
Methods. The articles were identied by a computerized search of the CINAHL
database and subjected to critical review.
Findings. The result of the search was that very few articles were found that
reported on a specic piece of research using the method. Methodological
discussions were more common and these were sometimes at a somewhat supercial
level without analysis or critique. The largest category of articles was concerned
with service development projects. The research-based articles were found to be
relatively unsophisticated in their use of the method, in particular in relation to data
analysis and social interaction within focus groups. Terms such as `content analysis'
and `grounded theory' were used in nonrigorous ways and incompatibility between
the underlying research approach and implementation of the method was identied
in the cases of phenomenology and grounded theory. Whilst selection of the focus
group method was often justied in terms of the benets that participant interaction
could yield, this interaction was rarely reported or discussed in the articles. One
author proposed a scheme for analysing this type of interaction, and this is
recommended to future researchers as a possible framework for interaction analysis.
Conclusions. The article concludes by calling for more in-depth consideration at the
research planning stages of the underlying assumptions of methodological approa-
ches that may be used to underpin focus group research and methods to be used to
analyse and report the data generated.
with the method and relying on secondary sources in nursing qualitative research method was not conrmed. There was
literature rather than citing the original developers of the indeed an increase in use of the method, but this had occurred
method or social science writings on the methodology. The more in relation to nursing education and management
most notable issue was the variety of data analysis methods in projects than to research studies. The next stage in the
use, some of which did not seem strictly appropriate to the attempt to understand how the method should be used in
method. Therefore it was decided to conduct a more system- research, and in particular at the data analysis stage, involved
atic analysis of the published material in relation to nursing detailed reading of the remaining 16 articles and following up
research. relevant references in the articles themselves. The rest of this
article is devoted to a consideration of the ndings from this
extended process.
Search strategy
2001 Blackwell Science Ltd, Journal of Advanced Nursing, 33(6), 798805 799
C. Webb and J. Kevern
800 2001 Blackwell Science Ltd, Journal of Advanced Nursing, 33(6), 798805
2 Methodological issues in nursing research Focus groups: a critique of nursing research1
calling them `grand tour questions'. Spradley's work is Krippendorff (1980) and Stewart and Shamdasani (1990).
grounded in an ethnosemantic and not a phenomenological `Latent' as well as `manifest' content may be identied using
approach (Spradley 1979). There is no discussion of phenom- this method, but this may also rely on quantication of data.
enology in the article and no reference to phenomenological Carey (1995, p. 492) points out that any attempt at counting
sources. This article therefore represents an example of the frequencies in this way is inappropriate `unless a question is
point made earlier about the adoption of a research approach directly asked or probed for across groups', and Robinson
without going back to primary sources, resulting in claims (1999, p. 909) agrees that `it is not appropriate to give
that do not stand up to any attempt at rigorous criticism. percentages'. In other words, quantication should only be
A confusion between methodology and methods or tech- used when all respondents (in the case of individual inter-
niques is also made by Nyamathi and Schuler (1990, p. 1283) views) or groups (in the case of focus groups) have been asked
when they write that focus groups are: the same question. Otherwise the results of counts are merely
serendipitous and `will generally not be meaningful' (Carey
a qualitative method which bears similarities to ethnography,
1995, p. 492).
grounded theory, phenomenology and participant observation.
Reports do not give details of whether focus groups were
This reveals a confusion between the concepts of methodo- conducted in the same way when more than one was used in a
logy and methods and reinforces the point about not relying particular study. Therefore it is plausible to assume that they
on secondary sources. This article by Nyamathi and Schuler were similarly conducted. Thus, as Robinson (1999, p. 909)
is frequently quoted as a principal source of information by points out, `each transcript must be looked at in the same
writers of the articles reviewed here without critical appraisal way, and all the transcripts must be used' so that analysis is
of its content (DiIorio et al. 1994, Macleod Clark et al. 1996, systematic. However, some reports claim to have used
McDaniel 1996, Stevens 1996, Reed & Payton 1997, Idvall `grounded theory' as a data analysis method and the next
& Rooke 1998, Sim 1998, Sloan 1998, Twinn 1998). section will consider the appropriateness of this.
Techniques of data analysis are seldom discussed in any detail According to Strauss and Corbin (1990, p. 23):
or at all in accounts of focus groups research (Carey 1995,
A grounded theory is one that is discovered, developed, and
Frankland & Bloor 1999) and this was also the case in the
provisionally veried through systematic data collection and analysis
nursing research focus groups identied earlier. Stevens
of data pertaining to that phenomenon. Therefore, data collection,
(1996, p. 172) claims that `any number of qualitative analysis
analysis, and theory stand in reciprocal relationship with each other.
strategies can be adapted to these purposes' and Sloan (1998,
p. 41) similarly states that `analysing data from focus groups The method is characterized by the use of theoretical
is essentially the same as analysing qualitative data from sampling, constant comparative data analysis, theoretical
other sources'. Carey (1995, p. 126) writes of analysis sensitivity, memo writing, identication of a core category
processes varying from the `less intense' (coding and categ- and the concept of `theoretical saturation'.
orizing) to the `more intense' (grounded theory). Apparently Four of the articles identied in the literature search for this
taking the same position, many authors write about using article claim to have used grounded theory as a method for
various techniques of cutting and pasting, both manual and data analysis. However, they do not report having used
computerized, in a process of coding, categorizing and concurrent data generation and analysis as would be required
8,9 identifying themes (e.g. Reiskin 1992, Fox 1993a, Fox in grounded theory. As the above criteria demonstrate, for
8,9 1993b, DiIorio et al. 1994, Idvall & Rooke 1998, Morrison
10,11 10,11 grounded theory method to be claimed, data should be
& Peoples 1999). Some writers recommend the use of subjected to constant comparative analysis as additional data
analysis grids drawn out on paper to aid organization of are collected. On the basis of the developing coding system
data and identication of themes (Miles & Huberman 1994, and increasing theoretical sensitivity developed by the resear-
Higginbottom 1998, Robinson 1999). cher as the analysis continues, theoretical sampling should be
Sometimes the analysis process is referred to as `content used to gather additional data in an attempt to corroborate or
analysis (McDaniel 1996, McFarland et al. 1998, Sim 1998), otherwise existing data and to elaborate on the categories and
but this term is more usually reserved for techniques used their properties and interlinkages. This process of iteration
to quantify qualitative data by counting frequencies of continues until a core category is identied and there is
occurrences of words, phrases or themes as discussed by theoretical saturation. That stage marks the end of data
2001 Blackwell Science Ltd, Journal of Advanced Nursing, 33(6), 798805 801
C. Webb and J. Kevern
collection because additional data is not adding to the data does not use the term explicitly. Instead she writes that
analysis. `purposively sampling within relevant categories will enhance
It is clear that, if several focus groups are conducted in a the likelihood of saturation of data' as will conducting `more
similar way and then the transcripts are subjected to a than one group for each salient segment (of the population)'
grounded theory type of analysis, this does not full the (p. 491). She also writes that:
criteria for grounded theory. Sim (1998) argues that data
Often it is useful for the guideline questions to be rened with
analysis of several groups should be done concurrently in
increased understanding of the topic as the study progresses.
order to determine when there is consensus on the issues
Although analysis across sessions will be somewhat limited by the
considered relevant by group members, and refers to this
lack of total comparability of data, the rened questions should
consensus as `saturation'. However, he does not discuss
provide more useful data (Carey 1995, p. 491).
grounded theory as such, and indeed writes that `analysis of
focus group data is likely to follow the same process as for This approach therefore closely parallels the grounded theory
other sources of qualitative data' and that `many of the method but does not use that term and does not speak of core
problems that arise in analysing focus group data are those category identication or theorizing that are basic features of
relating to qualitative data in general' (Sim 1998, p. 348). grounded theory.
Thus, he seems to miss the point that data collection methods Grounded theory was the approach chosen by Donovan
will vary in order to be consistent with a researcher's (1995) for a study of men during their partners' pregnancies.
underlying methodological approach. Also, he selectively This was the only nursing study traced which explicitly
adopts some but not all of the features of grounded theory described all stages of the method as being used in the
without explaining the reasons for this. research in an appropriate way. Men were recruited via a
An example of this claimed ground theory method being general practice and focus group data were collected with
used but only after data had been collected is provided by the them over a 6-month period. These data were supplemented
report of Goss (1998) who used ve focus groups to study by observations in antenatal classes conducted by midwives.
nurses working in women's health care and the topic of A research diary was also recorded. Donovan describes how:
violence against women. The report mainly discusses meth-
coding, analysis of data and theorizing occurred simultaneously
odological issues and very little detail is given of the
Logic diagrams and memos were written throughout the process to
substantive research project. However, Goss reports using
guide thinking and to record analytical insights and interpretations
`a four-step process of data analysis for focus group data'
the researcher employed the steps of reduction, selective sampling of
(p. 33) developed by Koniak-Grifn et al. (1994). Following
the literature, and selective sampling of the data for the purpose of
transcription and accuracy checking, step 2 `involves the
describing the analysis and presenting substantiation for the conclu-
techniques borrowed from grounded theory' of coding and
sions drawn (Donovan 1995, p. 712).
developing themes and categories. Step 3 uses `constant
comparative procedures, both within and among groups' and If a grounded theory method is to be used for data analysis,
this is continued `until saturation is reached'. Step 4 involves then, this must be built in to a study from its conception
validation using `an outside researcher familiar with the focus because data generation and analysis need to occur concur-
group method but uninvolved with the data collection' and rently. If this is not done, it is hard to see the justication for
`respondent validation' (p. 33). Thus it seems that all data calling an analytic method `grounded theory' rather than
were collected and the analysis was conducted afterwards `thematic analysis', which is what seems to have been used in
rather than concurrently. The method cannot then legitim- most of the articles traced for this review. Thematic analysis
ately be called `grounded theory', because the essence of the is described by Kitzinger and Barbour (1999, p. 16) as:
method is that it is `cumulative', as was discussed earlier and
Drawing together and comparing discussion of similar themes and
as Strauss and Corbin themselves explain:
examining how these related to the variation between individuals and
Each event sampled builds from and adds to previous data collection between groups.
and analysis. Moreover, sampling becomes more specic with time
because the analyst is directed by the evolving theory (Strauss &
Reporting focus group ndings: interaction
Corbin 1998, p. 203).
As the denitions of focus groups have shown, a key feature
12 Carey and Smith (1994) term grounded theory a `more
of the method is the exploitation of group interaction among
intense' method of analysing focus group data but Carey's
participants to stimulate the generation of data that might
later methodological article on the analysis of focus group
802 2001 Blackwell Science Ltd, Journal of Advanced Nursing, 33(6), 798805
2 Methodological issues in nursing research Focus groups: a critique of nursing research1
not otherwise emerge, for example in a series of single- Focus groups are different to other methods of data collection, and it
respondent interviews. This emphasis on interaction is `an is important to maintain a sense of the whole group within the
integral component that cannot be teased out' (Carey 1995, analysis (p. 150).
p. 488). Therefore, as Carey states:
This also presents a `challenge' in presenting data and this
An appropriate description of the nature of the group dynamics is should:
necessary to incorporate (sic) in analysis for example, heated
ideally contain examples of the discursive nature of the method, by
discussion, a dominant member, little agreement (p. 488).
using two or more participants in any quotations rather then (sic)
However, Kitzinger (1994, p. 104), in a review of over 40 presenting isolated excerpts from one individual (p. 151).
reports of focus group studies reports:
`Sequences of discussion' are given by Reed and Payton
I could not nd a single one concentrating on the conversation (1997) in their article discussing `issues of analysis and
between participants and very few that even included any quotations interpretation' in their study of the experiences of older
from more than one participant at a time. people moving into residential care. The sequences are used
to illustrate how consensus evolved as opinions were modi-
Similarly, there were very few examples in the articles
ed and developed through discussion. Using sequence
identied for this review in which there was any discussion
analysis of this kind also allowed the researchers to see
of interaction in the focus groups studied, and only one
how processes varied in their different groups. They also
(Twinn 1998) included details of the kind that Carey
comment on the dominance of one member in a group, where
suggests.
a manager dominated the discussion so much that it was
Twinn's (1998) article is a methodological discussion
more like an interview with one person. This raises questions
about using focus groups with Chinese populations and the
about the way the group was moderated and the skills of the
appropriateness of the method in this culture, as distinct from
moderator, which are discussed in a number of articles
the `Anglo-Celtic' groups with which they have previously
(Morgan 1995, Macleod Clark et al. 1996, Sim 1998,
been used. Twinn considers specically issues of `one member
Kitzinger & Barbour 1999).
dominance', the effect of `signicant group members',
Idvall and Rooke (1998) used `interaction quotations'
offering alternative opinions, the effect of the number of
(p. 518) to illustrate process aspects of their focus group
group members and the status and relationships of partici-
study of nursing care in surgical wards. This involved a
pants with the group leader. She gives quotations from the
coding system to identify individual speakers and the groups
data to support her conclusion that the project yielded rich
in which they participated. In addition an arrow was used
data, participants from that particular ethnic group were
between quotations to illustrate interactions. However,
willing to participate, and that therefore the method is
specic points identied from this interaction coding system
appropriate, with the proviso that groups and analysis are
are not discussed in the article and the conclusion drawn
conducted in the rst language of participants.
from this aspect is disappointing in view of the promise raised
In an article that uses examples of research into sexual
by the coding system:
health to illustrate the use of focus groups, Robinson advises
that: (It) can be considered an effective use of time to conduct focus group
interviews if you are interested in group rather than individual
The impact of the group dynamics and specic comments, jokes,
opinions (p. 518).
anecdotes, questions, censorship, changes of mind, deferring to the
opinion of others are all equally important (Robinson 1999, p. 909). The only specic guidance found in the articles traced for this
review was given by Stevens (1996). In an article reporting on
However, in the section of the article discussing her four
focus groups as a method for studying community health
examples she merely reports that the method was `encour-
issues, and in particular her own work with lesbians' health
aged interaction', was useful for studying `social process in
care experiences, she proposes the following questions as a
action' and that `reports of sexual activity were often
basis for analysing group interaction:
untruthful' (p. 911). Beyond this, however, no detail is given
about social interaction in the groups. How closely did the group adhere to the issues presented for
Macleod Clark et al's (1996) methodological discussion discussion?
article is based on their study of the perceptions of the
Why, how and when were related issues brought up?
philosophy and practice of nursing in Project 2000
programmes. They note that: What statements seemed to evoke conict?
2001 Blackwell Science Ltd, Journal of Advanced Nursing, 33(6), 798805 803
C. Webb and J. Kevern
What were the contradictions in the discussion? tions. The principal advantage claimed for focus groups,
namely the ability to use participant interaction to gain
What common experiences were expressed?
in-depth and rich data that would not be obtained through
Were alliances formed among group members? individual interviews, was rarely capitalized on. One author
proposed a list of questions to guide analysis of group
Was a particular member or viewpoint silenced?
processes and this is recommended to future researchers as a
Was a particular view dominant? possible aid to gaining the benets of this aspect of the
method.
How did the group resolve disagreements?
A more general conclusion is that nurse researchers should
What topics produced consensus? study in greater depth the methodological underpinnings of
their proposed methods and should discuss these fully in their
Whose interests were being represented in the group?
reports. Where the word limits of journals prohibit a full
How were emotions handled? discussion of methodology then a separate article may be
needed, and this should be based on consultation of primary
(Stevens 1996, p. 172)
sources. Whether articles are published in `academic' or
In the section of the article reporting on the `application of `professional' journals a rigorous approach to methodology is
focus group method' in her lesbian health project, a para- essential, not only to ensure the integrity of the research being
graph is specically devoted to describing interaction and reported but also because these articles will be consulted by
reporting jokes, disagreements, and a `growing consciousness nurses and students attempting to develop their research
of collective phenomena' (p. 173) as the discussion knowledge. Nursing research does not stand in isolation from
progressed. The groups also allowed members to gain research in other disciplines, particularly the social sciences.
`insights about their own experiences that they had not It is to be hoped that nursing scholarship will lead to
had before the discussion' (p. 174) and to share solutions to methodological learning and progress both for the discipline
problems of how to access the health care system. The focus of nursing and more widely, and not be conned to abstract
groups thus had a `consciousness raising' effect for partici- empiricist research of interest only to nurses. The social
pants. In further sections giving `selected ndings' Stevens sciences contribute to the knowledge base of nursing and this
gives more information on group processes in order to relationship should be reciprocal if nursing knowledge is to
contextualize quotations from the transcripts. Stevens, then, achieve higher status and wider recognition. This article is
goes further than any of the other nursing articles discussed intended to be a contribution towards this development.
both in giving explicit guidelines on analysing process aspects
of focus groups and also in actually discussing her data in
References
terms of group interaction. It would be useful for other
researchers to experiment with applying her scheme to their Asbury J. (1995) Overview of focus group research. Qualitative
own data and to modify and develop it on the basis of their Health Research 5, 414420.
Barbour R.S. & Kitzingen J. (eds) (1999) Developing Focus Group
own experiences.
Research. Politics, Theory and Practice. Sage, London.
Carey M.A. (1995) Comment: concerns in the analysis of focus group
data. Qualitative Health Research 5, 487495.
Summary and conclusions
Carey M.A. & Smith M.W. (1994) Capturing the group effect in
The CINAHL-based literature search that formed the basis focus groups: a special concern in analysis. Qualitative Health
for this discussion identied only a small number of articles Research 4, 123127.
Clarke A. (1999) Focus group interviews in health-care research.
reporting empirical research using focus groups in the period
Professional Nurse 14, 395397.
19901999. Scrutiny of these articles and discussions of the Davis-McFarland E., Trickey B., Reigart B., Shilling L., Wager K. &
method by nurse authors revealed disappointingly supercial West V. (1998) Using focus groups to identify lifestyle and health
and uncritical discussions and implementations of the issues in the elderly. Gerontology and Geriatrics Education 18,
method. In particular, methodological incompatibilities had 3350.
13 DiIorio C., Hockenberry-Eaton M., Maibach E. & Rivero T. (1994)
not been recognized between phenomenology and focus
Focus groups: an interview method for nursing research. Journal of
groups, and the requirements of a grounded theory approach Neuroscience Nursing 26, 175180.
to data generation and analysis had not been respected. It was Donovan J. (1995) The process of analysis during a grounded theory
common for authors to rely on secondary rather than primary study of men during their partners' pregnancies. Journal of
sources and this may help to account for the misinterpreta- Advanced Nursing 21, 708715.
804 2001 Blackwell Science Ltd, Journal of Advanced Nursing, 33(6), 798805
2 Methodological issues in nursing research Focus groups: a critique of nursing research1
Fox A. (1993a) Use of focus groups in program planning: the Tridec McDaniel R.W. (1996) Focus group research: the question of
experience. Beta Release 17, 1319. scientic rigor. Rehabilitation Nursing Research 5, 5359.
Fox D. (1993b) The focus group interview: a choice method of data Merton R.K., Kendall P. & Fiske P. (1956) The Focussed Interview.
14 collection. Michigan Nurse 66(9) 2930. Free Press. Glencoe, IL.
Frankland J. & Bloor M. (1999) some issues arising in the systematic Miles M. & Huberman A. (1994) An Expanded Sourcebook:
analysis of focus group materials. In Developing Focus Group Qualitative Data Analysis (2nd edn). Sage, Thousand Oaks, CA.
Research. Politics, Theory and Practice (Barbour R.S. & Kitzinger Morgan D.L. (1988) Focus Groups in Qualitative Research. Sage,
J. eds), Sage, London, pp. 144155. Newbury Park, CA.
Giorgi A. (1988) Validity and reliability from a phenomenological Morgan D.L. (1993) Successful Focus Groups. Sage, Newbury Park,
perspective. In Recent Trends in Theoretical Psychology (Baker CA.
W.J., Mos L.P., Rappard H.V. & Stam H.J. eds), Springer-Verlag, Morgan D.L. (1995) Why things (sometimes) go wrong in focus
New York, pp. 167176. groups. Qualitative Health Research 5, 516523.
Giorgi A. (1989) Some theoretical and practical issues regarding the Morse J.M. (ed.) (1991) Qualitative Nursing Research. a Contem-
psychological phenomenological method. Saybrook Review 7, porary Dialogue. Sage, Newbury Park, CA.
7185. Morrison R.S. & Peoples L. (1999) Using focus groups methodology
Giorgi A. (1997) The theory, practice and evaluation of the in nursing. The Journal of Continuing Education in Nursing 30,
phenomenological methods as a qualitative research procedure. 6265.
Journal of Phenomenological Psychology 28, 235260. Morse J.M. (ed.) (1994) Critical Issues in Qualitative Research
Goss G.L. (1998) Focus group interviews: a methodology for socially Methods. Sage, Newbury Park, CA.
sensitive research. Clinical Excellence for Nurse Practitioners 2, Nyamathi a. & Schuler P. (1990) Focus group interview; a research
3034. technique for informed nursing practice. Journal of Advanced
Gray-Vickrey P. (1993) Gerontological research. Use and application Nursing 15, 12811288.
15 of focus group. Journal of Gerontological Nursing 19(5), 2127. Reed H. & Payton V.R. (1997) Focus groups: issues of analysis and
Higginbottom G. (1998) Focus groups: their use in health promotion interpretation. Journal of Advanced Nursing 26, 765771.
research. Community Practitioner 71, 360363. Reiskin H. (1992) Focus groups: a useful technique for research and
Holloway I. & Wheeler S. (1996) Qualitative Research for Nurses. practice in nursing. Applied Nursing Research 5, 197201.
Blackwell Science, Oxford. Robinson N. (1999) the use of focus group methodology with
Idvall E. & Rooke L. (1998) Important aspects of nursing care in selected examples from sexual health research. Journal of
surgical ward as expressed by nurses. Journal of Advanced Nursing Advanced Nursing 29, 905913.
7, 512520. Sim J. (1998) Collecting and analysing qualitative data: issues
Kitzinger J. (1994) The methodology of focus group interviews: the raised by the focus group. Journal of Advanced Nursing 28,
importance of interaction between research participants. Sociology 345352.
of Health and Illness 16, 103121. Sloan G. (1998) Focus group interviews: dening clinical supervision.
Kitzinger J. (1995) Introducing focus groups. British Medical Journal Nursing Standard 12, 4043.
311, 299302. Spradley J.P. (1979) The Ethnographic Interview. Holt, Rinehart &
Kitzinger J. & Barbour R.S. (1999) Introduction: the challenge and Winston, New York.
promise of focus groups. In Developing Focus Group Research. Stevens P.E. (1996) Focus groups: collecting aggregate-level data to
Politics, Theory and Practice. (Barbour R.S. & Kitzinger J. eds), understand community health phenomena. Public Health Nursing
Sage, London, pp. 120. 13, 170176.
Koniak-Grifn D., Nyamathi R., Vasquez R. & Russo A. (1994) Risk Stewart D.W. & Shamdasani P.N. (1990) Focus Groups: Theory and
taking behaviors and AIDS knowledge: experiences and beliefs of Practice. Sage, Newbury Park, CA.
minority adolescent mothers. Health Education Research 9, 449 Strauss A. & Corbin J. (1990) Basics of Qualitative Research.
463. Grounded Theory Procedures and Techniques. Sage, Newbury
Kreuger R.A. (1994) Focus Groups: a Practical Guide for Applied Park, CA.
Research. Sage, Thousand Oaks, CA. Strauss A. & Corbin J. (1998) Basics of Qualitative Research.
Krippendorff K. (1980) Content Analysis: an Introduction to its Techniques and Procedures for Developing Grounded Theory.
Methodology. Sage, Newbury Park, CA. Sage, Thousand Oaks, CA.
Macleod Clark J., Maben J. & Jones K. (1996) The use of focus Twinn S. (1998) An analysis of the effectiveness of focus groups as a
group interviews in nursing research: issues and challenges. NT method of qualitative data collection with Chinese populations in
Research 1, 143153. nursing research. Journal of Advanced Nursing 28, 654661.
2001 Blackwell Science Ltd, Journal of Advanced Nursing, 33(6), 798805 805