Professional Documents
Culture Documents
PRACTICE
qualitative research.
This paper provides a pragmatic approach
using a form of thematic content
P. Burnard,1 P. Gill,2 K. Stewart,3 E. Treasure4 and B. Chadwick5 analysis. Approaches to presenting
qualitative data are also discussed.
The process of qualitative data analysis
is labour intensive and time consuming.
Those who are unsure about this
approach should seek appropriate advice.
This paper provides a pragmatic approach to analysing qualitative data, using actual data from a qualitative dental public
health study for demonstration purposes. The paper also critically explores how computers can be used to facilitate this
process, the debate about the verification (validation) of qualitative analyses and how to write up and present qualitative
research studies.
simply manage the data and make han- Effects of sweets and chocolates
dling of them easier. 4. Single item consequences Effects of junk food
For example, computer packages can
Effects of fizzy drinks
Childrens perception of food However, researchers wishing to use such health, and perhaps even as a result of
Positive notions of food and their software should first undertake appro- participation in the study.15
consequences priate training and should be aware that Some respondents may also want to
Negative notions of food and their most programmes often do not abide by modify their opinions on re-presenta-
consequences normal MS Windows conventions (eg, tion of the data if they now feel that, on
Peer influence most interview transcripts have to be reflection, their original comments are
Copying converted from MS Word into rich text not socially desirable. There is also the
Healthy/unhealthy foods format before they can be imported into problem of how to present such informa-
Effects of sweets and chocolates the programme for analysis). tion to people who are likely to be non-
Effects of junk food academics. Furthermore, it is possible
Food choices in school Verification that some participants will not recognise
Diet in childhood The analysis of qualitative data does, of some of the emerging theories, as each
Food preferences course, involve interpreting the study of them will probably have contributed
Expected diet as a grown up findings. However, this process is argu- only a portion of the data.16
Food choices and preferences of ably more subjective than the process The process of peer review involves
friendship groups normally associated with quantitative at least one other suitably experienced
Effects of fi zzy drinks data analysis, since a common belief researcher independently reviewing
Perceptions of adult/child diets amongst social scientists is that a defi ni- and exploring interview transcripts,
The need to be healthy as an adult. tive, objective view of social reality does data analysis and emerging themes. It
not exist. For example, some quantita- has been argued that this process may
Once this second, shorter list of cate- tive researchers claim that qualitative help to guard against the potential for
gories has been compiled, the researcher accounts cannot be held straightfor- lone researcher bias and help to provide
goes a stage further and looks for over- wardly to represent the social world, additional insights into theme and the-
lapping or similar categories. Informed thus different researchers may interpret ory development.14,16,17 However, many
by the analytical and theoretical ideas the same data somewhat differently.4 researchers also feel that the value of
developed during the research, these cat- Consequently, this leads to the issue of this approach is questionable, since it is
egories are further refined and reduced the verifiability of qualitative data anal- possible that each researcher may inter-
in number by grouping them together.4 ysis. pret the data, or parts of it, differently.8
A list of several categories (perhaps up to There is, therefore, a debate as to Also, if both perspectives are grounded
a maximum of twelve) can then be com- whether qualitative researchers should in and supported by the data, is one
piled. If we consider the above example, have their analyses verified or validated interpretation necessarily stronger or
we might eventually come up with the by a third party.13,14 It has been argued more valid than the other?
reduced list shown in Table 2. that this process can make the analysis Unfortunately, despite perpetual
This reduced list forms the fi nal cat- more rigorous and reduce the element debate, there is no definitive answer to
egory system that can be used to divide bias. There are two key ways of hav- the issue of validity in qualitative analy-
up all of the interviews.12 The next stage ing data analyses validated by others: sis. However, to ensure that the analysis
is to allocate each of the categories its respondent validation (or member check) process is systematic and rigorous, the
own coloured marking pen and then returning to the study participants and whole corpus of collected data must be
each transcript is worked through and asking them to validate analyses and thoroughly analysed. Therefore, where
data that fit under a particular category peer review (or peer debrief, also referred appropriate, this should also include the
are marked with the according col- to as inter-rater reliability) whereby search for and identification of relevant
our. Finally, all of the sections of data, another qualitative researcher analyses deviant or contrary cases ie, fi nd-
under each of the categories (and thus the data independently.13-15 ings that are different or contrary to the
assigned a particular colour) are cut out Participant validation involves return- main findings, or are simply unique to
and pasted onto the A4 sheets. Subject ing to respondents and asking them to some or even just one respondent. Quali-
dividers can then be labelled with each carefully read through their interview tative researchers should also utilise a
category label and the corresponding transcripts and/or data analysis for process of constant comparison when
coloured snippets, on each of the pages, them to validate, or refute, the research- analysing data. This essentially involves
are filed in a lever arch file. What the ers interpretation of the data. Whilst reading and re-reading data to search
researcher has achieved is an organised this can arguably help to refi ne theme for and identify emerging themes in
dataset, filed in one folder. It is from this and theory development, the process is the constant search for understanding
folder that the report of the fi ndings can hugely time consuming and, if it does and the meaning of the data.18,19 Where
be written. not occur relatively soon after data col- appropriate, researchers should also pro-
As discussed earlier, computer pro- lection and analysis, participants may vide a detailed explication in published
grammes can be used to manage this have also changed their perceptions reports of how data was collected and
process and may be particularly useful in and views because of temporal effects analysed, as this helps the reader to crit-
qualitative studies with larger datasets. and potential changes in their situation, ically assess the value of the study.
It should also be noted that qualitative these supporting chapters would also undertaking this process for the fi rst
data cannot be usefully quantified given be used to develop theories or hypoth- time, we recommend seeking advice from
the nature, composition and size of the esise about the data and, if appropri- experienced qualitative researchers.
sample group, and ultimately the episte- ate, to make realistic conclusions and
1. Spencer L, Ritchie J, OConnor W. Analysis: prac-
mological aim of the methodology. recommendations for practice and tices, principles and processes. In Ritchie J, Lewis
further research. J (eds) Qualitative research practice. pp 199-218.
WRITING AND PRESENTING London: Sage Publications, 2004.
2. Lathlean J. Qualitative analysis. In Gerrish K, Lacy
QUALITATIVE RESEARCH Example b (combined findings A (eds) The research process in nursing. pp 417-
There are two main approaches to
and discussion chapter): 433. Oxford: Blackwell Science, 2006.
3. Williams C, Bower E J, Newton J T. Research in
writing up the fi ndings of qualitative Copying friends primary dental care part 6: data analysis. Br Dent J
2004; 197: 67-73.
research.20 The first is to simply report In this study, as with others (eg Lud- 4. Pope C, Ziebland S, Mays N. Analysing qualitative
key findings under each main theme or vigsen & Sharma21 and Watt & Shei- data. In Pope C, Mays N (eds) Qualitative research
in health care. 2nd ed. pp 75-88. London: BMJ
category, using appropriate verbatim ham22), peer influence is a strong factor, Books, 1999.
quotes to illustrate those fi ndings. This is with children copying each others food 5. Ritchie J, Spencer L, OConnor W. Carrying out
qualitative analysis. In Ritchie J, Lewis J (eds)
then accompanied by a linking, separate choices at school meal times: Qualitative research practice. pp 219-262. London:
discussion chapter in which the fi nd- Girl: They say copy me and what I Sage Publications, 2004.
6. Glaser B G, Strauss A L. The discovery of grounded
ings are discussed in relation to existing have. theory: strategies for qualitative research. Chicago:
research (as in quantitative studies). The Interviewer: And do you copy them if Aldine Publishing Company, 1967.
7. Stewart K, Gill P, Chadwick B, Treasure E. Qualita-
second is to do the same but to incor- they say that? tive research in dentistry. Br Dent J 2008; 204:
porate the discussion into the fi ndings Girl: Yes. 235-239.
8. Seale C. Analysing your data. In Silverman D (ed)
chapter. Below are brief examples of the Interviewer: Why do you copy them if Doing qualitative research. pp 154-174. London:
two approaches, using actual data from they say that? Sage Publications, 2000.
9. Morse J M, Field P. Nursing research: the applica-
a qualitative dental public health study Girl: Because they are my friends. tion of qualitative approaches. Cheltenham:
that explored primary school childrens (Girl, school 1, age 7). Stanley Thornes, 1996.
10. Stewart K, Gill P, Treasure E, Chadwick B. Under-
understanding of food.10 Children also identified friendship standing about food among 6-11 year olds in
groups according to the school meal type South Wales. Food Cult Soc 2006; 9: 317-333.
Example a (the they have. Children have been known to
11. Burnard P. A method of analysing interview tran-
scripts in qualitative research. Nurse Educ Today
traditional approach): have school dinners, or packed lunches 1991; 11: 461-466.
12. Burnard P. A pragmatic approach to qualita-
FINDINGS if their friends also have the same.21 tive data analysis. In Newell R, Burnard P (eds).
Contrasts and contradictions Research for evidence based practice. pp 97-107.
Oxford: Blackwell Publishing, 2006.
The interviews demonstrated that chil- If this approach was used, the com- 13. Mays N, Pope C. Rigour and qualitative research.
dren are able to operate contrasts and bined findings and discussion section BMJ 1995; 311: 109-112.
14. Barbour R S. Checklists for improving rigour in
contradictions about food effortlessly. would simply be followed by a conclud- qualitative research: a case of the tail wagging the
These contradictions are both sophisti- ing chapter. Further guidance on writ- dog? BMJ 2001; 322: 1115-1117.
15. Long T, Johnson M. Rigour, reliability and validity
cated and complex, incorporating posi- ing up qualitative reports can be found in qualitative research. Clin Eff Nurs 2000;
tive and negative notions relating to in the literature.20 4: 30-37.
16. Cutcliffe J R, McKenna H P. Establishing the cred-
food and its health and social conse- ibility of qualitative research findings: the plot
quences, which they are able to fluently CONCLUSION thickens. J Adv Nurs 1999; 30: 374-380.
17. Andrews M, Lyne P, Riley E. Validity in qualitative
adopt when talking about food: This paper has described a pragmatic health care research: an exploration of the impact
My mother says drink juice because its process of thematic content analysis as of individual researcher perspectives within col-
laborative enquiry. J Adv Nurs 1996; 23: 441-447.
healthy and she says if you dont drink it a method of analysing qualitative data 18. Silverman D. Doing qualitative research. London:
you wont get healthy and you wont have generated by interviews or focus groups. Sage Publications, 2000.
19. Polit D F, Beck C T. Essentials of nursing research:
any sweets and youll end up having to go Other approaches to analysis are avail- methods, appraisal, and utilization. 6th ed. Phila-
to hospital if you dont eat anything like able and are discussed in the literature.23- delphia: Lippincott Williams & Wilkins, 2006.
25 20. Burnard P. Writing a qualitative research report.
vegetables because youll get weak. (Girl, The method described here offers a Nurse Educ Today 2004; 24: 174-179.
school 3, age 11 years). method of generating categories under 21. Ludvigsen A, Sharma N. Burger boy and sporty girl;
children and young peoples attitudes towards food
which similar themes or categories can be in school. Barkingside: Barnardos, 2004.
If this approach was used, the fi ndings collated. The paper also briefly illustrates 22. Watt R G, Sheiham A. Towards an understanding
of young peoples conceptualisation of food and
chapter would subsequently be followed two different ways of presenting qualita- eating. Health Educ J 1997; 56: 340-349.
by a separate supporting discussion and tive reports, having analysed the data. 23. Bryman A, Burgess R (eds). Analysing qualitative
data. London: Routledge, 1993.
conclusion section in which the fi nd- This analysis process, when done 24. Miles M, Huberman A. Qualitative data analysis.
ings would be critically discussed and properly, is systematic and rigorous 2nd ed. Thousand Oaks: Sage Publications, 1994.
25. Silverman D. Interpreting qualitative data: methods
compared to the appropriate existing and therefore labour-intensive and time for analysing talk, text and interaction. 3rd ed.
research. As in quantitative research, consuming.4 Consequently, for those Thousand Oaks: Sage Publications, 2006.