Content uploaded by Paul Gill
Author content
All content in this area was uploaded by Paul Gill
Content may be subject to copyright.
© 2008 Nature Publishing Group
Methods of data collection
in qualitative research:
interviews and focus groups
P. Gill,1 K. Stewart,2 E. Treasure3 and B. Chadwick4
• Interviews and focus groups are the most
common methods of data collection used
in qualitative healthcare research
• Interviews can be used to explore
the views, experiences, beliefs and
motivations of individual participants
• Focus group use group dynamic s to
generate qualitative data
IN BRIEF
PRACTICE
This paper explores the most common methods of data collection used in qualitative research: interviews and
focus groups. The paper examines each method in detail, focusing on how they work in practice, when their use
is appropriate and what they can offer dentistry. Examples of empirical studies that have used interviews or focus
groups are also provided.
INTRODUCTION
Having explored t he nature and pur-
pose of qualitative research in the previ-
ous paper, this paper explores methods
of data collection used in qualitative
research. There are a variety of methods
of d ata c ollection i n qualitat iv e research,
including observations, textual or visual
analysis (eg from books or videos) and
interviews (individual or group).1 How-
ever, the most common methods used,
particularly in healthcare research, are
inter views and focus groups.2,3
1. Qualitative research in dentistry
2. Methods of data collection in qualitative
research: inter views and focus groups
3. Conducting qualitative interviews with
school children in dental research
4. Analysing and presenting qualitative data
QUALITATIVE RESEARCH
IN DENTISTRY
1* Senior Research Fellow, Faculty of Health, Sport
and Science, University of Glamorgan, Pontypridd,
CF37 1DL; 2Research Fellow, Academic Unit of Primary
Care, University of Bristol, Bristol, BS8 2AA, 3Dean
and Professor of Dental Public Health; 4Profes sor of
Paediatric Dentistry, School of Dentistry, Dental Health
and Biological Sciences, School of Dentistry, Cardiff
Universit y, Heath Park, Cardiff, CF14 4XY
*Correspondence to: Dr Paul Gill
Email: PWGill@glam.ac.uk
Refereed Paper
DOI: 10.1038/bdj.2008.192
©British Dental Journal 2008; 204: 291-295
The purpose of this paper is to explore
these two methods in more detail, in
particular how they work in practice,
the purpose of each, when their use is
appropriate and what they can offer
dental research.
QUALITATIVE RESEARCH
INTERVIEWS
There are t hree fundamental t ypes of
research inter views: structured, semi-
structured and unstructured. Structured
interviews are, essentially, verbally
administered questionnaires, in which
a list of predetermined questions are
asked, with little or no variation and
with no scope for follow-up questions to
responses that warrant fu rther elabora-
tion. Consequently, they are relatively
quick and easy to administer and may be
of par ticular use if clarification of cer-
tain questions are required or if there are
l i ke l y t o b e li t er ac y o r nu m e ra c y pr o bl e ms
with the respondents. However, by their
very nature, t hey only allow for limited
participant responses and are, therefore,
of litt le use if ‘depth’ is required.
Conversely, unstructured interviews
do not reflect any preconceived theories
or ideas and are performed with lit tle or
no organisation.4 Such an interview may
simply star t wit h an opening question
su ch a s ‘ Can yo u t ell me about your e x pe-
rience of visiting t he dentist?’ and will
then progress based, primarily, upon the
initial response. Unstructured interviews
are usually very time-consuming (often
lasting several hours) and can be diffi cult
to manage, and to par ticipate in, as the
lack of predetermined inter view ques-
tions provides little guidance on what
to talk about (which many participants
find confusing and unhelpful). Their use
is, therefore, generally only considered
where significant ‘depth’ is required, or
where v ir tually nothing is known about
the subject area (or a different perspec-
tive of a k nown subject area is required).
Semi-structured inter views consist of
several key questions that help to defi ne
the areas to be explored, but also allows
the interviewer or inter viewee to diverge
in order to pursue an idea or response
in more detail.2 This interview format
is used most frequently in healthcare,
as it provides participants with some
guidance on what to talk about, which
many find helpful. The fl exibility of
this approach, particularly compared to
structured interviews, also allows for
the discovery or elaboration of informa-
tion that is important to participants but
may not have previously been thought of
as per tinent by the research team.
For example, in a recent dental public
heath study,5 school children in Cardiff,
UK were interviewed about t heir food
choices and preferences. A key fi nd-
ing that emerged from semi-structured
inter views, which was not previously
thought to be as highly infl uential as the
data subsequently confi rmed, was the
BRITISH DENTAL JOURNAL VOLUME 204 NO. 6 MAR 22 2008 291
PRACTICE
significance of peer-pressure in infl uenc- between the two, such as the purpose of a one-off or, if change over time is of
ing children’s food choices and prefer- the encounter, reasons for par ticipating, interest, repeated basis,4 for example
292 BRITISH DENTAL JOURNAL VOLUME 204 NO. 6 MAR 22 2008
ences. This finding was also established
primarily through follow-up question-
ing (eg probing interesting responses
with follow-up questions, such as ‘Can
you tell me a bit more about that?’) and,
therefore, may not have emerged in the
same way, if at all, if asked as a prede-
termined question.
The purpose of research interviews
The purpose of the research interv iew
is to explore the views, experiences,
beliefs and /or motivations of individu-
als on specific matters (eg factors that
influence their attendance at the den-
tist). Qualitative methods, such as inter-
views, are believed to provide a ‘deeper’
understanding of social phenomena than
would be obtained from purely quanti-
tative methods, such as questionnaires.1
Interviews are, therefore, most appro-
priate where little is already known
about the study phenomenon or where
detailed insights are required from
individual participants. They are also
particularly appropriate for exploring
sensitive topics, where participants may
not want to talk about such issues in a
group environment.
Examples of dental studies that have
collected data using inter views are
‘Examining the psychosocial process
involved in regular dental attendance’6
and ‘Exploring factors governing den-
tists’ treatment philosophies’.7 Gibson et
al.6 provided an improved understanding
of factors that influenced people’s regu-
lar attendance with their dentist. The
study by Kay and Blin khorn7 provided
a detailed insight into factors that infl u-
enced GDPs’ decision making in relation
to treatment choices. The study found
that dentists’ clinical decisions about
treatments were not necessarily related
to pathology or treatment options, as
was perhaps initia lly t hought, but also
involved discussions with patients,
patients’ values and dentists’ feelings of
self esteem and conscience.
There are many similarities between
clinical encounters and research inter-
views , in that both employ similar i nte r-
personal skills, such as questioning,
conversing and listening. However, there
are also some fundamental differences
roles of the people involved and how the
inter view is conducted and recorded.8
The primary purpose of clinical
encounters is for the dentist to ask the
patient questions in order to acquire suf-
ficient information to inform decision
making and treatment options. However,
the constraints of most consultations
are such that any open-ended question-
ing needs to be brought to a conclusion
within a fairly short time.2 In contrast,
the fundamental purpose of the research
inter view is to listen attentively to what
respondents have to say, in order to
acquire more knowledge about the study
topic.9 Unlike the clinical encounter, it
is not to intentionally offer any form of
help or advice, which many researchers
have neither the training nor the time
for. Research interviewing therefore
requires a different approach and a dif-
ferent range of sk ills.
The interview
When designing an inter view schedule
it is imperative to ask questions that
are likely to yield as much information
about the study phenomenon as possible
and also be able to address the aims and
objectives of the research. In a qualita-
tive interview, good questions should
be open-ended (ie, require more than a
yes/no answer), neutral, sensitive and
understandable.2 It is usually best to
start with questions that participants
can answer easily and then proceed to
more difficult or sensitive topics.2 This
can help put respondents at ease, build
up confidence and rapport and often
generates rich data that subsequently
develops the interview f urther.
As in any research, it is often wise to
first pilot the interview schedu le on sev-
eral respondents prior to data collection
proper.8 T his allows the researc h team to
establish if the schedule is clear, under-
standable and capable of answering the
research quest ions, and if, therefore,
any changes to the interview schedule
are required.
The length of interviews varies
depending on the topic, researcher
and participant. However, on average,
healthcare inter views last 20-60 min-
utes. Inter views can be perfor med on
exploring the psychosocial impact of
oral trauma on participants and their
subsequent experiences of cosmetic den-
tal surgery.
Developing the interview
Before an inter view takes place,
respondents should be informed about
the st udy details and given assur-
ance about ethical principles, such as
anonymity and confi dentiality.2 This
gives respondents some idea of what to
expect from the interview, increases
the likelihood of honesty and is also a
fundamental aspect of the informed
consent process.
Wherever possible, interviews should
be conducted in areas free from distrac-
tions and at times and locations that are
most suitable for participants. For many
this may be at their own home in the
evenings. Whilst researchers may have
less control over the home environment,
familiarit y may help the respondent to
relax and result in a more productive
interview.9 Establishing rapport with
participants prior to the interview is
also important as this can also have a
positive effect on the subsequent devel-
opment of the interview.
When conducting the actual inter view
it is prudent for the inter viewer to famil-
iarise themselves with the interview
schedule, so that the process appear s
more natural and less rehearsed. How-
ever, to ensure that the interview is as
productive as possible, researchers must
possess a repertoire of skills and tech-
niques to ensure that comprehensive
and representative data are collected
during the interview.10 One of the most
important skills is the ability to listen
attentively to what is being said, so that
participants are able to recount their
experiences as fully as possible, without
unnecessary interruptions.
Othe r importa nt skil ls include adopt ing
open and emotionally neutral body lan-
guage, nodding, smiling, looking inter-
ested and making encouraging noises
(eg, ‘Mmmm’) during the interview.2 The
strategic use of silence, if used appro-
priately, can also be highly effective at
getting respondents to contemplate t heir
responses, talk more, elaborate or clar if y
© 2008 Nature Publishing Group
PRACTICE
particular issues. Other techniques that use in public sector marketing, such as researcher gives due consideration to the
can be used to develop the interview the assessment of the impact of health impact of group mix (eg, how the group
BRITISH DENTAL JOURNAL VOLUME 204 NO. 6 MAR 22 2008 293
further include refl ecting on remarks
made by participants (eg, ‘Pain?’) and
probing remarks (‘When you said you
were afraid of going to the dentist what
did you mean?’).9 Where appropriate, it
is also wise to seek clarifi cation from
respondents if it is unclear what they
mean. T he use of ‘leading’ or ‘loaded’
questions that may unduly infl uence
responses should always be avoided
(eg, ‘So you think dental surgery wait-
ing rooms are frightening?’ rather than
‘How do you find the waiting room at the
dentists?’).
At the end of the interv iew it is impor-
tant to thank participants for their time
and ask them if there is anything they
would like to add. This gives respond-
ents an opportunity to deal with issues
that they have thought about, or think
are important but have not been dealt
with by the interviewer.9 This can often
lead to the discovery of new, unantici-
pated information. Respondents should
also be debriefed about the study after
the interview has fi nished.
All interview s shou ld be tape r ec orded
and transcribed verbatim afterwards, as
this protects against bias and provides a
permanent record of what was and was
not said.8 It is often also helpf ul to make
‘field notes’ during and immediately
after each interview about obser vations,
thoughts and ideas about the inter-
view, as this can help in data analysis
process.4,8
FOCUS GROUPS
Focus groups share many common fea-
tures with less structured inter views, but
there is more to them than merely col-
lecting similar data from many partici-
pants at once. A focus group is a group
discussion on a par ticular topic organ-
ised for research purposes. This discus-
sion is guided, monitored and recorded
by a researcher (sometimes called a
moderator or facilitator).11,12
Focus groups were first used as a
research method in market research,
originating in the 1940s in the work of
the Bureau of Applied Social Research
at Columbia University. Eventually the
success of focus groups as a marketing
tool in the private sector resulted in its
education campaigns.13 However, focus
group techniques, as used in public
and private sectors, have diverged over
time. Therefore, in t his paper, we see k to
descr ibe focus groups as they are used in
academic research.
When focus groups are used
Focus groups are used for generating
infor mation on collective v iews, and the
meanings that lie behind those views.
They ar e a lso useful in generatin g a rich
understanding of participants’ experi-
ences and beliefs.12 Suggested criteria
for using focus groups include:13
• As a standalone method, for research
relating to group norms, meanings
and processes
• In a multi-method design, to explore
a topic or collect group language or
narratives to be used in later stages
• To clarify, extend, qualify or
challenge data collected through
other methods
• To feedback results to research
participants.
Morgan12 suggests that focus groups
should be avoided according to the fol-
lowing criteria:
• If listening to par ticipants’ views
generates expectations for the out-
come of the research that can not
be fulfi lled
• If par ticipants are uneasy with each
other, and will therefore not discuss
their feelings and opinions openly
• If the topic of interest to the
researcher is not a topic the
participants can or wish to discuss
• If statistical data is required.
Focus groups give depth and
insight, but cannot produce
useful numerical results.
Conducting focus groups:
group composition and size
The composition of a focus group needs
great care to get t he best quality of dis-
cussion. There is no ‘best’ solution to
group composition, and group mix will
always impact on the data, according to
things such as the mix of ages, sexes and
social professional statuses of the par-
ticipants. What is important is that the
may interact with each other) before the
focus group proceeds.14
Interaction is key to a successful focus
group. Sometimes this means a pre-
existing group interacts best for research
purposes, and sometimes stranger
groups. Pre-existing groups may be eas-
ier to recruit, have shared experiences
and enjoy a comfort and familiarity
which facilitates discussion or the abil-
ity to challenge each other comfortably.
In health settings, pre-existing groups
can overcome issues relating to disclo-
sure of potentially stigmatising status
which people may fi nd uncomfortable in
stranger groups (conversely there may
be situations where disclosure is more
comfortable in stranger groups). In other
research projects it may be decided that
stranger groups will be able to speak
more freely without fear of repercussion,
and challenges to other participants may
be more challenging and probing, lead-
ing to richer data.13
Group size is an important considera-
tion in focus group research. Stewart and
Shamdasani14 suggest that it is better to
slightly over-recr uit for a focus group
and potentially manage a slightly larger
group, than under-recruit and risk hav-
ing to cancel the session or having an
unsatisfactory discussion. They advise
that each group will probably have two
non-attenders. The optimum size for
a focus group is six to eight par tici-
pants (excluding researcher s), but focus
groups can work successfully with as
few as three and as many as 14 partici-
pants. Small groups risk limited discus-
sion occurr ing, while large groups can
be chaotic, hard to manage for the mod-
erator and frustrating for par ticipants
who feel they get insuffi cient opportu-
nities to speak.13
Preparing an interview schedule
Like research interviews, the interview
schedule for focus groups is often no
more structured than a loose schedule of
topics to be discussed. However, in pre-
paring an inter view schedule for focus
groups, Stewart and Shamdasani14 sug-
gest two general principles:
1. Questions should move from general
to more specifi c questions
© 2008 Nature Publishing Group
PRACTICE
2. Question order should be relative to differences of opinions to be discussed should be identified in a way that makes
importance of issues in the research fairly and, if required, encourage reti- it possible to follow the contributions
294 BRITISH DENTAL JOURNAL VOLUME 204 NO. 6 MAR 22 2008
agenda.
There can, however, be some confl ict
between these two principles, and trade
offs are often needed, although often
discussions will take on a life of their
own, which will influence or determine
the order in which issues are covered.
Usually, less than a dozen predeter-
mined questions are needed and, as with
research interviews, the researcher will
also probe and expand on issues accord-
ing to the discussion.
Moderating
Moderating a focus group looks easy
when done well, but requires a complex
set of skills, which are related to the fol-
lowing principles:15
• Participants have valuable views
and the ability to respond actively,
positively and respectfully. Such an
approach is not simply a courtesy, but
will encourage fru itf ul discussions
• Moderating without participating: a
moderator must guide a discussion
rather than join in with it. Express-
ing one’s own views tends to give
participants cues as to what to say
(introducing bias), rather than the
confidence to be open and honest
about their own views
• Be prepared for v iews that may be
unpalatably cr itical of a topic which
may be important to you
• It is important to recognise that
researchers’ individual characteristics
mean that no one person will always
be suitable to moderate any kind of
group. Sometimes the characteristics
that suit a moderator for one group
will inhibit discussion in another
• Be yourself. If the moderator is com-
fortable and natural, participants will
feel relaxed.
The moderator should facilitate group
discussion, keeping it focussed without
leading it. They should also be able to
prevent the discussion being domi-
nated by one member (for example, by
emphasising at the outset the impor-
tance of hearing a range of views),
ensure that all participants have
ample opportunity to contribute, allow
cent participants.13
Other relevant factors
The venue for a focus group is impor-
tant and should, ideally, be accessible,
comfortable, private, quiet and free from
distractions.13 However, while a central
location, such as the participants’ work-
place or school, may encourage attend-
ance, t he venue may affect participants’
behaviour. For example, in a school set-
ting, pupils may behave like pupils, and
in clinical settings, participants may
be affected by any anxieties that affect
them when they attend in a patient role.
Focus groups are usually recorded,
often observed (by a researcher other
than the moderator, whose role is to
obser ve the interaction of the group to
enhance analysis) and sometimes vide-
otaped. At the start of a focus group, a
moderator should ack nowledge the pres-
ence of the audio recording equipment,
assure participants of confi dential-
ity and give people the oppor tunit y to
withdraw if they are uncomfortable with
being taped.14
A good quality multi-directional
external microphone is recommended
for the recording of focus groups, as
inter nal microphones are rarely good
enough to c ope w i t h the var iat ion in v ol-
ume of different speakers.13 If observers
are pr ese nt, the y shou ld be i nt r oduce d to
participants as someone who is just there
to obser ve, and sit away from the dis-
cussion.14 Videotaping will require more
than one camera to capture the whole
group, as well as additional operational
personnel in the room. This is, therefore,
very obtrusive, which can affect the
spontaneity of the group and in a focus
group does not usually yield enough
additional information that could not be
captured by an observer to make vide-
otaping worthwhile.15
The systematic ana lysis of focus
group transcripts is crucial. However,
the transcription of focus groups is more
complex and time consuming than in
one-to-one interviews, and each hour
of audio can take up to eight hours to
transcribe and generate approximately
100 pages of text. Recordings should be
transcribed verbatim and also speakers
of each individual. Sometimes obser-
vational notes also need to be described
in the transcr ipts in order for them to
make sense.
The analysis of qualitative data is
explored in the final paper of this series.
However, it is important to note that the
analysis of focus group data is different
from other qualitative data because of
their interactive nature, and this needs
to be taken into consideration during
analysis. The importance of the con-
text of other speakers is essential to the
understanding of individual contribu-
tions.13 For example, in a group situa-
tion, participants will often challenge
each other and justify their remarks
because of the group setting, in a way
that perhaps they would not in a one-
to-one interview. The analysis of focus
group data must therefore take account
of the group dynamics that have gener-
ated remarks.
Focus groups in dental research
Focus groups are used increasingly in
dental research, on a diverse range of
topics,16 illuminating a number of areas
relating to patients, dental services and
the dental profession. Addressing a spe-
cial needs population diffi cult to access
and sample through quantitative meas-
ures, Robinson et al.17 used focus groups
to investigate the oral health-related
attitudes of drug user s, exploring the
priorities, understandings and barriers
to care they encounter. Newton et al.18
used focus groups to explore barriers to
services among minorit y ethnic groups,
highlighting for the fi rst time differ-
ences between minority ethnic groups.
Demonstrating the use of the method
with professional groups as subjects in
dental research, Gussy et al.19 explored
the barriers to and possible strategies for
de velopin g a sh ared approach in pre ven-
tion of caries among pre-schoolers. This
mixed method study was ver y important
as the qualitative element was able to
explain why t he clinical trial failed, and
this understanding may help research-
ers improve on the quantitative aspect
of future studies, as well as making a
valuable academic contribution in its
own right.
© 2008 Nature Publishing Group
PRACTICE
2. Britten N. Qualitative interviews in healthcare. 10. Hammersley M, Atkinson P. Ethnography: princi-
CONCLUSION In Pope C, Mays N (eds) Qualitative res earch in ples in practi ce. 2nd ed. London: Routledge, 1995.
Inter views and focus groups remain the health care. 2nd ed. pp 11-19. London: BMJ Books, 11. Kitzinger J. The methodology of focus groups:
1999. the importance of interaction between research
BRITISH DENTAL JOURNAL VOLUME 204 NO. 6 MAR 22 2008 295
most common methods of data collec-
tion in qualitative research, and are now
being used with increasing frequency in
dental research, particularly to access
areas not amendable to quantitative
methods and/or where depth, insight and
understanding of particular phenomena
are required. The examples of dental
studies that have employed these meth-
ods also help to demonstrate the range of
research contexts to which interv iew and
focus group research can make a useful
contribution. The continued employment
of these methods can further strengthen
many areas of dentally related work.
1. Silverman D. Doing qualitative re search. London:
Sage Publications, 2000.
3. Legard R, Keegan J, Ward K . In-depth inter views.
In Ritchie J, Lewis J (eds) Qualita tive research
practice : a guide for social scienc e students and
researchers. pp 139-169. London: Sage Publica-
tions, 2003.
4. May K M. Interview techniques in qualitative
research: concerns and challenges. In Mor se J
M (ed) Qualitative nursing re search. pp 187-201.
Newbury Park: Sage Publications, 1991.
5. Stewart K, Gill P, Treasure E, Chadwick B. Under-
standing about food among 6-11 year olds in South
Wales. F ood Culture Society 2006; 9: 317-333 .
6. Gibson B, Drenna J, Hanna S, Freeman R. An
exploratory qualitative study examining the social
and psychological processes involved in regular
dental attendance. J Public He alth Dent 2000;
60: 5-11.
7. Kay E J, Blinkhorn A S. A qualitative investigation
of factors governing dentists’ treatment philoso-
phies. Br Dent J 1996; 18 0: 171-176 .
8. Pontin D. Interviews. In Cormack D F S (ed) The
research pro cess in nursing. 4th ed. pp 289-298.
Oxford: Blackwell Science, 2000.
9. Kvale S. Inter views. Thousand Oaks: Sage Publica-
tions, 1996.
participants. Sociol Health Illn 1994; 16: 103-121.
12. Morgan D L. The focus group g uide book. London:
Sage Publications, 1998.
13. Bloor M, Frankland J, Thomas M, Robson K. Focus
groups in social research. London: Sage Publica-
tions, 2001.
14. Stewart D W, Shamdasani P M. Focus grou ps. Theory
and practice. London: Sage Publications, 1990.
15. Krueger R A. Moderating focus groups. London:
Sage Publications, 1998.
16. Chestnutt I G, Robson K F. Focus groups – what
are they? Dent Update 2002; 28: 189-192.
17. Robinson P G, Acquah S, Gibson B. Drug users:
oral health related attitudes and behaviours.
Br Dent J 2005; 198 : 219-224.
18. Newton J T, Thorogood N, Bhavnani V, Pitt J, Gib-
bons D E, Gelbier S. Barriers to the use of dental
services by individuals from minority ethnic com-
munities living in the United Kingdom: fi ndings
from focus groups. Primary Dent Care 2001;
8: 157-161.
19. Gussy M G, Waters E, Kilpatrick M. A qualitative
study exploring barriers to a model of shared care
for pre-school children’s oral health. Br Dent J
2006; 201: 165-170.
© 2008 Nature Publishing Group
