THE USE OF FOCUS GROUPS (IM SOCIAL AIMD BEHAVIOURAL RESEARCH SOME METHODOLOGICAL ISSUES

Item

Title
THE USE OF FOCUS GROUPS (IM SOCIAL AIMD BEHAVIOURAL RESEARCH SOME METHODOLOGICAL ISSUES
extracted text
I
145 —

or controls.
qy, 19 (1).

' I,

THE USb OF FOCUS GROUPS (IM SOCIAL AIMD BEHAVIOURAL RESEARCHSOME METHODOLOGICAL ISSUES

ions for th'
jng

SDA-RF-CH-1B.1

youih

Mohammed Ejazuddin Khan, Martha Anker, ' Bella C Patel?
Sandhya Barge, ’ Hemlatha Sadhwani ■ & Hanjana Kahle "

nunizatioi
lion, 63

I
itions dan
adies din;
ation arm
s rotaviru'.
ion diale dt -

During recent years, the importance of qualitative
approaches in understanding social realities has
been increasingly recognized by social scientists as
well as by programme managers. Many researchers
have started questioning the adequacy of an ex­
clusively
quantitative
approach
in
explaining
changes in the social and demographic situation.
Among the various qualitative methods, "focus
group discussion" has become very popular and is
being extensively used in social and behavioural re­
search.
le focus group is an established method
in markt
esearch, its use in social science, demo
graphy or other related disciplines is rather new.

The focus-group session should be held in a natural
setting and be conducted in a relaxed manner. The
full discussion is tape-recorded. Apart from the par­
ticipants and moderator, a note taker also sits in the
session but does not participate in the discussion.
The note taker knows about the objectives and sub­
ject of inquiry, and is expected to be well trained in
observing and noting nonverbal group feedback,
such as facial expressions. Later the note taker also
transcribes the complete discussion based on notes
and tapes. These transcripts then serve as basic data
for analysis.

Following a brief description of the focus-group
•r- methodology, this article outlines the potential use
of focus-group discussions, their strengths and
weaknesses as well as methodological issues that
------- still need to be investigated in order to make use of
the full potentia1 T this method. The conclusions call
for more mves’
tion into the factors that influence
the outcome or u focus-group discussion, and warn
against using focus groups as a stand-alone, rapid
assessment method.

It is expected that the informal homogeneous group
setting, and the open-ended nature of questions, will
encourage the participants to feel free from various
constraints to which they are subject during indi­
vidual interviews. Thus it is believed that they ex­
press their views openly and spontaneously. The
moderator helps the participants to interact and this
interaction stimulates memories and
clings and
thus leads to a full in-depth discussion of the topic
at hand. These group dynamics distinguish focusgroup sessions from individual in-depth interviews
typical of ethnographic research?

udies: d< ■
Oxford Un:
...

iple size ,
■;v

& S'lns

dy. Journal
tour recall
laires '■nd
:es: a
n
Health Or
etrcspech
les co
r i; ques
s i le pkir
ie|. Bullen

me, 65 (2-

vVhat is a focns group?

ol study '
?n on die.'
'al of eo

A focus-group session is an in-depth discussion in
which a small number of people (usually 8-12) from
the target population, under the guidance of a fa­
cilitator (moderator) discuss topics of importance for
a particular study/project. It is basically a qualitative
method in which the moderator, with the help of
predetermined guidelines, stimulates free discussion
among the participants on the subject of inquiry.
The order in which the topics are covered is flexible,
but generally the discussion starts with more gen­
eral issues and slowly flows into more specific ones.
At th'- end, a few probing questions are sometimes
aske- :o reveal more in-depth information or to
clarify earlier statements or responses.

e (1): 83-

ties of •
1 diarrhons of ii<mai jour/:.i

I
I

Generally the p... ticipants are chosen purposively
and it is recommended that they hould be homo­
geneous with respect to cha
teristics which
mic;hi otherwise impede the free flow of discussion.
It is also considered desirable that the participants
should not know each other or the subject of the dis­
cussion in advance.

‘Associate. Population Council, Baroda, India.
'Statistician. Division of Epiderniolngical Surveillance and Hpairh
Situation and Trend Assessment. World Health Organization
Geneva.
c• Research
nesearcn Executive, Centre for Operations ReseaiL.'.
Reseaic!. o-.J Training
Baroda, India.

f

^Project Manager, Centre for Operations Research and Training
Baroda, India.

.4

mond.. 44 < 19911

• Research Associate, Operations Research Group, Baroda, India.
’ Knodel, J. & Pramualratana, A. Focns-group research as a means
of demographic inquiry. Population Studies Center, University of
Michigan, April 1987. (No. 87-106).

v- Wld hlth statist, quart, 44 (1991)

How can focus groups be used?
Available literature shows that the focus-group ap­
proach, like some other qualitative methods, could
effectively be used as follows.
•

As an idea-generation tool

Focus-group discussions could, for example, be used
by a health programme to find out what motivates
people to use a specific health product or health­
service facility, : to adopt better health-related prac­
tices. Such background information can be critical to
health planners who need to know how the popula­
tion views various health issues. In addition, focusgroup discussions with health-care providers can be
useful in pinpointing problems and in generating
ideas for improvem -nts in services.

•

In conjunction with a quantitative study

Focus-group discussions are often used as a com­
plement to a quantitative studv. helping to answer
such questions as "why?" or "Ituw?', rather than
how many?". They can also be used as a pre­
liminary step, providing background information,
and to generate hypotheses for field-testing. Tney
can also be used to refine a questionnaire, and to
ensure that the words and concepts correspond to
those commonly used by the target group.
Focus groups have also been used as a follow-up to
a quantitative study, to explain, expand and illum-

— 146 —
4
I

,

inate quantitative data, in order to gain some under­
standing about the reasons for certain findings.
In short, focus-group discussions, when used along­
side quantitative studies, can result in a much
greater understanding than either method used
alone.

•

As a primary data-collection method

Focus-group discussions can be used as a primary
data-collection method for some topics that cannot
easily be studied through quantitative methods
bo-us-group discussions are particularly suited to
subjects that are of a sensitive or personal naturefor example, Suyono et al. (1) covered abortion, and
KowaleskM(2) covered sexual behaviour among gay
men. Neither of the studies reported problems in
discussing these rather sensitive topics. In fact, the
former study in Indonesia found that participants
were much more willing to discuss abortion in the
tocus-group discussion than they were in survey in­
terviews. Group discussions suggested high aware­
ness r.
abortion and different techniques for
abortion, whereas sample-survey results indicated
low awareness of abortion. The researchers conC U; u Ju
SUrvey interviews, which are usually
watched by neighbours, are probably much less con­
ducive to eliciting information about sensitive topics
than are focus-group discussions which are away
trom home, among anonymous participants and in a
supportive setting.
However, care must be taken to treat the results of
ocus-group interviews with some caution, since
they can only suggest plausible answers, and cannot
indicative of the distribution of attitudes or beliefs
m the populat;nn.
Strengths and limitations of focus group interviews
The advantages of focus-group d -cussions have
oeen discussed extensively in the
erature (3-5, f-Q)
These advantages could briefly be
summarized as
follows.
Focus-group discussions offer many of the advan­
tages of qualitative studies without requiring fullscale anthropological investigations. They provide a
°f ins|ght into motivation, attitudes, feelings
we/f
and behaviour that cannot easily be obtained by
quantitative methods alone • This is p..
probably the
reason why the focus-group discussion approach
has become so popular in recent times.
The group sett, g is believed to be beneficial in
many situations. An informal, supportive group of
peope with similar backgrounds can often put
people at ease, and encourage them to express their
v.ews freely and frankly. It enables participants to
elaborate on ideas, and the group interaction can
stimulate memories and feelings. Because each par­
ticipant is relating to a group of people with similar
backgrounds, the likelihood of participants giving
answers they think will please the interviewer (a
common problem of surveys) is reduced. In addition
because of the interaction during focus-group dis­
cussion, the moderator has more of a chance to

V|assoff C. Contributions of the micro-approach
to social sciences
research. Report prepared for IDRC, 1987.

91

clarify the questions, and there is less
likelihood of “
icoo iii^ciiuuua of
queshons being misunderstood. F;..u;;y
U11C ,e(
Finally, the
relativeiy free format of the focus-group discussion
allows the moderator to pursue unexpected avenues '1
which are relevant to the topic at hand, but could W
not have been foreseen beforehand.
U

I

There are a number of limitations to focus-grouo
discussions. Firstly, a group setting is not alwavs
ideal for encouraging free expression. Sometimes
the group can inhibit discussion. For example
vlassoff
described
a
focus-group
discussion
amongst adolescent girls in India, during which the
girls were painfully shy, not wishing to discuss their
opinions in front of other people, despite extensive A
efforts to-create a relaxed setting conducive *o dis­
cussion.1’
In addition, care must
must be
be taken
taken in
in 'preparing
'preparing tran­
transcripts
from
taped
discussion.
Chances
of intro­
intro, .
■ '
------- Chances of
ducing error are j—l
"
particularly
high ■'
if the interview has
to be translated from the native language to the lam
guage of the investigator (a problem which is signifi­
cant in multilingual environments).
Focus-group discussions also have many of the limi­
tations of other qualitative methods. Their samples
are small and purposively selected and therefore do
not allow generalization to larger populations. In
addition, as with other qualitative methods, the
chances of introducing bias and subjectivity into the
interpretation of the data are high. Because of this it
is not appropriate to treat the findings from focusgroup discussions as though they were findings
from quantitative research. While the focus-qrouo
discussion can provide plausible insights aM ex­
planations, one should not extrapolate from * jeusgroup discussions to the distribution of responses in
a population. This tenet is not always followed. In
tact Merton, one of the founders of focus-group dis­
cussions, recently expressed his concern that "focusgroup research is being mercilessly misused as
quick-and-easy claims for
validity of the research
are not subjected to furthei. quantitative test" (6).
Methodological issues
Much has been written about the way to conduct
tecus-grouP interviews. However, this discussion has
tended to be superficial, with little empirical back,- >
and many basic questions remaining unanswered.
n21d tS^Ck
d'JJcusses some key issues that still
need to be addressed in order to further develop the
focus-group method.

f

he number of focus-group discussion sessions
Little is known about how many discussion sessions
!ILnteede|d
bG reasonab|y sure that all most
SJe at6d tO the sub<ect of inquiry have been
x p i o reel.

I
t
v
il
t
v
je
ft
a
if

Two aspects of this qjcctic
question need to be explored:
the number of target groups
, > needed and the number
of discussion sessions with each target group. As
discussed earlier, the p_r____
population should be divided
nto homogeneous subgroups> according to charteristics relevant to the research , such as users
and non-users, males and females, workinq
j women
'
and housewives, and geographical
areas. This will
enable the researcher to do a^
- a separate analysis for
groups whose behaviour is different from one
another. It will also help to create a supportive ambi-

D
ex
Ed

Rapp, trimest. statist, sanit. mnnd.. 44 >1991)

lV/c

A
Vi

147 —
of

>

ance for the discussion, as the group members will
have some characteristics in common. In addition,
participants should have similar socioeconomic
status, and possibly educational backgrounds as
well, so that they all feel on an equal footing in the
discussion. Other criteria for creating separate focus
groups might be cultural or religious differences,
gender or age, or any other characteristic which is
likely to stand in the way of free discussion.

*ljn

es
id

jl

Little is known about the number of focus-group
sessions needed for each subgrou Debus et al.1 re­
commend doing at least two focus group interviews
with each subgroup, to compare the results. Text­
books in market research advocate forming additional groups until no hew information comes to light.
Even if only two interviews per target group are car­
ried out, the number of interviews required may be
large. For example, a study of attitudes towards con­
traceptives in India might easily require 24 inter­
views p • geographical area since separate inter­
views would likely be required for males and fem­
ales, for younger and older age groups, and for dif­
ferent caste group (say three different castes). If
more than one geographical area is included in the
study, the number of sessions is multiplied accord­
ingly.

'

r
oas
in

es
do

I

..

For the time being, common sense and financial re­
sources are the only guiding principles. However,
this situation is far from ideal. What is needed is a
T- -methodical investigation into this area which will
permit the users of focu1 groups to make an inui._fQrmed decision on the
timal number of focus
groups for their purposes.

?x

;
In
iis
us
as
re”

>

UP

■W'? 1

Analysis of focus-group discussions
Focus-group discussic
provide a great deal of
data, including intervie notes and often transcripts
of the session. This information
needs
------ -------------- to be analysed and organized in an understandable fashion.
Content coding is often mentioned as the method of
choice. This consists of listening or watching the
tapes and reading the transcripts (if available) to
generate a list of key ideas" for each topic under dis­
cussion. Quotations and ideas are then placed under
the appropriate categories, which can be divided
into subcategories or combined into larger themes.

lr» order to see whether the transcripts are really
necessary, and whether different researchers would
, come to the same conclusions from the same interviews, one of the authors (M. E. Khan) carried out a
/? methodological experiment is part of a larger study.
t-A, The larger study conducted in collaboration with the
International Labour Organisation (ILO) Gene a, was
to evaluate family-planning programmes
t the
workplace, using focus-group interviews as one of
J/ its tools. In this study all the sessions (7 in all) were
both video- and audio-taped. Subsequently the
j -f videotapes along wi’h written transcripts and objectives of the study were provided to seven different professionals, to go through the materials and
® analyse them to draw their own conclusions about
'•c' the study.

ed
Stl1!
V”-

iOSl

sen

ber

13r

I
| >5'

will
for
one
nbi

| S-■
I
■
y

1199”

1»

Although the analysis is not yet final, initial observations suggest the following:

1

1 Debus' M- et aL Communication for cluld survival. Handbook for
®xcel,ence In focus-group research. Washington, D.C., Academy for
Educational Development, 1988.

nWld hlth statist, quart., 44 (1991)

•

•
•

•

Drawing conclusions from the videotape alone is
difficult and time-consuming.
Going through the transcripts is faster and easier.
If conclusions are drawn only on the basis of
seeing or listening to the tapes, there seems to
be a certain amount of variation in drawing conclusionjs or picking up their expression.
If transcripts are used for analysis, and content
analysis is done properly (i.e. care is taken to
note which views are expressed how many times
and by how many participants), the answers or
the conclusions drawn are fairly stable. Video­
tapes or observations of note takers add further
stability to the interpretations.

Practical applications in rural areas or urban slums
It must be remembered that focus-group discussions
were originally developed for market research in de­
veloped countries, where transport and com­
munication are relatively advanced. Therefore, it is
not surprising that some of the methods for con­
ducting focus-group discussions will need modi­
fication in the face of realities in developing country
environments.
This section is based upon recent experience in ap­
plying focus-group interviews in rural areas and in
urban slums of India.
In many instances it was not practical to follow the
usual guidelines, and a number of methodological
issues and problems came to light.
Should focus-group members kyow one another be­
forehand? It is usually recommended that focus
groups consist of individuals who are not acquainted
with one another. It is believed that this increases
the likelihood that group m mbers express them­
selves frankly. However, this . not a practical option
in many rural villages or urban slums where it is
generally very difficult to find people who are not ac­
quainted with one another. Based on experience in
carrying out focus-group discussions in India, we
feel that for topics which are not sensitive, the type
of informant does not make much difference, and
the usual rule of anonymity can be relaxed. How­
ever, in the case of sensitive issues, participants who
do not know one another provide better information
than acquainted ones.
Logistical problems in conducting focus-group inter­
views with women. Focus roup discussions usually
last for one or two hours. Group members are ex­
pected to concentrate on the topics being discussed.
Experience suggests that this is difficult for women
in the Indian context (and perhaps in other societies
where f ee movement of women is socially re­
stricted,'. Often, the women selected for the session
feel it necessary to bring someone with them, es­
pecially the ounger women who are frequently ac­
companied oy their mother-in-law or younger sisterin-law. In addition, mothers are often required to
leave the room to attend to some urgent work (e.g.
to take care of crying children) and subsequently
come back. This interrupts the discussion, and
makes it harder for respondents to follow. Avail­
ability of space where a focus group could be pri
vately conducted is a serious problem in some rural
areas and urban slums (unless the respondents are
ready to come to a community centre such as a
school or Panchayat hall). If the sessions are conduc­
ted in a private home, getting enough space and
privacy might be problematic.

■«

I

149 —

♦
•

RESUME

■

1

L'utilisation des groupes focaux en recherche sociale et comportementale:
problemes methodologiques
L'une des methodes qualitatives d'evaluation rapide
est celle des groupes focaux. Le groupe focal est
compose d'un petit nombre de personnes (en
general 8-12) provenant de la population cible, qui
se reunissent pour discuter de sujets presentant de
I'importance pour une etude ou un projet. En
general, les participants ne sont pas choisis au
hasard et il est recommande qu'ils constituent un
ensemble homogene pour eviter que certaines de
leurs caracteristiques individuelles ne genent le libre
cours de la discussion.
Les groupes focaux peuvent servir a faire naitre les
idees ou etre utilises conjointement avec une
methode quantitative, ou encore etre un moyen de
collecter des donnees primaires. Toutefois, dans le
cas de cette derniere utilisation, il conviendra de
traiter les resultats obtenus avec prudence.
Le principal avantage de la methode des groupes
focaux pour revaluation rapide est qu'elle apporte
une in ormation en profondeur sans qu'il soit besoin
de se livrer a des recherches anthropologiques en
vraie grandeur. Par son cote informel, le groupe
focal est cense mettre a I'aise ceux qui en font partie
et les encourager a s'exprimer en toute iiberte. La
>"m6thode comporte cependant un certain nombre de
'limitations. Le groupe focal represente un petit
' 6chantillon, dont la selection est orientee et qui, par
.-^consequent, ne permet pas k. generalisation a des
^■populations plus importantes. En outre, comme c'est
■le cas des autres methodes qualitatives, le risque est

C!

o
g

f"

ml. Suyono , H. et al . Family planning attitudes in
urban Indonesia: findings from focus-group re­
search. Studies in family planning, 12 (12): 433442 (1981).
|

:S

nd
ng

et

?x

''2. Kowalewski , M. R. Double stigma and boundary
maintenance; how gay men deal with AIDS.
Journal of contemporary ethnography, 17 (2):
127-138 (1988).
,.3. Freedman , R. The contribution of social science re’
search to population policy and family planning
■ x':
programme effectiveness. Studies in family plan­
ning, 18 (2): 57-82 (1987).

I

ns
rial
ex
he
uld

1991'

la sub-

La methode n'atteindra son plein developpement
que si Ton resout un certain nombre de questions
methodologiques. On ne sait guere combien de
seances de discussion sont necessaires pour etre
raisonnablement certain que la plupart des aspects
du sujet traite ont ete explores. On ne voit pas non
plus clairement quels sont les meilleurs moyens
d'analyser les debats d'un groupe focal ni dans
quelle mesure les resultats subissent I'influence des
opinions et des prejuges de I'analyste.
La methode des groupes focaux a ete utilisee avec
grind succes pour les etudes de marche dans les
pays develop^es. II va de soi que cette technique ne
peut etre adaptee aux zones rurales et aux quartiers
urbains desherites des pays en developpement sans
quelques changemers >
methodologie.
Les
methodes permettant d'assurer cette adaptation ne
sont pas encore entierement au point et il est necessaire dans ce domaine important de proceder a des
travaux exploratoires supplementaires.
En conclusion, la methode des groupes focaux peut
apporter un enrichissement considerable a la re­
cherche sociale et comportementale, mais il y aurait
lieu de poursuivre I . etudes methodologiques ex­
perimentales concernant son
utilisation pour
('evaluation rapide.

REFERENCES — REFERENCES

I

k

considerable d'introduire un biais et de
jectivite dans ('interpretation des donnees.

hlth statist, quart., 44 (1991)

4. Morgan , D. L. & Spanish , M. T. Focus
new tool for qualitative research.
sociology, 7 (3): 253-270 (1984).

groups: a
Qualitative

5. Folch -Lyon , E. et al . Focus group and survey re­
search on family planning in Mexico. Studies in
family planning, 12 (12): 409-432 (1981).
6. Merton , R. K. The focused interview and focus
groups — continuities and discontinuities. Public
opinion quarterly, 51: 550-566 (1987).
7. Fern , E. J. The use of focus groups for idea gener­
ation: the effects of group size, acquaintanceship,
and moderator on response quantity and quality.
Journal of marketing research, XIX: 1-13 (1982).

:yf

—- 148 —
Objections to tape recorder. Another problem en­
countered was that some participants objected to the
use of tape recorders. In such instances it was not
possible to record the session, and therefore it was
necessary to rely only on notes. This can impede the
analysis of the interview.
Homogeneity ot the group. Although at present in­
tragroup homogeneity is emphasized, our own ex­
perience reveals that in some cases heterogeneity
may also be useful. For example, in a focus-group
session consisting of women of lower-middle repro­
ductive age, initially we found it very difficult to
stimulate discusssion on the problems related to re­
productive health. But after a while an older woman
(mother-in-law of one of the participants) present in
the group started . Iking. This stimulated the young­
er women who then came forward with very useful
information.
The role of the moderator. The moderator is crucial
in focus-group research. It is the job of the mod­
erator to keep the group focused on the topic at
hand, to encourage - oup members to speak freely,
to ensure that no group member dominates the con­
versation and that all opinions are heard, to create a
supportive atmosphere, to probe when necessary,
and to listen well. However, not much is known
about the effect of the moderator's style on the re­
sults of interviews. For example, does an active
moderator get more and better information than a
quiet, laid-back moderator?; does a challenging
argumentative moderator evoke more or better re­
sponses than a polite friendly moderator? More ex­
perimentation is required with moderator style, in
order to be able to make informed choices on this
important issue. .

The ideal number of respondents within a focus i
group. Usually a focus group has anywhere between
8 and 12 respondents. However, lately there has
been an increasing trend to use minigroups with 4-6
respondents. It would be useful to compare results
from both types of group. Some work on this topic
has been done by Fern (7) who observed that the
number of ideas generated did not double as group
size increased from 4 to 8, and that the ideas prod­
uced in a group were not necessarily superior in
quality to those produced in individual interviews.

Conclusions
This article has shown that focus-group discussions
have considerable potential to be used as a com­
plementary approach to enrich social and beha ^ural research. However, its limitations need to
be appreciated and its indiscriminate use should be
discouraged. This art:cle has also demor strated that
a number of met
dological issues remain un­
answered. There are hardly any methodological
studies evaluating the trustworthiness and useful­
ness of the procedures. It is suggested that ex­
perimental studies should be undertaken to evaluate
qualitative approaches, particularly how focusgroups fare against other qualitative methods, and
how the findings of focus-group research are in­
fluenced by the various procedural differences raised
above. It is also important to experiment in using
the focus-group approach for rapid appraisal of
health-promoting behaviours related to selected dis­
eases. Unless attention is paid to strengthening the
methodology by undertaking evaluative experi­
mental studies, it is feared that the indiscriminate
use of focus groups ma' cause more harm than ben­
efit.

SUMMARY
The use of focus groups as a qualitative method for
rapid assessment is discussed. A focus-group ses­
sion is an in-depth discussion in which a small
number of people (usually 8-12) from the target
population discuss topics that are of importance for
a particular study or project. Generally the par­
ticipants are chosen purposively, and it is recom­
mended that they should be homogeneous with re­
spect to characteristics which might otherwise im­
pede the free flow of discussion.
Focus groups can be used for idea generation, in
’ h a quantitative method, or as a
conjunction with
primary data-collection method. However, if focus
groups are used as a primary data-collection
method, their results must be treated with caution.
The main advantage of using focus-group dis­
cussions during rapid assessment is that they pro­
vide in-depth information without requiring full-scale
anthropological investigations. The informal group
setting is believed to make people feel at ease, en­
couraging them to express their views freely. How­
ever, there are a number of limitations to focusgroup discussions. The samples are: small and pur­
posively selected, and therefore do .not allow gene­
ralization to larger populations. In addition, as with
other qualitative methods, the chances of intro­

ducing bias and subjectivity into the interpretation of
the data are high.
There are a number of methodological issues which
still need to be addressed in order to further develop
the method. Little is known about how many dis­
cussion sessions are neec I to be reasonably sur
that most aspects related io the subject of inquiry
have been explored. The best way to analyse focusgroup discussions, and the extent to .vhich the re­
sults reflect the opinions and biases ot the analyser,
are not well understood.

Les
idee
met
colk
cas
trait

Le t
foca
une
de
vrai
foot
et I
met
Hmi
ech;
con
pop
le c

1. i
i
5
Z

2. F
r

3. .

f

Focus-group discussions have been used very suc­
cessfully by market researchers in developed coun­
tries. Naturally, adapting this technique to rural and
slum areas in developing countries will involve
some changes in methodology. Methods of adapting
iscussions to rural and urban slum setfocus-grou
tings are sun not fully worked out, and more exploratory work in this important area is required.
The conclusions
conclusions indicate that focus-group dis­
The
cussions have considerable potential to enrich social
and behavioural research, and suggest that more ex­
perimental methodological studies in using the
focus-group approach for rapid assessment should
be undertaken.
Rapp, trimest. statist, sanit. mpnd.. 44 ’991)

.■;
.................... ...........................

L'un
est
com
gent
se r
i'imj
gent
base
ense
leur
cour

IV/d

PWF
Media
SDA-RF-CH-1B.1.pdf

Position: 2278 (4 views)