Journal of Advanced Nursing, 1994,20,716-721

The strengths and weaknesses of quantitative

and qualitative research: what method for
nursing?
U n d a T Carr RNMH RMN Dip N Cert Ed (FE) RNT
Lecturer, Department of Professional Development, Wealden College of Health and
Social Studies. East Surrey Hospital, Redhill, Surrey RHi 5RH, England

Accepted for publication 14 January 1994

CARR L T (1994) Journal of Advanced Nursing 20, 716-721

The strengths and weaknesses of quantitative and quahtative research- what
method for nursing?
The overall purpose of research for any profession is to discover the truth of the
disciphne This paper examines the controversy over the methods by which
truth IS obtained, by examining the differences and similarities between
quantitative and qualitative research The historically negative bias against
qualitative research is discussed, as well as the strengths and weaknesses of
both approaches, with issues highlighted by reference to nursing research
Consideration is given to issues of sampling, the relationship between the
researcher and subject, methodologies and collated data, validity, reliability,
and ethical dilemmas The author identifies that neither approach is superior to
the other, qualitative research appears invaluable for the exploration of
subjective experiences of patients and nxirses, and quantitative methods
facilitate the discovery of quantifiable information Combining the strengths of
both approaches m tnangulation, if time and money permit, is also proposed as
a valuable means of discovenng the truth about nursing It is argued that if
nursing scholars limit themselves to one method of enquiry, restrictions will be
placed on the development of nursmg knowledge

DEFINING QUANTITATIVE AND f ^'^ ^''°'^'f' '^^'^ Qualitative research also differs
r»TT AI TT AXTiriT vjfKv xvcw fr°™ quantitative approaches as it develops theory mduc-
tively There is no explicit mtention to coimt or quantify
Quantitative research is also described by the terms the findings, which are instead described m the language
'empiricism' (Leach 1990) and 'positivism' (Duffy 1985) employed durmg the research process (Leach 1990) A
It derives from the scientific method used m the physical qualitative approach is used as a vehicle for stud}Tng the
sciences (Cormack 1991) This reasearch approach is an empirical world from the perspective of the subject, not
objective, formal, systematic process m which numerical the researcher (Duffy 1987) Benoliel (1985), expands on
data are used to quantify or measure phenomena and pro- this asp»ect and describes qualitative research as 'Modes of
duce findings It describes, tests and exammes cause and systematic enquiry concerned with understanding human
effect relationships (Bums & Grove 1987), using a deduct- bemgs and the nature of their transactions with themselves
lve process of knowledge attainment (Duffy 1985) and with their surroundings'
Whereas quantitative methodologies test theory The aim of qualitative research is to describe certain
deductively from existing knowledge, through developing aspects of a phenomenon, with a view to explaining the
hypothesized relationships and proposed outcomes for subject of study (Cormack 1991) The methodology itself
study, qualitative researchers are guided by certam ideas, is also descnbed as phenomenology (Duffy 1985), or as a
perspectives or himches regarding the subject to be mvesti- humanistic and idealistic approach (Leach 1990), with its

716
 Quantitative or qualitative research for nursing*

ongrns lyu^ in the disciplines of history, philosophy, moeased likelihood of bemg generahzable The disadvan-
anthropology, soaology and psychology (Ckinnack 1991) tage, and a weakness of the quantitative approach, is tlut
This histoncal foundation, which is not that of the physi- random selection is timenionsummg, with the result that
cal science domain, has been cited as one of the great weak- many studies use more easily obtamed opportunistic
nesses of qualitative research, and is associated with samples (Duffy 1985) This inhibits the possibilities of gen-
the poor initial uptake of the approach withm nursmg eralization, especially if the sample is too small This is
(Bockmon k Rieman 1987) demonstrated m the study by Gould (1985) who mvesti-
gated nurses' knowledge of isolation procedures withm a
specific health distnct The study makes mteresting com-
HISTORICAL BIAS
ments, but It IS not possible to generalize from its findings
Histoncally the use of true expermients has contnbuted as the sample is too small
greatly to the universal knowledge now acquired, especi- Qualitative research, because of the ln-depth nature of
ally m the field of medicme The quanbtative methods studies and the analysis of the data required, usually
used produced legitimate scientific answers, and as a relates to a small, selective sample (Cormack 1991) A
result of this 'hard' data, action was generated and changes weakness of this can be the suspicion that the researcher
took place (Meha 1982) The qualitative approaches pro- could have been lnfiuenced by a particular predisposition,
duced 'soft' data which were, and are still described by affecting the generalizability of the small scale study
some, as bemg madequate m providmg answers and gener- (Bryman 1988) This suggests that qualitative research has
atmg any changes One can argue that the use of the labels a low population validity However, the strength of this
'hard' and 'soft' data suggests in itself that analysis by num- approach is seen when the sample is well defined, for then
bers IS of a superior quality to analysis by words (Comer It can be generalized to a population at large (Hmton 1987)
1991] Raggucci's (1972) ethnographic nursmg study demon-
Benoliel (1985) considers the role nursmg literature has strated the value of this approach m studymg the benefits
played m giving qualitative research a lower status The and practices of minonty ethnic groups
message, only 9 years ago, was that qualitative research is
pnmeinly for the discovery of knowledge to be tested, and
RELATIONSHIP BETWEEN RESEARCHER
was subsidiary to quantitative research Bockmon &
AND SUBJECT
Rieman (1987) discussed the difficulties qualitative
researchers had before the mid-1980s in achievmg publi-
Relationship in quanbtative research
cation traditional nursmg journals Histoncally, fundmg
for research was awarded mainly to quantitative research In quantitative research the investigator maintain a
reports ( D u ^ 1986), emphasizing the depth of acceptance detached, objective view in order to understand the facts
and respect for this particular method (Duffy 1986) The use of some methods may require no
Quabtative research thus has had a major obstacle to direct contact with subjects at all, as in postal question-
overcome m achievmg recogmtion for its contnbution to naire surveys It can be argued that even interview surveys
knowledge Evaluation of qualitative research has been require the researcher to have little, if any contact with
inhibited through lack of published papers It is because respondents, especially if hired staff carry out most of all
of the recent mcrease in nursmg publications usmg the the interviews (Bryman 1988) The strength of such a
qualitative methodology, that an anadysis of the strengths detached approach is avoidance of researcher mvolve-
and weaknesses of both quantitative and qualitative ment, guardmg agamst biasmg the study and ensuring
approaches can be conducted objectivity
Such an approach was successfully used m the West
Berkshire-based penneal management tnals of Sleep et al
SAMPLING (1984) This midwifery study was mdirectly controlled by
Sampling procedures for each methodology are complex the researchers whose mam involvement, other than ran-
and must meet the cntena of the data collection strategy domly allocating mothers to either the controlled or exper-
Both research approaches require a sample to be identified imental episiotomy group, was to analyse the data, once
which IS representative of a larger population of people or collected The findmgs of this study, through its objec-
objects Quantitative research demands random selection tivity, have contnbuted to knowledge withm this field
of the sample from the study population and the random Spencer (1983) a i ^ e s that little is denved from such an
assignment of the sample to the vanous study groups lnduect researcher-subject relationship especially m the
(Du% 1985) Statistical sampling relies on the study health care settmg His major cnticism is that the detached
sample to develop general laws which can be generalized approach treats the participants as though they are objects
to the larger population The advantage of results obtamed and, as such, places hospitals on par with car repair gar-
from random samplmg is that the findings have an ages Cormack (1991) also emphasizes the weaknesses of

717
 LT Carr

such an approach She argues that the research partici- IS achieved by the abihty of the researcher to manipulate
pants are usually kept m the dark about the study, and are an mdependent vanable m order to study its effects on the
often left untouched by the research itself but are expected dependent vanable
to transfer the findings mto practice These ai^uments are This strength can also be argued to be the weakness of
examples of the criticism that quantitative methods treat the quantitative method, especially where nursing
people merely as a source of data research is concemed The methodology dismisses the
expenences of the mdividual as unimportant, which is,
demonstrated m the Bockmon & Rieman study (1987), and
Researcher-subject relationship in qualitative
regards human bemgs as merely reactmg and responding
research to the environment (Cormack 1991) This causes diffi-
As with quantitative research, qualitative methodologies culties m nursmg research, because nursing uses an hohs-
also have supposed strengths and weaknesses regardmg tic view of people and their environment and, accordii^
the closeness of the relationship between researcher and to Bnones & Cecchini (1991), quantitative methods do not
respondent Duffy (1986) argues that a strength of such an permit this approach
mteractive relationship is that the researcher obtains first- The qualitative approach mcludes methods such as
hand expenence providing valuable meaningful data As groimded theory and ethnographic research (Denzm 1978)
the researcher and the subject spend more time together The strength of the methodology employed lies m the fact
the data are more likely to be honest and valid (Bryman that it has an holistic focus, allowing for flexibility and
1988) the attainment of a deeper, more valid understandmg of
Supporting this argument is the study by Baruch (1981) the subject than could be achieved through a more ngid
which revealed that time and the subsequent relationship approach (Du% 1986) It also allows subjects to raise
built between the researcher and the subjects was crucial issues and topics which the researcher rmght not have
for a genume understandmg of the dilemma faced by par- mcluded in a structured research design, adding to the
ents of sick or handicapped children This appears to be quality of data collected The study by Meha (1982) is a
a major strength of the qualitative approach itself, as good example of these strengths, and its findmgs have con-
Woodhouse & Lavmgwood (1991) pomted out m their tributed to the knowledge of student nurses' perspective
study of a multi-agency substance abuse project They on nursing
claimed that the approach, because of the mteractive A weakness of qualitative methodology, is the possible
method, far exceeded expected evaluation outcomes, by effect of the researchers' presence on the people they are
contributmg to empowerment, and enhanced communi- studying As previously highhghted, the relationship
cation and cleuification of roles among the partners between the researcher and participants may actually
mvolved m the project distort findings
The weakness of such a close relationship is the likeh-
hood that it may become pseudotherapeutic, complicating
DATA
the research process and extendmg the responsibilities of
the researcher (Ramos 1989) The possibility of becommg The data collected in quantitative research are, as men-
enmeshed with subjects could also lead to researchers tioned, hard and numencal The strength of producing
having difficulty in separating their own expenences from numbers as data is that this demonstrates an ordered
those of their subjects (Sandelowski 1986) resultmg m sub- system Such an approach could be viewed as being neces-
jectivity (Cormack 1991) In its most extreme form this is sary m an orgamzation as big as the NHS, for as Spencer
referred to as 'gomg native', where the researcher loses (1983) suggests, prepanng an off-duty rota for 5000
awareness of bemg a researcher and becomes a participant employees needs quantitative methods and a computer
(Bryman 1988) However this may not be entirely negabve This argument is also supported by Kileen's (1981) study
m that It facilitates a better understanding of the subject, regardmg new mothers where there was a need to use
as demonstrated by Oakley (1984) numencal data to identify the nursmg resources needed,
number of nurses mvolved, and what difference they made
to patient outcome, length of stay, cost-effectiveness of dis-
METHODOLOGY charge plannmg and the length of the tune patients stayed
The research processes used in the quantitative approach out of hospital before any re-admission
mclude descnptive, correlational, quasi-expenmental and The opposmg argument, suggesting the invalidity of
expenmental research (Cormack 1991) The strengths of numencal findings, is that data not displaying significance
such methods are that both true experiments and quasi- are often neglected, or alternatively attention is centred on
expenments provide sufficient information about the a nunonty of the respondents leaving the majonty imex-
relationship between the vanables under investigation to plored, m other words there are 'deviant cases' (Cormack
enable prediction and control over future outcomes This 1991) This therefore distorts the evaluation of data

718
 Quantitative or qualitative research for nursmg?

In contrast, the soft data collected in quahtative research (Sandelowski, 1986) Campbell & Stanley (1963) mamtain
identify and account for any 'deviant cases' (Connack that the more similar the research experiment is to the
1991) The nch data produced provide an illuminating natural setting the greater is the vahdify and thus general-
picture of the subject, with great attention often given to lzabihfy of the findings The field studies concemmg pen-
pomtii^ out intricate details Evidence of this is seen in neal management by Sleep et al (1984) (also. Sleep 1984a,
the study by Melia (1982) where student nurses' comments b) all contnbute to the scientific understanding of this
are quoted, enabling the reader to fully understand the aspect of nursing One reason that this can be claimed hes
subject being mvestigated m the fact that the studies took place m a climcal environ-
The comparative weakness of quahtative data concems ment, which mcreased validity
the likelihood that some researchers can become over- The strength of qualitative research is proposed m the
wtiehned by the data collected They may become con- claim that there are fewer threats to external validify,
fused by their mabihty to lumt the scope of the study, t)ecause subjects are studied m their natural setting and
concentratii^ on a few manageable areas (Bryman 1988) encounter fewer controlling factors compared with quan-
In this situation the research can tiecome poorly focused titative research conditions (Sandelowski 1986) The
and ineffective researchers also become so immersed m the context and
subjective states of the research subjects that they are able
to give the assurance that the data are representative of
Reliability
the subject being studied, as seen in Ocddey's (1984)
Quantitative research is considered more reliable than antenatal climc study Paradoxically, the closeness of
qualitative mvestigation This is because a quantitative researchers also threatens the validify of the study if they
approach aims to control or elinunate extraneous vanahles become unable to mamtam the distance required to
within the mtemal structure of the study, and the data descnbe or mterpret experiences in a meanmgful way, as
produced can also be assessed hy standardized testing discussed atrave (Hmton 1987) It is argued, however,
(Duffy 1985) This quantitative strength can be seen m the that this IS worth nsking t)ecause of the tugh level of
comparative analysis of patients' and nurses' perceptions validity achieved by employmg qualitative methodologies
about nursmg activities m a postpartum unit, conducted (Duffy 1985)
hy Morales-Maim (1989)
However one can question the reliahilify of quantitative
ETHICAL ISSUES
research, especially when the data have been stripped from
the natural context, or there have been random or acciden- Conceptually, the ethical considerations for both quanti-
tal events which are assumed not to have happened tative and qualitative research are the same safefy and
(Comer 1991) protection of human nghts These are mainly achieved by
The reliability of quahtative resesirch is weeikened by using the process of informed consent The utilization of
the fact that the process is under-standardized smd relies informed consent is problematic in quantitative research,
on the msights and the abilities of the observer, thus but practically impossible in qualitative methodologies in
makmg an assessment of rehabilify difficult (Dufiy 1985) which the direction that the research takes is largely
The study of Hind et al (1990) examined this issue and unknown (Ramos 1989) Munhall (1988) argues that m-
demonstrated that reliability could be assessed by usmg formed consent can be achieved in quahtative research
mdependent experts to examme various aspects of the pro- by re-negotiation when imexpected events occur, but one
cess of developing grounded theory However, one must can argue m tum that this places greater responsibihfy on
question the feasibilify of employing such a costly process, the researchers, as well as requinng them to possess a high
both m terms of time and money, to venfy the rehabilify level of skill, especially m negotiation
of a qualitative study The ethical weakness of quantitative research concems
the formulation of hypotheses In nursing there are
immense ethical considerations, especially for instance
Validity when it IS explained that improvements will occur m
Although quahtitabve methodologies may have greater patient care when a certam approach is adopted, and the
problems with rehabilify than quantitative methodologies, eventual findmgs of the research do not support this
the position is reversed when the issue is validify The Dewis (1989) used a qualitative approach m her study of
weakness m quantitative research is that the more tightly adolescents and young adults with spmal cord mjiines,
controlled the study, the more difficult it liecomes to con- because of the absence of specific previous research
firm that the research situation is like real hfe The very and the ethical dilemma of formulating a hypothesis
components of scientific research that demand control of on assumptions The quahtative approach, for this
vanables can therefore be argued as operating against reason alone, proved valuable for this particular nursing
external vahdify and subsequent generahzabilify study

719
 LT Carr

posed by both This research approach is called

DISCUSSION
tnangulation
For every strength there appears to be a correspondmg
weakness m both quantitative and quahtative research It
TRIANGULATION
IS this dilemma that has fuelled the debate over which
approach is supenor (Duffy 1986), and which method The mam research areas that tnangulation is concemed
should therefore be adopted for nursing research Nursmg with are issues of data, mvestigator, theory and method-
has a history of being divided, researchers m nursing can ology (Murphy 1989) Morse (1991) argues that tnangu-
ill afford to be divided in attitudes to methodologies for lation not only maximizes the strengths and mmimizes the
this could add to the confusion and the division of the weaknesses of each approach, but strengthens research
profession (Comer 1991) However, the author does not results and contnbutes to theory and knowledge develop-
suggest that ngid imiformify about methodology should ment Silva & Rothbart (1984) hold a different opinion,
be the aim of nurse researchers, as studies have arguing that a compromise resolution seems to ignore the
demonstrated that neither method has the upper hand or significance of work presented that acknowledges vanous
the complete set of answers philosophies of science as factors m research and theory
Choosing just one methodology narrows a researcher's development The literature demonstrates that there is no
perspective, and depnves him or her of the benefits of agreement between researchers about tnangulation This
building on the strengths inherent m a vanefy of research IS not surpnsing when there is no agreement either about
methodologies (Duffy 1986) Atwood (1985) disagreed quantitative or qualitative methods, employed within the
with this, and argued that nursing should adopt quantitat- approach
ive approaches to build nursing into a science He stated The tnangulation study conducted by Comer (1991)
that this would provide nursing with a useful theory base concerning newly roistered nurses' attitudes to and
with practical applications Since this argument was posed educational preparation for canng for patients with ceincer,
by Atwood in 1985, studies have demonstrated that the illustrates both the strengths and weaknesses of the
model of measurement, prediction and causal inference approach
does not easilyfita profession where health, illness adjust- The study revealed a ncher and deeper understanding
ment, recovery, participation and care are frequently the of the subject matter than would otherwise be possible
vanables to be measured, whilst assessing the impact of Quantitative and qualitative approaches were found to
nursing practice (Comer 1991) Relying solely on a quanti- complement each other while the inadequacies of each
tative approach to tmswer research questions has been seen were actually offset However, it also highlighted the time
to have senous limitations (Metcalfe 1983) Reliance solely and cost implications the volume of data produced was
on qualitative approaches has also been shown to have immense and an extremely broad knowledge base was
many limitations, although mainly of a different nature required to anedyse it, which meant that other researchers
(Kileen 1984) were contracted in to work on different parts of the analy-
This debate could be seen as advantageous to nursing sis These findings are similar to those of Murphy (1989)
Researchers are bemg forced to consider the controversial who used the method of tnangulation to study traumatic
issues of both methodologies, and this requires them to life events
have m-depth knowledge of epistemology and method- Considenng the evidence, it seems reasonable to suggest
ology and not to be restncted, as m the past, to the tradition that tnangulation is not the way forward for all nursing
of the physical sciences (Duffy 1985) Preference for a research but that it may help nursing to remove itself firom
specific research strategy is not just a technical choice, it the bipolar debate and restncUons, especially m the light
IS an ethical, moral, ideological and political activity of current financial constramts on health professions
(Moccia 1988) This debate unearths these issues m
relation to both approaches, allowing appropnate methods
to be adopted by researchers in order to answer questions CONCLUSION
and develop nursmg theones Although quantitative and qualitative methods sire differ-
Considermg the facts, it is argued that each approach ent, one approach is not supenor to the other, both have
should be evaluated in terms of its particular ments and recognized strengths and weaknesses and are used ideally
limitations, m the light of the particular research question m combmation It can therefore be argued that there is no
under study (Duffy 1987) However this implies that one best method of developing knowledge, and that
there are only technical differences between the two exclusively valuing one method restncts the abihty to
those of research strategies and data collection procedures progress beyond its inherent boundanes Recognizing the
(Bryman 1988) There is a suggested alternative to this, tension between researchers about quantitative and quali-
that of combimng the approaches, pulling on the strengths tative research, and attempting to understand it, may serve
of each method and therefore counteracting the lirmtations to create relevant and distinctive modes of enquiry m

720
 Qiantitative or quahtative research for nursing'

nursmg It may also help the imification rather than the Duffy ME (1987) Methodological tnangulation a vehicle for
division of nursmg scholars metging quantitative and quahtaUve methods /mctge 1B(3).
From examining research m nursmg, quahtative 130-133
approaches appear to be invaluable for the exploration of Gould D (1985) IsolaUon procedures in one health distnct
subjective expenences of patients and nurses, while quan- Nursing Times 81(7) 47-50
Hmton A (1987) Research Awareness 7 the EthnograpAjc
titative methods facilitate the development of quantifiable
Perspective Ashford, Southampton
information Combmmg the strengths of the methods m
Hmd P S , Scandrett-Hihden, & McCaulay L S (1990) Further
tnangulation, if time and money permits, results m the assessment of method to estimate rehahihty and validity of
creation of even richer and deeper research findings It qualitative research findings Journal of Advanced Nursing
seems that nursing research has the potential to provide a 15(4), 430-435
valuable resoiuxie for the health care system As nursmg Kileen M (1984) From hospitals to home continuous care for
discovers and uses different methodologies, it will assist new mothers and infants Nursing Management 15(3), 10-13
in creating the necessary baleuice m the knowledge Leach M (1990) Philosophical choice Nursing The Journal of
required to develop nursing as both a science and an art Clinical Practice, Education and Management 4(3), 16-18
Meha K M (1982) 'Tell it as it is' — qualitative methodology and
nursing research understanding the student nurse's world
References Journal of Advanced Nursing 7(4), 327-335
Metcalfe C (1983) A study of change m the method of organising
Atwood J R (1985) Advancing nursmg science quantitative the delivery of nursing care m a ward of a maternity hospital
approaches Western Journal of Nursing Research 6(3) Suppl, In Nursing Research Studies m Patient Care (Wilson-
9-15 Bamet J ed), John Wiley, Chichester, pp 119-140
Banich G (1981) Moral tales parents' stones of encounters with Moccia P (1988) A cntique of compromise beyond the methods
the health profession Sociology of Health and Illness 3(3), debate Advanced Nursing Science 10(4), 1-9
275-296 Morales-Mann E T (1989) Comparative analysis of the percep-
Benohel JQ (1985) Advancing nursing science qualitative tions of patients and nurses about the importance of nursmg
approaches Western Journal of Nursing Research 6(3) Suppl, activities m a postpartum unit Journal of Advanced Nursing
1-8 14(6), 478-484
Bockmon D F & Rieman D J (1987) Qualitative versus quanUtat- Morse J M (1991) Approaches to qualitaUve and quantitative
lve nursing research Holistic Nursing Practice 2(1), 71-75 methodological tnangulation Nursing Research 40(1),
Bnones T L & Cecctum, D (1991) Nursmg versus medical 120-123
research Heart and Lung 20(2), 206-207 Munhall P L (1988) Ethical considerations in qualitative research
Bryman A (1988) Quantity and Quality m Social Research Western Journal of Nursing Research 10(2), 150-162
Routledge, London Murphy S A (1989) Multiple tnangulation Application in a pro-
Bums N & Grove S K (1987) The Practice of Nursmg Research gramme of nursing research Nursing Research 38(5), 291-297
Conduct, Critique and Utilization WB Saunders, Oakley A (1984) Taking It Like A Woman Cape, London
Philadelphia Raggucci A T (1972) The ethnographic approach and nursmg
Camphell DT & Stanley JC (1963) Experimental and quasi- research Nursmg Research 21(6), 485-490
experimental design for research Rand McNally, Chicago Ramos MC (1989) Some ethical implications of qualitative
Clark E (1988) Research Awareness 9 The Experimental research Research in Nursing and Health 12(1), 57-63
Perspective Ashford, Southampton Sandelowski M (1986) The problem of rigor in qualitative
Cormack D F S (ed ) (1991) The Research Process in Nursing 2nd research Advances m Nursing Science 8(3), 27-37
edn Blackwell Scientific, Oxford Silva M C & Rothbart D (1984) An analysis of changing trends
Comer J (1991) In search of more complete answers to research in philosophies of science on nursing theory development and
questions Queintitative versus qualitative research methods is testing Advanced Nursing Science 6(2), 1-13
there a way forward'' Journal of Advanced Nursing 16(3), Sleep J, Grant J , Elboume D , Spencer J & Chalmers I (1984)
718-727 West Berkshire perineal management trial Rntish Medical
Denzm N K (1978) The Research Act A Theoretical Introduction Journal 289(6445), 587-590
to Sociological Methods McGraw-Hill, London Sleep J (1984a) Episiotomy in nonnal delivery Nursing Times
Dewis ME (1989) Spinal cord m)ured adolescents and young 80(47), 28-30
adults the meaning of hody changes Journal of Advanced Sleep J (1984b) Management of the permeum Nursing Times
Nursing 4(5), 389-396 80(48), 51-54
Duffy ME (1985) Designing nursmg research the qualitative- Spencer J (1983) Research with the human touch Nursing Times
quantitative dehate Journal of Advanced Nursmg 10(3), 29(12), 24-27
225-232 Woodhouse L D & Livingood W C (1991) Exploring the versatility
Duffy M E (1986) Quantitative and qualitative research antagon- of qualitative design for evaluating commimity substance abuse
istic or complementary' Nursing and Health Care 8(6), protection projects Qualitative Health Research 1(4), 434—445
356-357

721