The approriate uses of qualitative methods

HEALTH ECONOMICS

Health Econ. 8: 345–353 (1999)

METHODOLOGY

THE APPROPRIATE USES OF QUALITATIVEMETHODS IN HEALTH ECONOMICS

JOANNA COAST*Department of Social Medicine, Uni6ersity of Bristol, Bristol, UK

SUMMARY

Ontology, epistemology and methodology are not subjects frequently discussed in health economics, yet they areof great relevance to the question of how, or whether, to use qualitative methods as a means of examining certainissues. The paper discusses the nature of enquiry in health economics and then details the nature of qualitativemethods and the constructivist philosophy with which they are most commonly associated. The paper continuesby examining different areas in the study of economics: neo-classical positive economics, alternative approaches toexplanatory economics and normative welfare economics. For each area the philosophical approach is outlined asare the areas of research interest. Appropriate roles for qualitative methods within these philosophical approachesare then suggested. The paper concludes by warning that health economists should not use qualitative methodsnaively. They must be aware of the potential difficulties: both of inadvertently ending up outside the intendedresearch philosophy and of conducting research which is accepted by neither economists nor qualitativeresearchers. If, however, health economists are aware of ontological, epistemological and methodological issues,they can make an informed decision about the appropriateness of qualitative methods in their research and therebypotentially enhance their ability to answer the questions in which they are interested. Copyright © 1999 JohnWiley & Sons, Ltd.

KEY WORDS — ontology; epistemology; methodology; qualitative methods

INTRODUCTION

It is standard to consider health economics as thediscipline of economics applied to the topic ofhealth. Both the philosophical and methodologi-cal bases of health economics are, as a result,founded upon their similar bases in economics.Although there are discussions of philosophy andmethodology in economics [1–5], some of whichare extremely critical of the methodological posi-tion taken by economists [2,3], there has beenlittle examination within the sub-discipline ofhealth economics of these issues. Practitioners ofhealth economics could, however, by virtue of themultidisciplinary health services research environ-

ment in which they frequently find themselvesworking, usefully contribute to discussions ofmethod in economics.

Health services research has become a multidis-ciplinary enterprise primarily because of the com-plexity of health, health care and the organizationof health care services. Multidisciplinary researchentails not just practitioners of different disci-plines working alongside one another in a ‘purist’fashion, but also the opportunity to develop inno-vative ways of researching particular questions ina cross-disciplinary manner. In discussions ofmultidisciplinary work, however, issues of ontol-ogy and epistemology (see Appendix A for glos-sary) surface very quickly—far more quickly thanin similar discussions between economists. (In-

* Correspondence to: Department of Social Medicine, University of Bristol, Canynge Hall, Whiteladies Road, Bristol BS8 2PR,UK. Tel: +44 117 928 7213; fax: +44 117 328 7325; e-mail: [email protected]

CCC 1057–9230/99/040345–09$17.50Copyright © 1999 John Wiley & Sons, Ltd.

Recei6ed 31 July 1998Accepted 6 January 1999

J. COAST346

deed, issues of ontology, epistemology andmethodology seldom concern economists explic-itly as evidenced by Lawson’s comment that thetraining of economists [untypically for the socialsciences] tends to contain little if any methodolog-ical content [2].) Individuals working in healthservices research can be situated right along thecontinuum of paradigm from the positivist scien-tific quantitative approach (epidemiologists) tothe constructivist interpretive qualitative ap-proach (ethnographers /medical anthropologists).Health services research is still a very new areaand is utilizing many methods, although the ques-tion of how to use the methods to obtain acoherent whole is still an issue [6].

Health economics—as a component of healthservices research as well as economics—has theopportunity to draw on these innovative cross-disciplinary ways of working and there is thepotential to consider researching particular ques-tions in what are generally considered to be un-conventional ways, by using qualitative methods.Such methods include the use of interviews, focusgroups and observation, with the researcher aim-ing to obtain an overview of the area of study andhence to understand the ways in which peopleview, and act in, their everyday lives [7] (for moredetail see Appendix A: Qualitative methods).How such methods can be useful and appropriatein the study of health economics is, therefore, animportant issue to consider.

Health economists themselves work in a varietyof research areas. There is much work of anexplanatory nature examining a large variety ofissues relating to the nature of health care systemsas well as health more generally. The vast major-ity of this work is conducted within the neo-classi-cal framework, although explanatory work canalso be conducted within alternative frameworks.The other main area in which health economistsare currently working is within the arena of nor-mative (welfare or extra-welfare) economics. Herea variety of issues are being explored including alarge body of work relating to preference elicita-tion and the nature of utility functions among keyplayers in the health care sector.

The appropriate role for qualitative work foreach of these groups of health economists is likelyto depend on the research questions which theyaim to answer which will, in turn, be heavilyinfluenced by the philosophy of the researcher.Explicit consideration of the ontological and epis-

temological bases for these different areas of workis, therefore, required in order to consider whatrole qualitative methods might appropriately play:whether, and how, they could enhance the moreconventional use of quantitative methods.

This paper will begin by detailing the nature ofqualitative research methods as well as the con-structivist philosophy with which they are mostfrequently associated. From this discussion it isapparent that such a philosophy is very differentfrom that usually associated with the discipline ofeconomics. The paper then discusses three areasof work for health economists: explanatory workconducted within a neo-classical paradigm; ex-planatory work conducted using alternative ap-proaches; and normative research conductedwithin the framework of welfare or extra-welfareeconomics. For each of these areas of work, thephilosophical basis is examined as is the nature ofresearch questions studied and the role that quali-tative methods might play in enhancing the abilityof researchers to answer important research ques-tions within that philosophy. Finally, the conclu-sion to the paper stresses the importance forhealth economists of awareness of ontological,epistemological and methodological issues inmaking an informed decision about the appropri-ateness of qualitative methods in their research, aswell as the dangers of incorporating these meth-ods naively.

CONSTRUCTIVIST PHILOSOPHY ANDQUALITATIVE RESEARCH METHODS

‘Quantitative methods have developed largely to con-firm or verify theory, whereas qualitative methodshave been developed to discover theory’ (Swansonand Chapman, p. 75 quoting Mullen and Iverson(1986, p. 150)) [8].

Qualitative research is particularly linked with thedisciplines of sociology and anthropology and isdescribed using many terms, including ethnogra-phy, phenomenology, grounded theory, ethno-science, hermeneutics (although some linkspecifically to method, e.g. ‘grounded theory’, andothers to the philosophical basis, e.g. ‘phe-nomenology’). The basic aim of qualitative re-search is, however, to try to ‘grasp phenomena insome holistic way or to understand a phe-nomenon within its own context or to emphasisethe immersion in and comprehension of human

Copyright © 1999 John Wiley & Sons, Ltd. Health Econ. 8: 345–353 (1999)

QUALITATIVE METHODS IN HEALTH ECONOMICS 347

meaning ascribed to some set of circumstances orphenomena’ (Lincoln [9], p. 376). The aim is todevelop theory inductively [10].

Qualitative research is generally conductedwithin a constructivist paradigm. It is not ac-cepted that there is a single reality which can beknown. The position taken is instead relativist,with a belief in multiple realities which are them-selves constructed entities [9]. These realities takethe form of ‘multiple mental constructions’ whichare ‘dependent for their form and content on theindividuals who hold them and on the groups towhich these individuals belong’ (Lincoln [9], p.379). Context, hence, becomes extremely impor-tant in considering these multiple realities, and,more generally, in qualitative research as a whole.

The corresponding epistemology is that re-search findings are created by the interaction be-tween the investigator and the investigated [9].Hence, qualitative methodology requires interac-tion between the investigator and the investigatedwith the aim of exploring individual views andconstructions. The investigator should be open tothe views of the world elicited. Miller and Crab-tree [11] conceptualize the mode of enquiry thus:

‘[The] constructivist inquirer enters an interpretivecircle and . . . must always be rooted to the context.No ultimate truth exists; context-bound construc-tions are all part of the larger universe of sto-ries . . . [The constructivist inquirer performs] anongoing iterative dance of discovery and interpreta-tion’ (p. 13).

As with quantitative enquiry there are stan-dards of rigour which qualitative research shouldattempt to meet [9,12,13]. Because the researcheris not trying to maintain an objective stance,however, the aim is not to eliminate bias but toshow how aspects of the background and motiva-tions of the researcher may have influenced theresults obtained.

Qualitative research is not, however, confinedto use within a constructivist mode of enquiry. Infact, as Lincoln points out, qualitative methodsare merely methods and can, therefore, be usedwithin conventional scientific models of enquiry[9]. As an example she points to the work of Milesand Huberman [7], who categorize themselves asrealists and hence outside the constructivist mode.These researchers advocate qualitative methodsnot only for developing hypotheses, but also sug-gest that there is the potential to use such meth-ods for testing hypotheses [7]. In fact, they

suggest, frameworks for analysis of qualitativedata can be theory driven and to choose to ignorethese conceptual strengths can be self-defeating.

Using qualitative methods outside of construc-tivism implies that although the method can cap-ture holistically the particular phenomenon ofinterest, the analyst does not treat the data in thisway. Instead the data are used reductively; forexample, one such method (content analysis)would be to categorize the data using pre-con-ceived coding schemes and count the number ofoccurrences of each category. With such methods,although the qualitative data can provide addi-tional understanding without losing the claim toscientific validity and reliability, much of thepower and subtlety that they can potentially havemay be lost [9]. Nevertheless, the view acrossmuch of the applied social sciences, as well as inhealth services research, is that the two ap-proaches can complement one another, allowingresearch to benefit from the respective strengthsof each [14]. The strength of qualitative research isits ability to aid understanding, provide explana-tions and explore issues, particularly those of acomplex nature. Its weakness, in comparison toquantitative research, is that it does not provideempirical data which are statistically generalizableto whole populations.

RESEARCH IN HEALTH ECONOMICS

Neo-classical positi6e (health) economics

As already stated, the disciplinary basis of healtheconomics is economics and we must, therefore,look to the philosophy and methodology of eco-nomics to understand the basis of health econom-ics. Health economics as an explanatory science isbased almost in its entirety not just on economics,but on one particular branch of economics (albeitthe main one): orthodox neo-classical economictheory. Neo-classical economic theory comprisestheories deduced from axioms and assumptionsabout behaviour with the intention of providingprediction and explanation. Although neo-classi-cal economics does not reach back so far, Mill’s[15] 1836 essay ‘On the definition of politicaleconomy and the method of investigation properto it’ provides a basic explanation of method as itis still largely understood in economics.


J. COAST348

The ontology and epistemology of positivismhave been influential, as have the views of Popper[1,16]. There is believed to be a single realitywhich is, in principle at least, knowable. Theassumption is that within this single reality thereis a set of laws (constant conjunctions of events[2]) which can be uncovered by research. Neo-classical economic theory takes the form of a setof predictions deduced or inferred from higherlevel principles or assumptions: explanation isgenerally assumed to be the reverse of prediction.The relationships developed within economic the-ory take the form ‘when x then y’ [2]. The unit ofanalysis is the individual: it is assumed that thepreferences of individual consumers and firms canprovide the basic building blocks for the develop-ment of theory (methodological individualism)[17].

Hypotheses are developed and are then tested.During this testing, the investigator and the inves-tigated are perceived to be independent entities.During his/her investigation, the goal of theeconomist is to be objective, studying the objectwithout influencing or being influenced by it. Theaim is to obtain findings which can be asserted tobe ‘true’ [9], in order to confirm or reject thetheory. In fact, economic method is heavilyweighted, at least in theory, towards the falsifica-tion of hypotheses. The falsificationist views ofPopper [1]—the notion that science is character-ized by the formulation and testing of proposi-tions which can, in principle, be falsified withempirical evidence—have exerted enormous influ-ence over the subject. (Although critics havepointed out that, despite vast amounts of evidencethat much of mainstream economics is not reflec-tive of reality [2,18], economists do not generallyin practice accept that their theories have beenfalsified. They proceed instead by developing re-finements of mathematical formulations [18],specifying further assumptions and attempting toendogenize more and more variables [2].) Theideal method for health economists workingwithin this philosophy to determine whether theirtheories are true would be to conduct controlledexperiments in order to eliminate bias and arriveat the objective truth. Of course, in practice this isseldom possible. Economists have, therefore,tended to use statistical techniques with the aim ofreducing the potential for bias in, mainly observa-tional, data.

Health economists working within thisparadigm and aiming to explain the workings ofhealth care systems or of individuals’ health be-haviour, similarly, tend to be interested in re-search which tests theories derived from standardaxioms about consumer and supplier preferences.There may be little that qualitative methods canoffer in terms of testing theory as such methodsdo not offer generalizable results and are in-evitably context bound. Quantitative methods aremore likely to provide the researcher with objec-tive data across statistically representative samplesof the population and could, therefore, be ex-pected to provide stronger evidence. There may,however, be appropriate roles for qualitative re-search even within this paradigm. Whilst mosttheory development in neo-classical economics isderived deductively from standard axioms withthe incorporation of additional assumptions,qualitative methods have been used, albeit infre-quently, in the more inductive development oftheory related to the operation of the market (seefor example [19,20]). Such theory developmenthas stayed within a predominantly neo-classicalparadigm by using preferences and utility func-tions as the starting point for the development oftheory. Qualitative methods may also be used toenhance quantitative studies, either as a precursorto the quantitative analysis or as an explanatoryadjunct. Within health services research, qualita-tive methods have been used in the developmentof quantitative survey tools. For example, semi-structured interviews are often conducted withpatients suffering from particular conditions priorto the development of health outcome question-naires for use in clinical trials. Such a method ofdeveloping questionnaires ensures both that rele-vant concepts have been captured and that thelanguage used in questionnaires is comprehensibleto survey respondents. Similarly, within healthservices research, qualitative methods are usedalongside experimental studies to improve under-standing of the quantitative results obtained.Thus the objective data are provided by quantita-tive enquiry, but where its results appeared con-fusing or unintuitive, qualitative research can beused to question its meaning. This role for quali-tative research has the advantage that the meth-ods used provide a powerful tool forunderstanding [7].

Such roles may be of great benefit where ex-planatory empirical work in health economics is



conducted in the neo-classical paradigm. It mayimprove the precision of results obtained as wellas providing greater understanding, but withoutquestioning the basic axioms within which re-searchers are working.

Alternati6e explanatory approaches to (health)economics

Mainstream orthodox economics is increasinglysubject to attacks from both inside and outsidethe discipline. These attacks centre on the predic-tive failure of economics and generally questionthe reality of the view of the world that is pro-vided by economists (see for example [2,3,18,21]).Various ‘solutions’ are offered. Although somequestion whether reality is actually an issue foreconomists (for example, see Rosenberg’s [3,22]notion that economics can be conceptualized as abranch of applied mathematics), for others thetask of making economics more responsive toproblems in the real world is a major issue.

Lawson identifies the problem with mainstreamorthodox economics as being primarily related toits positivist deductive ontology. He describes theform of explanation in economics as requiringuniversal regularities in the form of ‘whenever xthen y’ [2,21]. In neo-classical economics the ax-ioms of orthodox theory, whilst being formulatedat a relatively high level of generality, are inter-preted as secure event regularities in the realworld [21]. The nature of knowledge supported bymainstream economics is that reality is that whichis given in experience; thus the task of science is toseek empirical evidence of event regularities [21].The problem identified by Lawson is that theexistence of such event regularities in the socialrealm of the real world is highly suspect. Hencemainstream economics is destined for failure.

Lawson’s suggested alternative for economics isto take a different ontological perspective: that ofcritical realism [2,23]. Here the notion is thatstructures, mechanisms, powers and tendenciesexist which are irreducible to events but whichunderlie these events and govern or facilitatethem. Causal laws do not relate to event regulari-ties, but instead to ‘tendencies which may or maynot produce events which may or may not beobserved’ (Wainwright [24], p. 1264). Explanationin this ontology is, therefore, concerned withproviding an account of the structures, mecha-nisms, powers and tendencies which together pro-

duce the phenomenon of interest (the event) [21].Prediction is not a goal of such research.

Whilst Ormerod accepts that economics canprovide useful insights at the micro-level, his viewis that macro-level economics cannot be deducedfrom the extrapolation of the behaviour of singleindividuals. An example of such extrapolation inhealth economics would be the suggestion con-tained in papers by Culyer [25] and Lindsay [26],produced during the early days of health econom-ics, that the existence of the NHS in the UK canbe explained by the notion of consumers maximiz-ing a utility function which contains some elementof altruism. It is Ormerod’s [18] view that:

‘For all its apparent mathematical sophistication, thecore model of theoretical economics, that of compet-itive general equilibrium, is premised upon an en-tirely faulty view of the modern world. Behaviouralprecepts derived from an autonomous, deterministicindividual on a desert island, the idealised RationalEconomic Man of economic theory, will not apply tohuman beings en masse in a large, modern economy’(p. 208).

Ormerod does not offer a specific solution, butsuggests that economics as a discipline is close tosubjects such as palaeontology where data areincomplete, subject to error and only obtainedfrom one history. In these disciplines, he suggests,research consists of ‘careful collection and siftingof data’ [18] (p. 210) with theories built aroundthe facts rather than ‘pursuing abstract theorieson how a rational world ought to operate’ [18] (p.210).

These recent challenges do not provide the onlychallenges to mainstream economics. Longstand-ing alternative schools of thought (includingHayek’s subjectivism [2] and institutional eco-nomics [17]) explicitly depart from the deductivistmode characteristic of mainstream economics.For example, institutional economics reconstructstheory as a set of patterns which fit together. Theinstitution (rather than the individual) is the mainunit of analysis. (The preferences of individualsare not seen as causal factors, because they areguided in their behaviour by institutions; hencestarting a theory by looking at the behaviour ofindividuals is pointless.) Because institutionaleconomists are concerned with patterns, the evi-dence they require is different from that requiredby mainstream economists. Specifically they areconcerned with obtaining structural evidence andthe method by which evidence is obtained is by


J. COAST350

checking different sources of evidence—for exam-ple historical and case studies—in order to evalu-ate the plausibility of initial interpretations [17].

Alternative philosophical bases for the study ofeconomics have in common a concern with expla-nation in the real world, rather than the use of‘neat’ deductive models with the potential forincreasingly sophisticated mathematical manipu-lation. Qualitative approaches, which lend them-selves to the development of theory grounded inreal experience, potentially have a role to play inthe development of theory relevant to these alter-native approaches. The more piecemeal nature ofexplanation which would inevitably arise is un-likely to be considered problematic in these alter-native approaches, in contrast to the neo-classicalframework.

The use of qualitative methods in these alterna-tive approaches is, however, likely to be verydifferent from attempts to use qualitative methodsfor theory development within a neo-classicalframework (which would start from an assump-tion that individuals’ preferences are all impor-tant). From these alternative perspectives(subjectivism provides the exception) the types ofquestions posed are likely to be very different:theory about the workings of the economy isunlikely to be built (solely) from information,however well-embedded in context, about individ-ual’s preferences. Lawson [2], for example, arguesfor the reality of social structures that are notirreducible to individuals. For instance, to returnto the example of theories explaining the existenceof the NHS in the UK, ‘alternative’ theorieswould not attempt to explain the NHS’s existenceand the form which that service takes by consider-ation of individual preferences. Instead it wouldbe more likely to consider the historical and insti-tutional conditions which led to the developmentof such a system. Consideration of individuals’views about the health service would be the wrongstarting point for this type of research.

Normati6e welfare or extra-welfare (health)economics

So far the discussion in this paper has related topositive, or explanatory, economics. Many healtheconomists, however, conduct work primarily innormative economics, either from a welfarist oran extra-welfarist perspective. Here the aim is todetermine how policy should proceed, and whilst

normative economics is allied with neo-classicalpositive economics in the importance it assigns tothe preferences of consumers, it is also allied witha particular ethical theory, in which well-being isidentified with the satisfaction of preferences [27].This is a value judgement which may be contro-versial [27], but if the concern of policy-makers isto maximize welfare through the satisfaction ofpreferences (or to maximize preferences for healthoutcomes as in the extra-welfarist perspective [28])then the explanatory power or otherwise of posi-tive economic theory is, in a sense, unimportant,as is its philosophical foundation. Healtheconomists working in such a policy environmentmay have the very pragmatic aim of providinginformation that will enable policy-makers tomaximize the preferences of, for example, patientsor citizens (perhaps by developing appropriateincentive schemes for agents). In order to maxi-mize these preferences or to develop such incen-tive schemes, information will be needed aboutwhat these preferences are, and thus the researchthat interests health economists working in thisarea is likely to include provision of informationabout these preferences. Indeed this has been afruitful area of research for many healtheconomists.

In identifying preferences, qualitative methodsmay be appropriately used in a number of ways.Without necessarily taking a constructivist view ofthe world (with its associated concept of multiplerealities) methods such as semi-structured inter-views or focus groups can be used to talk toindividuals and identify themes relating to theirpreferences. Using qualitative methods in such away could prove helpful in enabling researchers torelate preferences to context and may also providea better understanding to policy-makers of howindividuals might receive benefit or disbenefitfrom the health care system. Qualitative methodsmay also be appropriately used in developingsurvey instruments to utilize with preference elici-tation techniques such as conjoint analysis [29].This approach avoids restricting, to those sug-gested by the researcher, the sources of utilityfrom amongst which survey respondents canchoose, but maintains the advantages of the useof a survey which can be sent to a representativesample of sufficient size to achieve statistical sig-nificance. Qualitative methods can also providean ideal method for explaining results obtained byquantitative techniques, and this method has in-



deed been used by Robinson et al. [30] to explainconfusing results obtained during a large-scalesurvey using visual analogue scales and timetrade-off for the elicitation of preferences for dif-ferent health states from a population sample.Similarly the method has been used by O’Hanlonet al. [31] in exploring issues relating to theEuroQol.

CONCLUSION: APPROPRIATE ROLES FORQUALITATIVE RESEARCH

This paper has examined the ontology and episte-mology generally associated with qualitativemethods. Qualitative methods are drawn from avery different ontology and epistemology fromthat generally associated with explanatory neo-classical economics, and hence much of the workof health economists. There are, however, otherareas of work in health economics: explanatorywork conducted outside a neo-classical frame-work has not dominated, but normative (policy)work within the welfarist or extra-welfaristparadigms provides a considerable proportion ofthe work of health economists. As has been dis-cussed, the roles which can appropriately beplayed by qualitative research are likely to varysignificantly across these different areas, in partbecause of the different philosophical approachesand in part because the questions in which re-searchers are likely to be interested will differ.

Qualitative methods are certainly a useful toolfor those who are concerned in obtaining explana-tions for phenomena which are inevitably context-dependent. They can also be helpful when usedalongside quantitative methods either for develop-ment of quantitative instruments or additionalunderstanding. If health economists do choose touse these methods, however, they must be awareof where they are appropriate and where they arenot, the sorts of questions they can help in an-swering and those where the methods will be lessuseful, and how they should be used to answerspecific types of question. Such awareness comesin part from a consideration of the philosophyassociated with different research perspectives,and in part from the particular research questionof interest (itself likely to be heavily influenced bythe philosophy of the researcher).

There is great danger for health economists inincorporating these methods naively. First, there

is the possibility of inadvertently ending up out-side the intended research philosophy. For exam-ple, neo-classical health economists naivelypursuing qualitative methods could find them-selves outside their philosophical framework alto-gether. As Miles and Huberman [7] (themselvesrealists rather than relativists) state: ‘As qualita-tive researchers collect data, they revise theirframeworks—make them more precise, replaceempirically feeble bins with more meaningfulones, and reconstrue relationships’ (p. 20). Meth-ods designed for building theory grounded inreality may inevitably come up with somethingother than that derived on the basis of logicalassumptions and notions of rationality! Second,undertaking qualitative research within a pre-de-fined framework may lead to the potential diffi-culty that research would please neitherqualitative researchers, who may criticize it forbeing restricted by the theory to which it sub-scribes, nor economists, who may find the workinsufficiently rigorous and objective, and who, insome cases, may be unable to accept challenges toaxioms held dear.

If, however, health economists are aware ofontological, epistemological and methodologicalissues, they can make an informed decision aboutthe appropriateness of qualitative methods intheir research. Such appropriate use of qualitativemethods can only enhance the ability of healtheconomists to answer the questions in which theyare interested.

ACKNOWLEDGEMENTS

I would like to thank the following for helpful comments onearlier drafts of the paper: Jenny Donovan, Richard Smith,Tessa Peasgood and two anonymous referees. They bear noresponsibility for the final version.

APPENDIX A: GLOSSARY

Constructi6ism : also referred to as ‘naturalism’,this is the notion that research should be carriedout in non-contrived settings [9]. It has its ownparticular interpretations of ontology, epistemol-ogy and methodology—see main text.

Deducti6e : a type of theory generation, wheretheory is developed on the basis of logic combinedwith axioms and assumptions.


J. COAST352

Epistemology : view of the relationship betweenthe knower and the to-be-known or knowable [9].

Hayek’s subjecti6ism : ‘The scientist is concernedwith individual beliefs and attitudes only and theaim is not to explain such beliefs but merely tograsp them’ (Lawson [2], p. 142).

Inducti6e : a type of theory generation, wheretheory is developed on the basis of particularexperience.

Methodology : view of the means by which theinquirer goes about finding out about the to-be-known [9].

Ontology : view of the form and nature of real-ity [9].

Positi6ism : the notion that there is a singlereality, which can be researched on the basis oflived experience.

Qualitati6e methods : include methods such asin-depth, semi-structured interviews, focusgroups, participant observation: ‘The quest for auseful organizational map of qualitative methodsis not unlike the quest for the Holy Grail. Themethods derive from multiple disciplines and fromat least 20 or more diverse traditions, each with itsown particular language’ (Miller and Crabtree[11], p. 13). Miles and Huberman [7] set out thegeneral principles which describe the nature ofqualitative research:

� ‘Qualitative research is conducted through anintense and/or prolonged contact with a‘‘field’’ or life situation . . . ’

� ‘The researcher’s role is to gain a ‘‘holis-tic’’ . . . overview of the context under study:its logic, its arrangements, its explicit and im-plicit rules’.

� ‘The researcher attempts to capture data onthe perspective of local actors ‘‘from the in-side’’, through a process of deep attentiveness,of empathetic understanding (verstehen) andof suspending or ‘‘bracketing’’ preconceptionsabout the topics under discussion’.

� ‘Reading through these materials, the re-searcher may isolate certain themes and ex-pressions that can be reviewed withinformants, but that should be maintained intheir original forms throughout the study’.

� ‘A main task is to explicate the ways people inparticular settings come to understand, ac-count for, take action, and otherwise managetheir day-to-day situations’.

� ‘Many interpretations of this material are pos-sible, but some are more compelling for theo-

retical reasons or on grounds of internalconsistency’.

� ‘Relatively little standardized instrumentationis used at the outset. The researcher is essen-tially the main ‘‘measurement device’’ in thestudy’.

� ‘Most analysis is done with words. The wordscan be assembled, subclustered, broken intosemiotic segments. They can be organised topermit the researcher to contrast, compare,analyse and bestow patterns upon them’ (pp.6–7).

Realism : an ontological view in which it isbelieved that there is a single knowable reality.There are different versions of realism: empiricalrealism is associated with positivism and assumesthat reality is available through experience; criti-cal realism takes the view that reality is not onlyconstituted by events ‘but also (irreducible) struc-tures, mechanisms, powers, and tendencies that,although perhaps not directly observable, never-theless underlie actual events and govern or facili-tate them’ (Lawson [21], p. 13).

Relati6ism : an ontological view opposed to re-alism in which it is assumed that there is no singlereality.

REFERENCES

1. Blaug, M. The methodology of economics. Or howeconomists explain, 2nd edition. Cambridge: Cam-bridge University Press, 1992.

2. Lawson, T. Economics and reality. London: Rout-ledge, 1997.

3. Rosenberg, A. Economics—mathematical politics orscience of diminishing returns? Chicago: Universityof Chicago Press, 1992.

4. McCloskey, DN. The rhetoric of economics. In:Hausman, D.M. (ed.) The philosophy of economics.An anthology, 2nd edition. Cambridge: CambridgeUniversity Press, 1994, pp. 395–445.

5. Friedman, M. The methodology of positive eco-nomics. In: Hausman, D.M. (ed.) The philosophy ofeconomics. An anthology, 2nd edition. Cambridge:Cambridge University Press, 1994, pp. 180–213.

6. Black, N. Why we need qualitative research. Jour-nal of Epidemiology and Community Health 1994;48: 425–426.

7. Miles, M.B. and Huberman, A.M. Qualitati6e dataanalysis: an expanded sourcebook. London: SagePublications, 1994.

8. Swanson, J.M. and Chapman, L. Inside the blackbox: theoretical and methodological issues in con-



ducting evaluation research using a qualitative ap-proach. In: Morse, J.M. (ed.) Critical issues inqualitati6e research methods. Thousand Oaks: Sage,1994.

9. Lincoln, Y.S. Sympathetic connections betweenqualitative methods and health research. Qualita-ti6e Health Research 1992; 2: 375–391.

10. Strauss, A. and Corbin, J. Basics of qualitati6eresearch. Grounded theory procedures and tech-niques. London: Sage, 1990.

11. Miller, W.L. and Crabtree, B.F. Primary care re-search: a multimethod typology and qualitativeroad map. In: Crabtree, B.F. and Miller, M.L.(eds.) Doing qualitati6e research. London: SagePublications, 1997, pp. 3–28.

12. Altheide, D.L. and Johnson, J.M. Criteria for as-sessing interpretive validity in qualitative research.In: Denzin, N.H. and Lincoln, Y.S. (eds.) Hand-book of qualitati6e research. London: Sage, 1994.

13. Naidoo, J. and Orme, J. Qualitative and quantita-tive research: an opportunity to restore the balance.Critical Public Health 1998; 8: 93–95.

14. Dingwall, R., Murphy, E., Watson, P., Greatbatch,D. and Parker, S. Catching goldfish: quality inqualitative research. Journal of Health Ser6ices Re-search and Policy 1998; 3: 167–172.

15. Mill, J.S. On the definition and method of politicaleconomy. In: Hausman, D.M. (ed.) The philosophyof economics. An anthology, 2nd edition. Cam-bridge: Cambridge University Press, 1994, pp. 52–68.

16. Hausmann, D.M. Introduction. In: Hausmann,D.M. (ed.) The philosophy of economics. An anthol-ogy, 2nd edition. Cambridge: Cambridge Univer-sity Press, 1994, pp. 1–50.

17. Dugger, W. Methodological differences betweeninstitutional and neoclassical economics. In: Haus-man, D.M. (ed.) The philosophy of economics. Ananthology, 2nd edition. Cambridge: CambridgeUniversity Press, 1994, pp. 336–345.

18. Ormerod, P. The death of economics. London:Faber and Faber, 1994.

19. Mannion, R. and Smith, P. Trust and reputation incommunity care. In: Anand, P. and McGuire, A.

(eds.) Changes in health care. Reflections on theNHS internal market. Basingstoke: Macmillan,1997, pp. 141–161.

20. Coase, R.H. The nature of the firm. Economica1937; November: 386–405.

21. Lawson, T. A realist perspective on contemporary‘economic theory’. Journal of Economic Issues 1995;24: 1–32.

22. Rosenberg, A. If economics isn’t science, then whatis it? In: Hausman, D.M. (ed.) The philosophy ofeconomics. An anthology, 2nd edition. Cambridge:Cambridge University Press, 1994: 376–394.

23. Bhaskar, R. Reclaiming reality. London: Verso,1989.

24. Wainwright, S.P. A new paradigm for nursing: thepotential of realism. Journal of Ad6anced Nursing1997; 26: 1262–1271.

25. Culyer, A.J. and Simpson, H. Externality modelsand health: a Ruckblick over the last 20 years.Economic Record 1980; 56: 222–230.

26. Lindsay, C.M. Medical care and the economics ofsharing. Economica 1969; November: 351–362.

27. Hausman, D.M. and McPherson, M.S. Economics,rationality and ethics. In: Hausman, D.M. (ed.)The philosophy of economics. An anthology, 2ndedition. Cambridge: Cambridge University Press,1994, pp. 252–277.

28. Culyer, A.J. The normative economics of healthcare finance and provision. In: McGuire, A., Fenn,P. and Mayhew, K. (eds.) Pro6iding health care: theeconomics of alternati6e systems of finance and de-li6ery. Oxford: Oxford University Press, 1991, pp.65–99.

29. Ryan, M. Using consumer preferences in health caredecision making. The application of conjoint analy-sis. London: Office of Health Economics, 1996.

30. Robinson, A., Dolan, P. and Williams, A. Valuinghealth status using VAS and TTO: what lies behindthe numbers? Social Science and Medicine 1997; 45:1289–1297.

31. O’Hanlon, M., Fox-Rushby, J. and Buxton, M.J. Aqualitative and quantitative comparison of the Eu-roQol and time trade-off techniques. InternationalJournal of Health Sciences 1994; 5: 85–97.


Documents

The approriate uses of qualitative methods