[rn] decrop 1999 - triangulation in qualitative tourism research

Upload: rafael-pantarolo-vaz

Post on 11-Oct-2015

10 views

Category:

Documents


0 download

TRANSCRIPT

  • Tourism Management 20 (1999) 157161

    Note

    Triangulation in qualitative tourism research

    Alain Decrop*

    Department of Business Administration, University of Namur, Rempart de la Vierge, 8, 5000 Namur, Belgium

    Abstract

    Qualitative research in tourism is often blamed for missing the tenets of good science. There are two major reasons for this. On onehand, positivism is still the prevailing paradigm in many areas of tourism research. On the other hand, qualitative researchers oftenfail to explain how and why their methods are sound. This results in confusion and misunderstandings. In this paper, basic criteria toassess the trustworthiness of a qualitative study are listed, and triangulation is proposed as a way to implement them. Rening theconcepts of corroboration and validation, triangulation consists of strengthening qualitative ndings by showing that severalindependent sources converge on them, or at least, do not oppose them. Denzins four basic types of triangulation (i.e. data, method,investigator and theoretical triangulation) are described and illustrated by appropriate tourism examples. ( 1999 Elsevier ScienceLtd. All rights reserved.

    Keywords: Tourism; Triangulation; Positivism; Interpretivism

    1. Introduction

    Recently, we presented a paper about a groundedtheory of Belgian holiday makers (Decrop, 1998) to col-leagues. While the focus of the study was on the holidaymakers decision-making processes, the discussion quick-ly evolved to the epistemological and philosophicaldebate between positivism (more inclined towards quant-itative research) and interpretivism (whose topics arebetter approached by qualitative methods). One col-league asked to what extent a qualitative approach wasmore than just pre-scientic. This question gave rise tothis paper since this is a typical reaction which confrontsa large number of qualitative researchers in tourism.

    Nowadays, qualitative methods are widely used inmarket research and are gaining large acceptance in thesocial sciences. In travel and tourism research, anthro-pologists and sociologists have been turning to qualitat-ive approaches for a long time (e.g. Boorstin, 1964;Cohen, 1972, 1973; Graburn, 1976; MacCannell, 1973,1976; Smith, 1977). This is not the case for researchersfrom economy, geography, psychology or marketing.Riley (1996) notes that the majority of tourism market-ing research has relied on structured surveys and

    *Tel.: 0032 81 72 48 82; fax: 0032 81 72 48 40; e-mail: [email protected]

    quantication (p. 22). The subordinate and exploratorynature of qualitative research is explicitly recognized:qualitative techniques are used to provide informationfor developing further quantitative research. The bestindication of the prevalence of quantitative research isfound in tourism journals (for a review, see Riley & Love,1997). Among the major journals, only the Annals ofourism Research oer enough space to stand alonequalitative articles. Elsewhere, qualitative methods areoften used as a forerunner to quantitative techniques.There are more reasons for this lack of consideration ofqualitative research.

    A rst explanation is the persistence of the dominationof the positivist paradigm in many areas of tourismresearch. Positivism considers reality to be objective,tangible and single. Interest is focused on what is general,average and representative so that statistical generaliz-ation and prediction are possible. Positivists do not havea lot in common with interpretivists claim for a multipleand socially constructed reality and their focus on what isspecic and unique in order to understand and generateinterpreted meaning. They further blame qualitative re-search for lacking rigor and validity.

    But problems related to qualitative approaches do notonly result from the endurance of the positivist paradigm.Interpretivist researchers, on their side, often fail to ex-plain and justify how and why their qualitative ap-proaches are sound. The methodological introduction of

    0261-5177/99/$ see front matter ( 1999 Elsevier Science Ltd. All rights reserved.PII: S 0 2 6 1 - 5 1 7 7 ( 9 8 ) 0 0 1 0 2 - 2

  • papers is mostly limited (sometimes by space constraints)to describing the research design or (and) mentioningreliability and validity criteria, but without showing howthese criteria are implemented. Moreover, positivists lackfamiliarity with qualitative terms and interpretivists failto explain them, which results in confusion and miscon-ceptions. For example, consider how well-known con-stant comparison (Strauss & Corbin, 1990) is whencompared to factor analysis!

    In this paper, we do not want to go into the qualitativevs. quantitative research debate. We think that method-ological eclecticism is desirable: research questions or,more precisely, the relationship between the knowledge(phenomenon) and the knower (person or thing possess-ing the knowledge) must direct the choice of appropriateresearch design and methods. In this paper, we want tohelp bridging the gap between positivist and inter-pretivist tourism researchers by:

    1. Listing the criteria by which a qualitative studystrustworthiness can be assessed;

    2. Proposing triangulation as a way to make qualitativendings more sound, and to gain larger acceptance ofqualitative tourism studies.

    Before going into this, it should be pointed out that thoseissues of trustworthiness and triangulation are not rel-evant for any qualitative researcher. Qualitative researchentails dierent paradigms. Denzin and Lincoln (1994)outline ve moments that include dierent sets of on-tological, epistemological and methodological stances.While corresponding to an historical evolution, these vemoments are still in operation and all are legitimate invarious disciplines. Issues of trustworthiness and tri-angulation are only relevant to researchers operating inthe second and parts of the third moments, that is inpost-positivism, interpretivism and constructivism.Soundness and triangulation are a non-issue for re-searchers operating in other paradigms (feminism, (post-)structuralism, critical theory, cultural studies,2). Theexample of Hirschman (1992) writing about her ownstruggles with cocaine addiction while still being anacademic speaks for itself.

    2. Criteria of trustworthiness

    Qualitative approaches are criticized because of theirlack of rigor and credibility. Both reliability and validityare put into question since homogeneity of data andcoecients of determination cannot be computed. It isnot the usefulness of qualitative data that is at stake here,but rather the criteria by which the trustworthiness ofa qualitative study can be judged. This issue of soundnessgoes beyond the quantitative/qualitative debate: All re-search must respond to canons that stand as criteria

    against which the trustworthiness of the project canbe evaluated (Maxshall & Rossman, 1995, p. 143).Lincoln and Guba (1985) have developed four precisecriteria for qualitative inquiry that parallel the quantita-tive terminology:

    1. Credibility (internal validity): How truthful are par-ticular ndings?

    2. Transferability (external validity): How applicable arethe research ndings to another setting or group?

    3. Dependability (reliability): Are the results consistentand reproducible?

    4. Conrmability (objectivity): How neutral are the nd-ings (in terms of whether they are reective of theinformants and the inquiry, and not a product of theresearchers biases and prejudices)?

    The previous criteria are very useful in establishingcanons for qualitative research. However, they must notbe limited to contemplation, they rather need to be im-plemented in research designs. This is done throughtriangulation. Next to triangulation, there are other tech-niques that help enhancing the trustworthiness of quali-tative analysis, like testing rival explanations, looking fornegative or atypical cases, or keeping methods and datain context (Patton, 1990). The credibility of the researcheris also at stake because it aects the way ndings arereceived. Issues of training, experience, perspective,status, and presentation of self in the research projectshould be addressed.

    3. Triangulation

    Based on the triangle analogy, triangulation impliesthat a single point is considered from three dierent andindependent sources. Derived from topography and rstused in the military and navigation sciences, the concepthas been fruitfully adapted to social science inquiry.Campbell and Fiske (1959) introduced the concept, asa synonym for convergent validation, in the presentationof a multimethod/multitrait matrix. Authors like Webbet al. (1966), and Jick (1979) rened triangulation asmixing qualitative and quantitative methods, advocatingthat both should be viewed as complementary instead ofrival camps. Later, triangulation received more attentionin qualitative research as a way to ground the acceptanceof qualitative approaches (Denzin, 1978; Rossman& Wilson, 1985).

    Triangulation means looking at the same phenom-enon, or research question, from more than one source ofdata. Information coming from dierent angles can beused to corroborate, elaborate or illuminate the researchproblem. It limits personal and methodological biasesand enhances a studys generalizability. Denzin (1978)

    158 A. Decrop / Tourism Management 20 (1999) 157161

  • identies four basic types of triangulation. By combiningdata sources, methods, investigators, and theories, tri-angulation opens the way for richer and potentially morevalid interpretations. The researcher can also guardagainst the accusation that a studys ndings are simplythe artifact of a single method, a single data source, ora single investigators bias (Henderson, 1991, p. 11).Triangulation should be considered from the very start ofthe research design, but it is in data analysis that thestrategy of triangulation really pays o (Patton, 1990,p. 464).

    3.1. Data triangulation

    Data triangulation involves the use of a variety of datasources in a study. There are more ways to achieve this.First, dierent types of material can be collected. Next toprimary data resulting most often from interviews or(and) observations, secondary data are an importantsource of information for the qualitative researcher.Documents can be of multiple types (textbooks, novels,promotional material, minutes of meetings, newspapers,letters, etc.). They are found in libraries, shops, documen-tation centers, and in personal, institutional or organiza-tional archives. Next to written material, photographs,lms, videos and music are valuable documents. Datatriangulation has often been used, while not always inexplicit terms, in recent qualitative tourism research. Forexample, Markwell (1997) uses data from photographsand travel diaries, next to observation and interviewtranscripts, to investigate the dimensions of photographyin a nature-based tour. In their study of tourism market-ing images of industrial cities, Bramwell and Rawding(1996) triangulate data from committee papers, promo-tional brochures and structured interviews.

    A second way to triangulate data is to write eldnotes during and immediately after each interview orobservation session. These notes are especially useful asthey shed additional light on the textual content orindicate specic questions that do not directly appearin the interview transcripts. Things to be observed arenot limited to verbal activities. Non-verbal behavior,communicational aspects (audience reaction) and globalelements (group behavior, body gestures, combined ver-bal and non-verbal, etc.) often give precious clues if notdirect insight. Elements of the environment (weather,atmosphere, setting, furniture, etc.) can also be usefullydescribed.

    3.2. Method triangulation

    Method triangulation entails the use of multiplemethods to study a single problem. This can be dierentqualitative methods or a combination of qualitative andquantitative techniques (but not in a hierarchical orderlike qualitative exploration and quantitative inference).

    Since each method has its own limits and biases, andsingle methodologies result in personal biases, usingmultiple methods paves the way for more credible anddependable information. Projective techniques areparticularly useful to triangulate the more conventionalinterviewing and observational methods. Projectivetechniques are indirect means of qualitative questioningthat enable the informant to project beliefs and feelingsonto a third person, to an inanimate object, or into atask situation (Haire, 1950). The basic assumption isthat the informant interprets information on the basisof his own preoccupations, needs, and values, whichcan be hidden or latent. Projective techniques arecharacterized by the ambiguity of the proposed material(questionnaires, pictures, drawings, etc.) or situations(stories, events, etc.). Again, examples of methodtriangulation are found in the recent tourism researchliterature: Markwell (1997) used both participant obser-vation and semi-structured interviews; Corey (1996), andMarti (1995) carried out focus groups (with content anal-ysis) and structured questionnaires (with quantitativeanalysis), and Dann (1996), open-ended interviews andphoto-driving.

    3.3. Investigator triangulation

    Investigator triangulation is concerned with using sev-eral dierent researchers to interpret the same body ofdata. Next to the investigators subjective understanding,gender, race and culture can also bias qualitative analy-sis. This type of triangulation takes a lot of extra eortand time. The best solution is to work in a team. Indepen-dent investigators can also be asked to examine a part ofthe data and to conrm or invalidate prior interpreta-tions. This should meet the dependability requirement,i.e. that, under the same circumstances, the same inter-pretation would occur.

    Investigator triangulation should also include mem-ber checking, i.e. informants are invited to read theirtranscripts and (or) a summary of the analysis providedby the researcher, and to comment on it. Any remark,disagreement with interpretations, or additional informa-tion should be reintroduced in the analytical process.Again, this may enhance the credibility of the analysis.A last type of investigator triangulation is letting anauditor (devils advocate) regularly review the datagathering and analysis processes to conrm adherence tosound research practices. This is to ensure a consistencyof rigor in the qualitative research process (Lincoln& Guba, 1985). In his investigation of travel-related pres-tige, Riley (1995) took particular care in triangulatinginvestigators. On one hand, he invited interviewees todetermine whether his ndings were accurate reectionsof their conversation. On the other hand, he asked a neu-tral qualitative analyst to review his transcripts to con-rm the plausibility of his interpretations.

    A. Decrop / Tourism Management 20 (1999) 157161 159

  • 3.4. Theoretical triangulation

    Theoretical triangulation involves using multiple per-spectives to interpret a single set of data. One mightexamine interviews with holiday makers from dierentdisciplinary angles: anthropology, psychology, sociology,marketing or economics. Even within one eld, multipletheoretical perspectives may be considered: so in psy-chology, behavioral, cognitive or Gestalt psychology.Related to the above disciplines, dierent research per-spectives (e.g. ethnography, phenomenology, symbolicinteractionism, etc.), which have their own theoreticalimplications, could be triangulated.

    In the inductive analytical process of building atheory, multiple sources of evidence should be broughttogether to dene a construct or a causal relation.One should wonder if each new data chunk corroboratesor opposes the emerging theory. This is in line withGlaser and Strauss (1967) principle of permanentlyasking questions and making comparisons. Confrontingemerging hypotheses with existing theories (nomatter whether they are based on quantitative orqualitative approaches) and searching for alternativeexplanations further help to make conclusions moresound. In their investigation of residents attitudestowards a new tourist enclave in Puerto Rico, Hernandezet al. (1996) confronted emerging ndings to threetheoretical frameworks for understanding locals reactionto tourism.

    4. Conclusion

    Qualitative research is often qualied as bricolage orart, in contrast with quantitative research, which ishonored as being rigorous and scientic. In this paper,we have shown that this is not the right debate. If weaccept the principle that science is not a question ofnumbers but of reasoning, a qualitative study can be assound as a quantitative one. The tourist researcher mustnot only be conscious of the criteria which make a quali-tative study trustworthy, but s/he has to implement them.Triangulation can help this. Rening the earlier conceptsof corroboration and validation, triangulation consists ofconrming qualitative ndings by showing that indepen-dent sources converge on them, or at least, do not opposethem. Denzins (1978) four basic types of triangulationwere described and illustrated in this paper. This list isnot exhaustive: triangulation is above all a state of mind,which requires much creativity from the researcher. Thesearch for convergence is the motto, in order to makepropositions more sound and valid. Moreover, authorsshould mention explicitly how the data are triangulatedin the methodological discussion of their papers. Thiswould help convincing readers that qualitative inter-pretive research is not only an art but also a science.

    Two additional points were addressed in this paper.First, issues of trustworthiness and triangulation are ormay be not relevant for qualitative researchers operatingin paradigms other than post-positivism, interpretivismand constructivism. Second, triangulation is not the onlycriterion for good qualitative research and it does notensure that ndings are interesting and give appropriateanswers to research problems.

    References

    Boorstin, D. J. (1964). he Image: A Guide to Pseudo-events in America.New York: Harper.

    Bramwell, B., & Rawding, L. (1996). Tourism marketing images ofindustrial cities. Annals of ourism Research, 23, 201221.

    Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminantvalidation by the multitrait-multimehod matrix. Psychological Bull-etin, 30, 81105.

    Cohen, E. (1972). Towards a sociology of international tourism. SocialResearch, 39, 164182.

    Cohen, E. (1973). Nomads from auence: Notes on the phenomenon ofdrifter tourism. International Journal of Comparative Sociology, 14,89103.

    Corey, R. J. (1996). A drama-based model of traveler destination choice.Journal of ravel and ourism Marketing, 5(4) 122.

    Dann, G. M. (1996). Toursts images of a destination - An alternativeanalysis. Journal of ravel and ourism Marketing, 5(1/2), 4155.

    Decrop, A. (1998). Personal aspects of consumers decision making pro-cesses revisited through a grounded theory of Belgian vacationers.Paper presented at the Symposium on the Consumer Psychology ofTravel, Hospitality, and Leisure, Research, 1719 August, Universityof Hawaii at Hilo.

    Denzin, N. K. (1978). he research act: A theoretical introduction tosociological methods. New York: McGraw-Hill.

    Denzin, N. K., & Lincoln, Y. S. (1994). Entering the eld of qualitativeresearch. In N. K. Denzin, & Y. S. Lincoln (Eds.), Handbook ofqualitative research (pp. 117). Thousand Oaks: Sage.

    Glaser, B., & Strauss, A. (1967). he discovery of grounded theory.Chicago: Aldine.

    Graburn, N. H. (1976). Ethnic and tourist arts. Berkeley, CA: Universityof California Press.

    Haire, M. (1950). Projective techniques in marketing research. Journalof Marketing, 14, 649656.

    Henderson, K. A. (1991). Dimensions of choice: A qualitative approachto recreation, parks, and leisure research. State College: VenturePublishing.

    Hernandez, S. A., Cohen, J., & Garcia, H. L. (1996). Residents attitudestowards an instant enclave resort. Annals of ourism Research, 23,755779.

    Hirschman, E. C. (1992). The consciousness of addiction: Towardsa general theory of compulsive consumption. Journal of ConsumerResearch, 19, 155179.

    Jick, T. D. (1979). Mixing qualitative and quantitative methods:Triangulation in action. Administrative Science Quaterterly, 24,602611.

    Lincoln, Y. S., & Guba, E. (1985). Naturalistic inquiry. Beverly Hills:Sage.

    MacCannell, D. (1973). Staged authenticity: arrangements of socialspace in tourist settings. American Sociological Review, 79, 589603.

    MacCannel, D. (1976). he tourist: A new theory of the leisure class. NewYork: Schocken Books.

    Markwell, K. W. (1997). Dimensions of photography in a nature-basedtour. Annals of ourism Research, 24, 131155.

    160 A. Decrop / Tourism Management 20 (1999) 157161

  • Marshall, C., & Rossman, G. B. (1995). Designing qualitative research.Thousand Oaks: Sage.

    Marti, B. E. (1995). Marketing aspects of consumer purchasingbehavior and customer satisfaction abroad the Royal Viking Queen.Journal of ravel and ourism Marketing, 4(4), 109116.

    Patton, M. Q. (1990). Qualitative evaluation and research methods.Newbury Park: Sage.

    Riley, R. (1995). Prestige worthy leisure travel behavior. Annals ofourism Research, 22(3), 630649.

    Riley, R. (1996). Revealing socially constructed knowledge throughquasi-structured interviews and grounded theory analysis. Journal ofravel and ourism Marketing, 5(1/2), 2140.

    Riley, R. W., & Love, L. L. (1997). he state of qualitative tourismresearch. Unpublished working paper.

    Rossman, G. B., & Wilson, B. L. (1985). Numbers and words: Combin-ing quantitative and qualitative methods in a single large-scale evalu-ation study. Evaluation Review, 9, 627643.

    Smith, V. L. (1977). Hosts and guests: he anthropology of tourism.Philadelphia, PA: University of Pennsylvania Press.

    Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Groundedtheory procedures and techniques. Newbury Park: Sage.

    Webb, E., Campbell, D. T., Schwartz, R. D., & Sechrest, L. (1996).nobtrusive measures; Non-reactive research in the social sciences.Chicago: Rand McNally.

    A. Decrop / Tourism Management 20 (1999) 157161 161