Building Theories from Case Study Research

BuildingTheoriesfromCaseStudyResearchAuthor(s):KathleenM.EisenhardtSource:TheAcademyofManagementReview,Vol.14,No.4(Oct.,1989),pp.532-550Publishedby:AcademyofManagementStableURL::16/03/201011:38YouruseoftheJSTORarchiveindicatesyouracceptanceofJSTOR'sTermsandConditionsofUse,availableat://=aom.EachcopyofanypartofaJSTORtransmissionmustcontainthesamecopyrightnoticethatappearsonthescreenorprintedpageofsuchtransmission.JSTORisanot-for-profitservicethathelpsscholars,researchers,andstudentsdiscover,use,andbuilduponawiderangeofcontentinatrusteddigitalarchive.Weuseinformationtechnologyandtoolstoincreaseproductivityandfacilitatenewformsofscholarship.FormoreinformationaboutJSTOR,[e-mail]laboratingwithJSTORtodigitize,preserveandextendaccesstoTheAcademyofManagementReview.(e.g.,parsimony,logicalcoherence),andconvincinggroundingintheevidencearethekeycriteriaforevaluatingthistypeofresearch.Developmentoftheoryisacentralactivityinorganizationalresearch.Traditionally,authorshavedevelopedtheorybycombiningobserva-tionsfrompreviousliterature,commonsense,andexperience.However,thetietoactualdatahasoftenbeentenuous(Perrow,1986;Pfeffer,1982).Yet,asGlaserandStrauss(1967)argue,itistheintimateconnectionwithempiricalrealitythatpermitsthedevelopmentofatestable,rel-evant,andvalidtheory.Thispaperdescribesbuildingtheoriesfromcasestudies.Severalaspectsofthisprocessarediscussedintheliterature.Forexample,GlaserandStrauss(1967)detailedacomparativemethodfordevelopinggroundedtheory,Yin(1981,1984)describedthedesignofcasestudyresearch,andMilesandHuberman(1984)codi-fiedaseriesofproceduresforanalyzingquali-tativedata.However,confusionsurroundsthedistinctionsamongqualitativedata,inductivelogic,andcasestudyresearch.Also,thereisalackofclarityabouttheprocessofactuallybuildingtheoryfromcases,especiallyregard-ingthecentralinductiveprocessandtheroleofliterature.GlaserandStrauss(1967)andmorerecentlyStrauss(1987)haveoutlinedpiecesoftheprocess,buttheirsisaprescribedformula,andnewideashaveemergedfrommethodolo-gists(e.g.,Yin,1984;Miles&Huberman,1984)andresearchersconductingthistypeofre-search(e.g.,Gersick,1988;Harris&Sutton,1986;Eisenhardt&Bourgeois,1988).Also,itap-pearsthatnoonehasexplicitlyexaminedwhenthistheory-buildingapproachislikelytobefruitfulandwhatitsstrengthsandweaknessesmaybe.Thispaperattemptstomaketwocontributionstotheliterature.Thefirstisaroadmapforbuild-ingtheoriesfromcasestudyresearch.Thisroadmapsynthesizespreviousworkonqualita-tivemethods(e.g.,Miles&Huberman,1984),thedesignofcasestudyresearch(e.g.,Yin,1981,5321984),andgroundedtheorybuilding(e.g.,Gla-ser&Strauss,1967)andextendsthatworkinareassuchasapriorispecificationofconstructs,triangulationofmultipleinvestigators,within-caseandcross-caseanalyses,andtheroleofexistingliterature.Theresultisamorenearlycompleteroadmapforexecutingthistypeofre-searchthanhasexistedinthepast.Thisframe-workissummarizedinTable1.Thesecondcontributionispositioningtheorybuildingfromcasestudiesintothelargercontextofsocialscienceresearch.Forexample,thepa-perexploresstrengthsandweaknessesoftheorybuildingfromcasestudies,situationsinwhichitTable1ProcessofBuildingTheoryfromCaseStudyResearchStepActivityReasonGettingStartedDefinitionofresearchquestionFocuseseffortsPossiblyaprioriconstructsProvidesbettergroundingofconstructmeasuresNeithertheorynorhypothesesRetainstheoreticalflexibilitySelectingCasesSpecifiedpopulationConstrainsextraneousvariationandsharpensexternalvalidityTheoretical,notrandom,samplingFocuseseffortsontheoreticallyusefulcases-i.e.,thosethatreplicateorextendtheorybyfillingconceptualcategoriesCraftingInstrumentsMultipledatacollectionmethodsStrengthensgroundingoftheorybyandProtocolstriangulationofevidenceQualitativeandquantitativedatacombinedSynergisticviewofevidenceMultipleinvestigatorsFostersdivergentperspectiv