www.sjm06.com

Serbian JournalojManagement 5 (1) (2010) 175 - 185

Letter to Editor:

THE ROLE OF CASE STUDY METHODIN MANAGEMENT RESEARCH

SerbianJournal

ofMana·gement

Isidora Djuric, Borde Nikolic and Milovan Vukovic*

University ofBelgrade, Technical Faculty in Bar, VJ 12, 19210 Bar, Serbia

(Received 2 September 2009; accepted 12 January 2010)

Abstract

This paper examines some of the issues which arise from management research which developstheory from case studies. It raises some fundamental questions which arise when case material isused in management research. For example, confusion surrounds the distinctions between qualitativedata, inductive logic and case study research. Although every researcher has his/her preferredapproach, it is concluded that case studies may be built up in a number of ways: from, On the onehand, deep single case studies to multiple case studies using comparative logic, on the other. Betweenthese two extremes are a number of hybrid methods which use both approaches.

Keywords: Case study, management research, inductive logic, comparative logic.

1. INTRODUCTION

Case study is an ideal methodology whena holistic, in-depth investigation is needed(Feagin et aI., 1991). Case studies have beenused in various investigations, particularly insociological studies, but increasingly, ininstruction. Yin (1984), Stake (1995), andothers who have wide experience in thismethodology have developed robustprocedures.

Whether the study is experimental orquasi-experimental, the data collection andanalysis methods are known to hide some

* Corresponding author: mvukovic@tfbor.ac.rs

details (Stake 1995). Case studies, on theother hand, are designed to bring out thedetails fi:om the viewpointof the panicip~nts

by using multiple sources of data. The term"case study" is often used loosely, as animpressive but redundant synonym for"study". But the difference between a casestudy and a study is methodologicallysignificant. Indeed, the distinction helps toclarify the nature of comparative researchitself.

A case is an instance ofa. more generalcategory. To conduct a case study is thereforeto investigate something, which has

Case studies are a strategy for selecting atopic more than a technique for conductingresearch. In practice, they are normallymulti-method, using the range of techniquesin the managerial scientist's tool-kit: readingthe academic literature, examiningsecondary· documents (for example,newspapers), searching for primary material(for .example, .. lJnpubli$hed reports) andideally. .pOllducting .... interview withparticipaptS andoJherohservers. Scholars ofcases engage in "soaking and poking,marinating themselves in minutiae" (King etaI., 1994).

Case studies aim to provide a descriptionwhich is both rounded and detailed, a goalwhich anthropologist Clifford Greetz (1973)famously defined as "thick description".Because of this rounded character, casestudies are often contrasted with researchusing a single, systematic technique. Samplesurveys, fqi· iJistance,offer one particularform ·ofevidence ob. the topic. Compared tothe single slice of dataprovided by a survey,case students look through multiple lenses,mixing history and analysis, specific detailand wider implications, in an oftencompelling combination.

Achieving deep understanding in CSRusually involves the use of multiple researchmethods across multiple time periods(Denzin, 1978). Triangulation often includes:

1,\ ,.,

176 I.DuriO SJM 5 (1) (2010) 175 - 185

significance beyond its boundaries. For form not of relating cases to abstract theory,.---------.instanGG,--lawyer-s-s-tucly-eases-which-are-but-simply-ordrawiITg~a1falogles ·De-tween t~h~e-----'-----­

taken to define a legal principle with widec~sesth~;Inselves. In the absence ofapplicability; sociologists study particular overarching theory, case studies are thecommunities to cast light on general issues in building blocks from which we construct ourtheir discipline. A project turns into a case understanding of the managerial world.study only when it becomes clear what thestudy is a case of.

Case study research (CSR) .is· not 2. RESEARCH STEPS REQUIRED TOsampling research; that is a fact asserted by ACHIEVE DEEP UNDERSTANDINGall the major researchers in the field,

. including Yin, Stake, Feagin and others.However, selecting cases must be done so asto maximize what can be learned in theperiod of time available for the study.

In a case study, we seek to deepen ourunderstanding of process which has alreadybeen accepted within the discipline assignificant. Thus one practical advantage ofconducting a case study is that there is sureto be some interest in the fmdings. Comparedwith studies, case studies provide intellectualgearing, making a contribution to a widerdebate as well as offering a rounded accountof a particular subject. They offer a doublereturn on the research investment.

While case studies are sometimes treatedas week sibling in the family of researchstrategies, it would be pretentious to adoptsuch a lofty attitude. Management is anuntidy subject, in which cases arec-ontinually changing in path-dependentWayS, often influencing each other21'S theyevolve. By nature, the subject inatterofmanagement is data-rich and theory-poor;fotthis reason, cases are and will remain themajor route to understanding. Unlike, say,economics, we do not have a singletheoretical model to underpin our research.We must proceed by inspecting cases rather·than by making deductions from firstprinciples. In consequence, muchcomparative management analysis takes the

3.3. Deviant Cases

3.2. Prototypical Cases

TVDe DefinitionRepresentative Typical of the category

Prototypical Expected to become typical

Deviant An exception to the normCrucial Tests a theory in the least

favorable conditionsArchetvoal Creates the category

Deviant case studies are based on adifferent logic from both representative andprototypical designs. The purpose of adeviant case study is to cast light on the

3.1. Representative Cases

I.Duric/ SJM 5 (1) (2010) 175 -185 177

direct observation-by the researcher Table 1. Some Types ofCase Study (Haguewitlrintl1eenvU'onrrrenfsoflne-case; . et ai., 1998)

probing by askingcas'eparticipantsfor explanations and interpretations 'fbf"operational data" (Van Maanan, 1979); and

analyses of written documents andnatural sitesoccurtfug in case enVironmehts.

Operational data includes spontaneousconversations of participants in a case,activities engaged in and observed by theresearcher, and documents written by theparticipants. "Presentational data" are theappearances and answers to inquiries thatinformants strive to establish- and maintain:in the eyes of the fieldworker, outsiders andstrangers in general, work colleagues, closeand intimate associates, and to varyingdegrees, themselves (Van Maanan, 1979).

3. TYPES OF CASE STUDIES 'ttf.

The second type of case study is theprototypical form. Here, a topic is chosen notbecause it is representative but because itlsexpect¢d to ,become so: "their present is ourfuture" (Rose, 1991). Studying an earlyexample may help us to understand aphenomenon of growing significance. Thuswe will look closely to the new publicmanagement experience in Slovenia, sincethat country has traveled furthest down theroad of economic reform.

The danger with the prototypical case isThe first and most common form of case that it involves a bet on the future: what if the

study is the representative case - the study of prototype turns into a dud? Also, innovatorsa typical, standard example of a wider are by nature un-representatives; they·oftencategory (Table 1). This is the workhorse of posses unusual enthusiasm and experiencecase study designs, as useful as it is difficulties to those confronting theirundramatic. As Peters (1998) says "one very imitators. Yet by the same token, thevalid reason for doing a case study is to prototypical case study does offercollect information on the topic in question, opportunities for lesson;'drawing: laterespecially while the case is still in progress". adopters can learn from the mistakes of theExamining the managerial process wbile it is innovator. He~e the prototype still. exertsstill in progress may be especially valuable . influence but in a negative rather than ain comparative studies of~ for instance;· _positive way.~orporative social responsibility, ·given thatthe -researcher maybe less familiar with thenational setting of the case than he <>r shewould be for their own country.

J

178 /.DuriO SJM 5 (1) (20/0) /75 -/85

exceptional and the untypical. Deviant casesare often used to tidy up our understandingof exceptions and anomalies. Normalscience, suggests Khun (1970), proceeds inexactly this way, with researchers seeking toshow how apparent paradoxes can beresolves within a dominant intellectualtradition.

But deviant cases can also be ofconsiderable value in identifying underlyingcauses (Kazancigil, 1994). This is becausethey can provide the variation without whichwell-founded explanation is impossible. Ifwe want to ague that X causes Y, we mustcome up with cases ofnot.,.X and show thatthey not lead to not-Y. While deviant casesalways attract interest, the danger is that theybecome over-studied. The exceptional isalways more exotic than the typical.

X3.4. Crucial Cases

The crucial case study is sometimescommended, but less often used, in economyand management science. The idea here isthat if a proposition can be shown to workwhen conditions are least favorableforjtsvahdity; .it is likel~· to. be valid in all' othercirc\1Illstl,IDces.as well. If liberal economies,let us say, are now consolidating in countries(such as the post-communistic countries ofEastern and Central Europe) which have noprevious experience of that form ofeconomy, we can be sure that the modemmove toward liberal economy is significant.Alternatively, a proposition which .fails towork even in the most favorable conditionscan quickly be dismissed.

Similarly, if post-material values arenowhere to be found among graduates in thewealthiest countries, then the theory of post­materialism is no good. Depending onexpectations, we can set out either to support

a theory by showing its value in unfavorableconditions (a "lcast favorable" dt:sign) 01' todisprove a theory be showing it fails even infavorable circumstances (a most favorabledesign).

Eckstein (1975) is a strong advocate ofcrucial case studies. He suggests that "asingle crucial case may certainly score aclean knockout over a theory". But his thesiSis limited in three ways. First, common sensesuggests we should not place too muchweight on anyone case, no matter howcrucial it appears to be. Second, the crucialcase design assumes a narrow theory-testingrole for case studies, which, as we havesuggested, is often inappropriate for.comparative politics. Third, crucial casestudies involve a risky bet on the results. Ifthe fmdings run counter to expectations, wehave learned nothing at all of significancebeyond the case.

3.5. Archetypal Cases

The idea here is thata c.ase.generat&sthe .category of which it is then taken, insomewhat misleading way, as representative.Take the Industrial Revolution. This episodealtered the whole concept of technologicaldevelopment, reconstructing the' idea asprogressive, modernizing force. In this way,the Industrial Revolution made possible 'allthe technological revolutions whichfollowed.

The various uses and types of case studiesare also discussed by Lijphart. Lijphart'sterminology for identifying the various typesis similar to other widely accepted notionssuch as Ekcstein's, with two exceptions:Lijphart does not designate a separatecategory for Eckstein's 'plausibility probe';and Lijphart adds a quite important type ofcase study, the analysis of the 'deviant' case,

I.DuriO SJM 5 (l) (20/0) /75 - 185 179

for which Eckstein does not make explicitprovision. The similarities and ditlerencesbetween these two listings of types of casestudies are illustrated in Table 2:

Table 2. Additional Typologies ofCaseStudies: Lijphart vs. Eckstein (George &Alexander, 1979).

Lijphart Eckstein

atheoretical case studyconfigurative-

idiographic

interpretative case studydisciplined-

configurativehypothesis-generating

heuristiccase study

" plausibility probe'theory confinning' case

studycrucial case

'theory infinning' casestudy

deviant case study ?

4. THE PRECEPT OF CASE ANALYSIS

Social science methodology is anchoredby a number of basic precepts that are rarelyquestioned by practitioners. One precept thatis central to the logic of analysis is the ideaof having cases. Social scientists use tennslike "N of cases", "case study," and "sampleof cases" with relatively little considerationof the possihle theories and meta-theoriesembedded in these tenns or in the methodsthat use cases and inake conventional fonnsof analysis possible. For example, a studythat uses interviews of employees toconstruct a picture of the infonnalorganization ofa finn looks superficially likeone that uses interviews of employees toaddress variation in job satisfaction. Bothstudies use interviews of employees as theprimary data source, but the first is about thefinn as a whole, while the second is aboutemployees' subjective states. It is argued here

that the tenn "case" and the various tennslinked to the idea of case analysis are notwell defined in social science, despite theirwidespread usage and their centrality tosocial scientific discourse.

To the question "What is a case?" mostsocial scientists would have t6 give multipleanswers (Ragin & Becker, 1992). A case maybe theoretical or empirical or both; it may bea relatively bounded object or a process; andit may be generic and universal or specific insome way. Asking "What is a case?"questions many different aspects ofempirical social science.

4.1. The Comparativ Method - CaseStudy with other Method.~

The discussion about the tenn "case"presented in this paper had its origins inother works. The peculiar status ofthe "case"was clear in work The Comparative Method(Ragin, 1987). In that work Ragin showedhow conventional variable-orientedcomparative work (e.g., quantitative cr6ss,.national research), as compared with case- I\/f/oriented comparative work, disembodies and , 'Iobscures cases. In most variable-orientedwork, investigators begin by de.fhrltigth~ / 9\

pfobletnirta way that allows examination of ';Y 2/~ ..many cases (conceived as empirical units or ~~~observations); then they specify the relevant@' WQ~ '"variables, matched to theoreticalconbepts; ~~Jand, finally, they gather information on thesevariables, usually one variable at a time - notone case at a time. From that point on, thelanguage of variables and the relationsamong them dominate the research process.The resulting understanding of theserelations is shaped by examining patterns ofcovariation in the data set, observed andaveraged across many cases, not by studyinghow different features or causes fit together

180 I.DuriO SJM 5 (1) (2010) 175 - 185

Understanding Case conceptioncases Specific General

As empirical 1. Cases are 2. Cases areunits found objectslAs theoretical 3. Cases are 4. Cases areconstructs made conventions

The first dichotomy (whether the questionof cases involves empirical units ortheoretical categories) is common indiscussions of social science methodologyand overlaps with the philosophicaldistinction between realism andnomt'nalt'sm. Realists believe that there arecases (more or less empirically verifiable assuch) "out there". According to nominalists,cases are theoretical constructs that existprimarily to serve the interests ofinvestigators. A realist sees cases as eithergiven or empirically discoverable. Anominalist sees cases as the consequences oftheories or of conventions.

The second dichotomy concerns thegenerality of case categories. Are casedesignations specific (e.g the "authoritarianpersonality") and developed in the course ofresearch (e.g., through in-depth interviews orhistorical research) or are they general (e.g.,individuals, families, cities, firms) andrelatively external to the conduct ofresearch? In many areas of research, genericunits are conventionally treated as cases, andcase categories are neither found nor derivedin the course of research. They exist prior toresearch and are collectively recognized asvalid units by at least a subset of socialscientists.

Specific case categories, by contest,emerge or are delineated in the course of theresearch itself. What the research subject is a"case of' may not be known until after most

in individual cases. Table 3. Conceptual map for answers to-------------''Fhe~alternative,case.,{)rienty.chapproac}r-c-ItWhat··is caTase?'1

focuses on cases, not variables. However,what is a case? Comparative social sciencehas a ready-made, conventionalized answerto this question: Boundaries around placesand time periods define cases (e.g., Serbiaafter the fall of Communism). Incomparative and historical social science,there is a long tradition of studyingindividual countries or sets of theoreticallyor empirically related countries conceived ascomparable cases. The conventionalizednature of the answer in macro social inquirymade it simple to contrast variable-orientedand.~ase-{)riented approaches. It could just aseasily be argued, however, that not countriesbut rather parallel and contrasting eventsequences are cases, or that generic macrosocial processes, or historical outcomes, ormacro-level narratives are cases. "What is acase?" is problematic even where researchersare confronted at every tum by big, enduring,formally constituted macro-social units suchas countries.

The problem of "What is a case?" is evenmore crucial when the contrast betweenvariable-oriented and case-orientedapproaches is transferred to other researchdomains, because in most research areas theanswers are less conventionalized. Is a socialclass a case or a variable? This is not atrivial question for scholars interested insocial movements and the future ofinequality.

There are different approaches to answerthe question "what is a case?". To understandthem, consider two key dichotomies in howcases are conceived (Table 3):

whether they are seen as involvingempirical units or theoretical constructs, and

whether these, in turn, are understoodas general or specific.

I.Durie / SJM 5 (1) (2010) 175 -- 185 181

of the empirical part of the project is designations on existing definitions present-c6mpleted:~T(ra''lifiiited~exfent,tIiissecona~~~iri''=~researen-literatures.· 'A~=re'searcIler ...dichotomy overlaps with the qualitative- interested in explaining contemporaryquantitative divide in social science. The international inequality, for example, wouldcases of quantitative research tend to exist as accept nation-states (as conventionallyconventionalized, generic categories defmed) as appropriate cases for his or herindepeIld~nt ofany particular research effort. analysis. Often coupled with this view is anThe cas~sof qualitativete-seai"ch' tend to 'instium:ental attitude toward cases - they {>f~,.;-j..--'coaJ,esceasspecific categories in the course exist to be manipulated by investigators. I l'-~Cof.the-research. Cell 3: Cases are made. Researchers in .

The cross-tabulation of these two this quadrant see cases as specific theoreticaldichotomies (Table 3) yields four possible constructs which coalesce in the course ofstarting points for answering the question the research. Neither empirical nor given,I) ~ "What is a case?". Consider the nature of they are gradually imposed on empiricalr"cases" from the perspective of each cell of evidence as they take shape in the course ofthe cross-tabulation the research. A cell-3 investigator interested

J ,1, Cell 1: Cases are found. In the first in leadership, for example, would study/Jf-~ C, quadrant, researchers see cases as many possible instances of leadership. This

empirically real and bounded, but specific. investigation might lead to an identificationThey must be identified and established as of an important subset of instances withcases in the course of the research process. A many common characteristics, which mightresearcher may believe that "world systems" be conceived, in tum, as cases of the same(networks of interacting and inter~ep~~clent. thing (e.g., as cases of "authoritarianhuman societies) are fundameritally leadership" or as cases of "democraticimportant empirical units for un<fep:;tancling leadership"). Interaction between ideas andthe history of human social organization and evidence results in a progressive refinementtherefore may seek to determine the of the case conceived as a theoreticalempirical boundaries of various historical construct. At the start of the research, it mayworld systems (verifiable, e.g., through notbe at all clear that a case can or will beevidence of trade in bulk goods between discerned. Constructing cases does not entailpeoples of differing cultures). Researchers determining their empirical limits, as in cellwho approach cases in this way see I, but rather pinpointing and then~..assessment of the empirical bounding of demonstrating their theoretical significance. Ccases as an integral part of the research Cell 4: Cases are conventions. Finally, inprocess. the fourth quadrant, researchers see cases as

Cell 2: Cases are objects. In the second general theoretical' constructs, butquadrant, researchers also view cases as nevertheless view these constructions as theempirically real and bounded, but feel no products of collective .scholarly work andneed to verify their existence or establish interaction and therefore as external to anytheir empirical boundaries in the course of particular research effort. A researcher, forthe research process, because cases are example, might conduct research ongeneral and conventionalized. These "industrial societies," recognizing that theresearchers usually base their case assignment of empirical cases to this

182 I.DuriO SJM 5 (1) (2010) 175 - 185

theoretical category is problematic-and that definitions and practices (cell 4) could lead--the-theor~tiealeategotyitselfexists-primarily researchers~--to~----intensifr-tlr~fr~-'-~mp'irjcal"

because of collective scholarly interest. In efforts and to define cases and theirthis view, cases are general theoretical boUndaries in a more inductive manner (cellconstructs that structure ways of seeing I). In fact, most research involves mu;ltip'l~

social life and doing social science. They are useS of cases, as specific or generalthe collective products of the social scientific theoretical categories and as specific orcommunity and thus shape and constrain the general empirical units. These mtiltiple usespractice of social science. occur hecause research combines theoretical

This fourfold division of case conceptions and empirical analysis, and the two kinds ofis not absolute. A researcher could both use - analyses need not use parallel cases or units.conventionalized empirical units,_~cceptij}g The point of Table 4 is not to establishthem as empirically yalid (cell 2), and try to boundaries between different kinds ofgenerate n~w theoretical Gategones ,or case research, but to establish a conceptua,rmapcpnstructs (cell 3) .in the course of his or her for linking different approaches to theresearch. Frustrations with conventional case question of cases.

Table 4. Situating the Comparative Method as oj1971: Lijphart's Scheme (Collier, 1971)

Case Study Method Comparative Method Experimental Statistical MethodMethod

Merit: Defined as: Merit: Merit:Permits intensive Systematic analysis ofsmall Eliminates rival Assesses rivalexamination of cases number'of cases ("small-1'(" explanation through explanations througheven with limited analysis) experimental control statistical controlresources

Merit: lIiherent Problem: Inherent Problem:Inherent Problem: "Given inevitable scan;:ity of Experimental control Difficult to collectContributes less to time, energy, and financial is impossible for adequate informationbuilding theory than resources, the intensive many or most topics in a sufficient numberstudies with more analysis of a few cases may relevance to field of of cases, due to limitedcases be more promising than the comparative politics time and resources

Types of Casesuperficial statistical analysis

Studies:of many cases" (Lijphart,

1. Theoretical1971, p. 685)

2. Interpretive Inherent Problem:3. Hypothesis- Weak capacity to sort ourgenerating trivial explanations,4. Theory-confirming specifically, the problem of:5. Theory-infirming "many variables, few cases"(i.e., case studies that

Potential Solutions:weaken a theoryI. Increase number of casesmarginally)2. Focus on comparable cases6. Deviant case

studies 3. Reduce number ofvariables:

a. Combine Variablesb. Employ more

parsimonious theory

I.DuriO SJM 5 (I) (2010) 175 -185 183

4.2. Synopsis of Lijpbart analysis. ''Fheintensivecomparativeanalysis----~- ofaTew-c:astSsrriaYbemoie~promIsin~-rliaii-a

Lijpart defmed the comparative method asthe analysis of a small number of cases,entailing at least two observations, but lessthan about twenty. The central goal of hispaper is to assess the comparative method inrelation to the experimental, statistical, andcase study approaches. He evaluates thesedifferent approaches by two criteria: (1) howwell they achieve the goal of testing theorythrough adjudicating among rivalexplanations, and (2) how difficult it is toacquire the data needed to employ eachmethod (Table 4).

The case study method has the merit ofallowing the scholar with relatively modesttime and resources to assess at least one casewith care, but the opportunities forsystematically testing hypotheses are farmore limited than with the other methods.Yet, case studies do make a contribution, andLijphart offers a suggestive typology of thedifferent ways case studies can be used informing and testing theories.

The comparative method, as defmed byLijphart, has an intermediate status on bothofhis criteria. It provides a weaker basis thathe experimental or statistical method forevaluating hypotheses, specifically becauseof the many variables, small~N problem. Yetit offers a stronger basis for evaluatinghypotheses than do case studies. Even withthe problem of having-more variables thancases, the comparative method allowssystematic comparison which, ifappropriately utilized,can contribute to theassessment of alternative explanations.

Lijphart therefore views the comparativemethod as suitable in research based onmodest resources, and he suggests thatstudies using the comparative method mightoften serve as a first step toward statistical

more superficia1 statistiCal analysis of manycases.

With regard to the large number ofvariables, he suggests two approaches: First,analysts can focus on "comparable cases",that is, on cases that (are matched on manyvariables than are not central to the _study,thus in effect "controlling" - for thesevariables, and differ in terms of the keyvariables that are the focus of analysis,thereby allowing a more adequateassessment of their influence. Hence, theselection of cases acts as a partial substitutefor statistical or experimental control.Second, analysts can reduce th~ number ofvariablesejther by combining variables orthrough theoretical parsimony, that is,through the careful elaboration of a theorythat focuses on a small number ofexplanatory factors (Collier, 1971).

5. CONCLUSION

The case study methodology has beensubjected to scrutiny and criticism at varioustimes since the 1930's. As a research tool, ithas not been a choice that is listed in themajor research texts in thc social sciences.However; case study is a reliablemethodology when executed with due care.The literature, while not extensive, containsspecific guidelines for researchers to followin carrying out case studies.

CSR _ is focused on describing,understanding, predicting, and controllingthe individual (i.e. process, animal, person,household, organization, group, industry,culture, or nationality). Anyone orcombination of the following purposes mayserve as the major objective of CSR:

184 I.DuriO SJM 5 (1) (2010) 175 - 185

description, ..Ulltf,frsta.mlillg, prediction, or","",co.1Jtrol~,However,·in'most"',situations,'~deep

understanding of the actors, interactions,sentiments, and behaviors occurring forspecific processes through time should betaken as the principal objective by the casestudy researcher.

A mental model of a process provided bya participant interviewed in a case study is anlfll#C r~presep.tatiQIl of reality. Theinterpretation of same process provided bythe case study researcher isaneticrepresentation of reality. Etic representationin CSR often includes description andexplanation of ernic meaning as well asbuilding composite accounts of the processbased on data from triangulation.Triat};gulatioll includes: direct observation bythe re.~earcherwithinthe enviromne.n,ts of thecase, and, probing by askiIlg caseparticipants for explanations and

interpretations of operational <.lata, andanalyses' 'ofwritten"documents"~and natural····siteso~c.urriIlg i,Ilcase enviroPIIlents.

References

Bargh, J. (2002) Losing consciousness:automatic influences on consumer judgment,behavior, and motivation, Journal ofConsumer Research, (29): 280-285.

Bargh, J. A., Chen, M., & Burrows, L.(1996) Automaticity of social behavior:Direct effects of trait construct andstereotype activation on action. Journal ofPersonality and Social Psychology, 71 (2):230-244.

Collier, D. (1971). The ComparativeMethod: Two Decades of Change. Pp. 7-31in Rustow & Erickson (Eds), ComparativePolitical Dynamics. New York: HarperCollins.

.YJIOrA METO,[(E CTY,[(lIJE CJIYqAJA YMEHAI,IMEHT lICTPAlKlIBAIbY

*HcuAopa nypuli, noplie HUKOJluli U MUJlOBaH ByKOBHIi

YHtl6ep3Umem y Seozpaoy, TexHUtlKU paKyJlmem y Sopy, BJ 12,19210 Sop, Cp6uja

II3BO~

OBaj pa~ npoylJaBa HeKe o~ eJIeMeHaTa KOjH npoHCTIfqy H3 MeHaI,IMeHT HCTpa)KIIBaH>a ca QH.JbeM

pa3Boja TeopHje nyreM c~Hja cnyqaja. OB~e ce nOKpeliy o~peijeHa 4>YH~aMeHTaJIHanHTaH>a Koja

npOHCTHlJy H3 ynoTpe6e C~HjCKOf MaTepHjana TOKOM MeHaI,IMeHT HCTpa:>KHBaH>a. Ha npHMep,

KOH4>y3Hja Koja OKpy:>Kyje pa3JIHKOBalbe KBaJIHTaTHBHHX no~aTaKa, HH~YKTHBHe JIOfHKe H

HCTpa:>KHBaH>a no npHHQHny cry~Hje cJIyqaja. liIaKO CBaKH HCTpa:>KHBalJ HMa cBoj npHopHTeTHH

npHcryn, ~OJIa3H ce ~o 3aK.JbyqKa ,na ce c~Hje cnyqaja MOry CalJHHHTH Ha BHIIIe HatJHHa: o,n, C

je,nHecTpaHe, Iloje,nHHalJHHX cry,nHja ,no cry,nHje BHIIIecTpyKHX cJIylJajeBa Y3 ynoTpe6y

KOMnapaTHBHe JIOrHKe, C ,npyre CTpaHe. 113Meijy OBa ,nBa eKCTpeMHa cnyqaja nocTojH BeJIHKH 6poj

xH6pH,nHHX MeTo,na Koje KopHCTe o6a npHcryna.

IGoy'me pe'lu: C~Hje cJIyqaja, MeHaI,IMeHT HCTpa:>KHBaH>a, HHJlYKTHBHa JIOrHKa, KOMnapaTHBHaJIOfHKa.

I.Durie / SJM 5 (I) (2010) 175 - 185 185

Denzin, N.K. (1978).The Research Act, and the .~oD;lparative.Metl1od.Am~ricflQ.

McGraw;;Hill~NewYork,NY----PoliilcafSClence -Revi-ew~--o65: 68i~693~-- ----Eckstein, Harry (1975). Case study and Ragin, C. (1987). The Comparative

Theory in Political Science. Pp. 127 in F. Method: Moveing Beyond Qualitative andGreenstein & N.W.Polsby ( Eds.), Handbook Quantitative Strategies. Californiaof Political Science, (7). Reading, MA: University Press.Addison-Wesley. Ragin, c., & Becker, H. (1992). "What is

Feagin, J., arum, A., & Sjoberg, G. A Case?", Exploring the Foundation of(1991). A case for case study. Chapel Hill, Social Inquiry. Cambridge University Press.NC: University of North Carolina Press. Rose, R. (1991) Comparing Forms of

George, A. (1979). Case Studies and Comparative Analysis. Political Studies, 39:Theory Development: The Method of 446-462.Structured Focused Comparsion. Pp. 26 in Stake, R. (1995). The art of case research.Lauren Paul (Ed), Dipolmacy: New Newbury Park, CA: Sage Publications.Approaches in History, Theory, and Policy. VafiM,&li1la,n, L(~~79) The fact offjctionThe Free Press, New York. in organizational ethnography.

Giampietro, G., Jaber F. G., & Silverman, Administrative Science Quarterly, 24: 539­D. (2004). Qualitative Research Practice. - 5,50~

London and Thousand Oaks, CA: Sage. Weick, K.E. (1995). Sensemaking inGreetz, Clifford (1973), Thick Organizations. Sage, Thousand Oaks, CA.,

Description: Toward an Interpretive Theory Woodside, A.G., & Wilson, E.J. (2002)of Culture.In The Interpretation of Cultures, Respondent inaccuracy. Journal ofed. Clifford Greetz. New York: Basic Books. Advertising Research 42: 7-18.

Hague; Rod, Harrop Martin and Breslin Woodside, A.G., & Wilson, E.J.(2000)ShaUll (199&), COIQ.parative Government and Constructing Thick descriptions ofmarketersPolitics: An Introduction, fourth edition, 'and buyers' decision processes in business­Macmillan. to-business relationships. Journal of

Kazancigil, A. (1994). The Deviant Case Business & Industrial Marketing, 15 (5):in Comparative Analysis: High Stateness in 354-369.Comparative Analsysis. Pp. 213-238 in M. Yin, R. (1984). Case study research:Dogan & A. Kazancigil (Eds.), Comparing Design and methods (1st ed.). Beverly Hills,Nations: Concepts, Strategies, Substance in, CA: Sage Publishing.Cambridge, Mass and Oxford. Blackwell http://www.emeraldinsight.com/researchr

King, G., Sidney V, & Keohane, R.O. egister.(1992). Scientific Inference in QualitativeResearch. Unpublished manuscript,Department of -Government, HarvardUniversity.

Levy, S. (1988). Information technologiesin universities: An institutional case study.Unpublished doctoral dissertation, NorthernArizona University, Flagstaff.

Lijphart, A. (1971). Comparative Politics