Research Challenges in Cross-Cultural International Business Management

8/12/2019 Research Challenges in Cross-Cultural International Business Management

1/6

AdvancesInM anagement Vol.3 (4)April (2010)Case Study:Research ChallengesinCross-cultural In ternational BusinessM anagement: The IssueofCultural Construct Equ ivalency

Scroggins Westey A.'*, Rozett Etizabeth J.', Guo A im in \ S ebestov Jarmila^ and Velo Veronica1. Departmen tof Managemeni. Mis.souri Stale University, 901S. National Avenue. Springfield. MO 65897,USA2. Departmenlof Business Administration. Henan UniversityofFinanceand Economics (HUFE),80 Wenhua Road, 2 iengzhou, Henan. CHINA

3. DepartmentofManagement andBusiness, Silesian University inOpava, Univerzi tni nm. 1934/3, 73340 Karvin, CZECH R EPUB LIC4. Management andOrganization Depar tment ,ESCRennes SchoolofBusiness,2 rue Robert D'Arbrissel CS 76522, 35065 Rennes Cedex, FRANCE* wesscroggins@missour istate.edu

AbstractAs business becomes more global in nature

management researchers are increasingly interested inconducting cross-cultural research. Although cross-cultural management research has provided us with anunderstanding of management practice in variouscultures the nature of the research presents researcherswith numerous challenges. One such challenge is theissue of cultural construct equivalency. Culturalconstruct equivalencyconcert .> the validity of measuresacross cultures. Oftentimes construct equivalency isassumed to exist in the measures used. This paperpresents initial research data from the EWORLDProject an international entrepreneurship researchproject. The data are used to demonstrate the constructequivalency problem and to show that methodscommonly employed by researchers do not necessarilysolve the problem.

Keywords: International busines.s, Cross-cultural research,Entrepreneurship.IntroductionCross-cultural management research presentsnumerous challengesand difficulties. Oneofthese challengesistheissueof cultural construct equivalency. Constructsaremental categories or models that represent phenomenonandhelp usexplain thethings thatweexperience. Constructsarepsychological in nature and individuals develop constructs,schmas and schema prototypes for all objects, entitiesandissues of which they are cognizant. The management andorganizational sciences are cofnplete with constructs ofinterest to researchers that representor help explain behaviorin work organizations. International m anagement researchhasoften focused on cultural constructs such as those identifiedby Hofstede''* and later used by House, Hanges, Javidan,Dorfman and Gupta^. Hofstede'** identified the culturalconstructs of individualism, power distance, risk tolerance,masculinity and long-term orientation that have been usedasa model to exatnine cultures. House et al have used theseconstructs and others to investigate the construct ofleadership in a cross-cultural context. Other constructs ofinterest to international management researchers include

employee perceptions andattitudes toward v arious personnelselection tests and staffing a cti vi tie s' ' '^ '^ There has alsobeen an interest in examining various en trepreneurialconstructs as managem ent researchers become m oteinterestedinresearching international entrepieneu rship6 1 1 1 8While research involving constructs is common ininternational business research, at issue is whether these

constructs of interest are universal or equivalent in natureacross cultures. Central to this issue is the issue of theconstruct validityof the measures employed.If the natureofaconstruct changes across cultures, then it isquestionableas towhether ameasureof that construct developed in onecultureis a construct valid measure of the same construct in adifferent culture. This is the issue of cultural constructequivalency. C ultural co nstruct, equivalency exists whenindividuals in one cultural understand a research item inexactly the same manner as individuals in another culture'^.Management researchers conducting international cross-cultural research often assume construct equivalency'. Theassumption is that theories, constructs and even researchmethodologies are etic in nature, being universallyapplicable. This ignores the emic nature of many researchconstructs, that many constructs may bespecific oruniquetoa local culture and understood differently in that culturedueto differing cultural dynamics.Conceptual ndFunctional Construct Equivalence:Twobasic dimensions of construct equivalence have beenidentified'*. Conceptual equivalence exists when constructshave similarity of meaning and structure across culturesandindividualsinthose cultures groupandcategorize informationin similar w ays. Douglas andCraig''use abicycletoillustratethese forms ofequivalency.Theconstruct of a bicycle likelyhas conceptual equivalency across cultures. In most cultures,individuals are likely to conceptualize the structure of abicycle similarly and will distinguish and categorize itseparately from other structures with wheels, such as atricycle or rriotorcycle, However, the cross-culturalconceptual equivalency of constructs more psychological innature, suchasleadership, is more questionable. Houseet alhave provided supportfor thelackof concepttial equivalencyof the leadership construct across cultures. People in differentcultures have very different ideasof what leadershipis.

53


2/6

Advances In Management Vol. 3 (4) April 2010)The bicycle illustration provided by Douglas andCraig* is a good example of how a construct can lackfunctional equiva lency a cross cultures In the United States, abicycle is used recreationally. In China, however, the bicycleis still used by many as a major source of transportation. Thebicycle has a different function or purpose in the two cultures.Therefore, although Americans and Chinese may sharesimilar concepta structures of a bicycle, the bicycle is likelyunderstood differently in the two cultures due to its differentfunctions. Internationa managem ent researchers mustaddress the issues of conceptual and functional constructequivalency when measuring constructs cross-culturally.They need to provide evidence that the constructs are similaror identical across cultures, because construct equivalency isnecessary for the construct validity of measures whenidentical measures are used in the different cultures. If theconstructs- differ either co nceptually or functionally, it ishighly unlikely that the same measure will be a constructvalid measure of the construct in each culture. Therefore,results obtained in such a cross-cultural study is suspectwithout evidence that construct equivalency exists.

Emic and Etic Uses of Tests in Research: The constructequivalency problem often arises when measures validatedand used in an emic context are used to measure the sameconstruct in another cultural context. An emic approach totest development is one in which a measure is developed andvalidated to measure a construct in a specific culture. Themeasure is culture specific, based on data obtained fromindividuals in the one culture only. Then, this measure is usedin an etic fashion. Etic refers to totality, or a researchapproach in which measures are used universally as if theywere culture-free. The measure is used to measure the sameconstruct in another culture. Researchers often assumeconstruct equivalency and, therefore, the construct validity ofthe measure across cultures. Scales developed and validatedby management researchers in the United States are oftenused in this etic fashion, although they are developedemically. Measures of leadership, commitment and job fit areexamples of measures developed emically in the UnitedStates and then used in an etic manner. When measures aredeveloped emically and then applied etically, the constructvalidity of the measure cross-culturally is suspect.Oftentimes, the attem pts; to show evide nce of

construct equivalency in research go no further than thetranslation and back-translation of research tools into thelanguages of the participants involved. For example,researchers may create an English version of a questionnairefor use to measure certain constructs. They may translate thatquestionnaire into Chinese to use to collect data in China.Then, the questionnaire will likely be back-translated fromChinese into English and the two English versions compared.If the translations are the same, construct equivalency isassumed and the researchers assume that they are usingidentical measures across the different cultures. However,translation and back-translation do not guarantee either

conceptual or functional construct equivalency andquestionnaires used in cross-cultural research that have onlybeen subjected to this process will often lack constructequivalency and validity^. The result is a "pseudo etic"approach in which the questionnaire is wrongly assumed tobe measuring the same thing across cultures. Cross-culturalcomparisons made and conclusions drawn from data obtainedfrom the questionnaire will likely lack validity and bemeaningless. In this study, issues of semantics and changes inthe construct of interest across cultures contributed to theconstruct equivalency issue. Translation and back-translationitself was not sufficient to resolve these issues. Researchersmust take additional steps beyond the simple translation andback-translation of questionnaires to establish the constructvalidity of a measure cross-culturally.Evidence from the E\yORLD Project: The EWORLDProject, an international entrepreneurship research projectbeing conducted in approximately 40 countries takes animplicit/attribution theory approach to the study ofcomparative cross-cultural entrepreneurship. It is argued thatcultural factors^'' affect the perceptions and attributions madeof entrepreneurs in a specific country and contribute to thdevelopment of implicit cultural entrepreneurial prototypesacross countries. Different entrepreneurial prototypes willexist in countries based upon specific cultural factors anddynamics. It is important for entrepreneurs in a given cultureto match the prototype of the successful entrepreneur for thatculture.

The theories that guide the advancement of theEWORLD theoretical framework are an assimilation ofimplicit leadership theory'" and the value-belief theory ofculture'". Additionally, the model relies heavily on the workof House et al' in the theoretical model used in the GLOBEproject. Implicit leadership theory purports that individualshave implicit beliefs, convictions and assumptions concerningattributes and behaviours that differentiate leaders fromsubordinates and effective leaders from non-effective ones.The beliefs and assumptions are called the implicit leadershiptheory. The EWORLD research applies this same concept fothe entrepreneurship area. In essence, it is proposed thatindividuals have implicit beliefs about entrepreneurs as well.That is, entrepreneurial qualities are attributed to individualsand hence, those same individuals are accepted as successfulentrepreneurs. These qualities or implicit entrepreneurshiptheories (IET) influence the actions and effectiveness ofentrepreneurs.

The goal of the EWORLD Project is to identify theimplicit cultural entrepreneurial prototypes that exist indifferent cultures. More specifically, the research aims toidentify the content and structure of cultural perceptions ofthe entrepreneur, or the psychology of entrepreneurship indifferent cultures. This is a conceptual construct equiyalencyissue. Entrepreneurship is to some degree an attribution basedon the degree to which the individual fits/matches the cultural54


3/6

dvances In Management Vol. 3 (4) April 2010)entrepreneurial prototype. It is the implicit culturalentrepreneurial prototype that is the construct of interest inthis research and that is used here to demonstrate theconstruct equivalency problem and challenges faced byinternational researchers. Data are presented from the CzechRepublic, France and China to demonstrate the constructequivalency issue in this paper.MethodologyParticipants: Participants completed a 115 itemquestionnaire developed to assess their perceptions of theentrepreneur in their respective country culture. Sample sizesfor each country were small at this point and the analysesconducted and results reported here are preliminary ande.xploratory. Participants were entrepreneurs or employees ofentrepreneurs in entrepreneurial firms. In France, students inentrepreneurship degree programs were also included in thesample. The sample sizes in the Czech Republic, China andFrance were N = 36, 100 and 63 respectively.Procedure: A questionnaire was developed to measureperceptions of the entrepreneur (entrepreneurial prototypes)in the countries participating in the EWORLD Project. Astandard classical test theory approach was followed in thequestionnaire development process^^''''^ The process beganwith conducting entrepreneurial literature reviews and focusgroups in numerous EWORLD participating countries. Focusgroups literature reviews were conducted to identifyperceptions of the characteristics, traits and behaviours ofentrepreneurs in the different countries as reported by focusgroup participants. This provided initial information andinsight into the content of the implicit entrepreneurialprototypes in the different cultures. A list of 115entrepreneurial characteristics, traits, and behaviours obtainedfrom the focus groups and literature revievys served as itemsfor a questionnaire used to obtain a larger amount ofquantitative data to examine the entrepreneurial prototype.Each of the 115 characteristics, traits and behaviours wereplaced on the questionnaire and participants in the variouscountries were asked to rate the characteristics and traits on a1-7 scale regarding their importance for the entrepreneur topossess in their specific country. Following standard researchprotocol, the questionnaire was translated into the languagesof the various countries and back-translated into English tocreate equivalent versions of the questionnaire in the differentlanguages. The data from each country were then individuallysubjected to factor analysis to identify the factor structure ofthe entrepreneurial prototype for each country sample. Dataare presented from the Czech Republic, France and China todemonstrate the construct equivalency issue here.Results and Discussion

The use of the questionnaire cross-culturallypresented several construct equivalency problems for theEWO RLD Project. T he first challenge occurred in ' thequestionnaire development stage and demonstrates howtranslation and back translation will not necessarily resolve

construct equivalency problems. The first challenge was theissue of semantics. Identical words have different meaningsacross cultures. Although translation and back-translationmay produce equivalent or identical translations of aquestionnaire, individuals in different cultures may interpretthe words on that questionnaire differently. In such cases, it isunlikely that the same construct is being measured by thequestionnaire items.

An item from the EWORLD questionnairedemonstrates this problem. Item 106 (see Table 1) asksparticipants to rate the importance of masculinecharacteristics for the entrepreneur in their culture. Table 1presents item 106 from both the English and French versionsof the questionnaire. Although the items are supposed to beidentical or equivalent translations in the two languages, theproblem is created by the semantic meaning of the word ag gressive in the item description. In the United States, theword aggressiv e can be used positively or negatively,depending on the context. In the questionnaire, it is intendedas a positive trait, a masculine trait that may make it morelikely that the entrepreneur will create and capitalize onmarket opportunities. In the item description, it is listed alongwith com petitiveness and is likely to be perceived in theUnited States as a part of the competitive spirit and therefore,positive. In France, however, the word aggress ive has anegative connotation. According to the EWORLD Frenchcollaborators, the word aggre ssive in France means to behostile and mean, and French participants are likely tointerpret the word in this way. Th is is ' a very differentmeaning than the spirit of competition that the item wasoriginally intended to measure in the English version.Therefore, although identical translations of the questionnairewere created, this in no way guaranteed that identicalconstructs were being measured, as some words on thequestionnaire convey different meanings across cultures. '

Table 1EWO RLD Project Questionnaire ItemItem 106 on EnglishQuestionnaire106. MasculinecharacteristicsItem 106 on Frenchquestionnaire '106. Caractristiquesmasculines

Shows competitiveness,aggressiveness andassertivenessMontre de lacomptitivit, del'agressivit et del'assurance

As second construct equivalency issue is created dueto the constructs themselves differing across various cultures.Both the content and structure of the implicit prototypeschange. People across cultures have different perceptions andideas about the nature of entrepreneurship and entrepreneurs.If the construct itself varies cross-culturally, then it isunlikely that an identical measure will be a construct validmeasure of the construct in each setting. Initial quantitative 55 )


4/6

Advances In Management Vol. 3 4) April 2010data from the EWORLD Project obtained in the CzechRepublic, China and France, demonstrate how the implicitentrepreneurial prototype differs in these countries. Forsimplicity, only the first 10 items on the EW OR LDquestionnaire are used to demonstrate the problem see tables2 - 4 for the items).

Table 2Czech Republic Factor Structure

PositiveIndependentRuthlessAnticipatoryRisktakerTrustworthyLoyalAutocraticIntelligentDecisive

Factors1-.299.481*-.505.345-.510.326.617*-.189.674*.266

2.083.386.514*.380

.547*.440.524.005-.421.246

3 .397*.171.026-.024-.199-.348.010.775*.043.682*

4-.312-.216.503.646*-.157

-.575*.091-.002.180-.110* indicates highest factor loading for each questionnaire item.

Table 3Chinese Factor Structure

PositiveIndependentRuthlessAnticipatoryRisk takerTrustworthyLoyalAutocraticIntelligentDecisive

Factors1

.319.585*-.489*.475*

.047.444*.663*-.378.497

.568*

2.406.140.406.445.419-.401-.316.635*.198.501

3-.490*-.522-.330.369.566*.247.065.166-.045.055

4.301.229-.333-.256.176.351-.185.231

-.651*.172* indicates highest factor loading for each questionnaire item.

Pilot data from these countries were subjected tofactor analysis to identify the factor structure ofentrepreneurial perceptions in each country. Although eachquestionnaire was an exact translation of the original Englishversion, tables 2 - 4 show very different factor structures anditem factor loadings obtained for these countries. Thisindicates that although the same questionnaire language wasused, the entrepreneurial construct has different content andstructure across the different cultures and therefore thequestionn aire items characteristics and traits) correlatedifferently and load differently with each other on the factors.Therefore, it cannot be safely assumed that the sameconstructs are being measured across the two cultures andthat construct equivalency ex ists. Both the Czech and Ch inesefactor structures include four factors. But as is seen in tables2 and 3, the items load vary differently on these factors. TheFrench factor structure is even more different. The items load

on only three factors. These different factor structures providevidence that although identical language is being used, thimplicit entrepreneurial prototype construct is different acroscultures. It is unlikely that one questionnaire used eticallywill be valid for measuring the construct emically, or in aspecific cultural context. Both the problem of semantics andof the changing nature of the construct of interest acroscultures is a problem of conceptual construct equivalency. Acan be seen, the translation and back-translation procedurecommonly employed in cross-cultural research do not addresand resolve problems such as these.

Table 4French Factor Structure

PositiveIndependentRuthlessAnticipatoryRisktakerTrustworthyLoyalAutocraticIntelligentDecisive

Factors1.803*.253

-.358.718*.485

.780*.554-.448.755*.646*

2.060

.641*

.592*-.166.182.075.202

.668*-.023.277

3-.118.291.109.386

.530*-.380-.658*-.265-.096.217

* indicates highest factor loading for each questionnaire itemThere are implications of these results for research iinternational management. Research methodology musinclude much more than the mere translation and back

translation of questionnaires. Oftentimes, this is the extent othe steps taken to measure constructs cross-culturally. But ashown here, this alone does not address the construcequivalency problem. It is only a first step in the procesTranslation processes at best represent a pseudo-etic approacand research that that employs nothing more than this tmeasure constructs cross-culturally will likely be based omeasures that lack construct validity and cannot measure thequivalent constructs across cultural settings. Thereforresearch results presented in such studies should be suspect.International cross-cultural research must includboth qualitative and quantitative methods and more comple

analytical and statistical techniques to establish culturaconstruct equivalency. Emic approaches to cross-culturconstruct measurement must be taken. Researchers mubegin with qualitative methodology to gain an understandinof the nature of the construct of interest in each of thcultures in which research will be conducted. Qualitativmethods can include ethnography, focus groups and aexamination of cultural artifacts that will provide insight inthe nature of the con struct.

Quantitative methods such as confirmatory factanalysis and structural equation modeling must then b56 )


5/6

Advances In Management. Vo l. 3 (4) April 2010)employed to examine the nature and structtire of the constructin each culture. These methods and statistical procedures willena ble t he , r|eiir.Cher_ to iden tify the latent variab les thatunderlie.-the manifest questionnaire items. If different factorstructures indicate that the nature of the construct is differentacross cultures, then it is likely that different versions of aquestionnaire will need to be developed to measure the sameconstruct in each culture. The combined qualitative andquantitative approaches reflect an emic approach to scaledevelopment and construct measurement and increase thelikelihood that a questionnaire will be developed that will bea construct valid measure of the construct of interest for eachcountry included in the research.

A need for different questionnaire versions tomeasure entrepreneurial prototype perceptions in the CzechRepublic and France is demonstrated here. If we wanted todevelop a questionnaire to measure the degree of fit betweenan individual entrepreneur and the cultural entrepreneurialprototype in each country, then we would need aquestionnaire that measured the four latent constructs(factors) found in the Czech data for use in the CzechRepub lic and a different questio nnaire that measured jus t thethree latent constructs (factors) found in the French data foruse in France. The use of the same questionnaire to measureentrepreneurial prototype fit (an etic approach) would likelylead to either contamination of measurement in France ordeficiency of measurement in the Czech Republic. Eithermeasurement contamination or deficiency will decrease theconstruct validity of the measure.

The results presented here were obtained from veryinitial quantitative data collected as part of the EWORLDProject. Although a more detailed demonstration of theconstruct equivalency problem may have been presented withconfirmatory factor and structural equations analyses, webelieve that the demonstration of the problem with initialexploratory analy.ses and small sample sizes indicates theprobable magnitude of the problem in cro.ss-cultural studies.Future international management research must continue toinvestigate the construct equivalency problem and emicapproaches.References1. Boyacigillcr N.A. and Adler N.J., The parochial dinosaur:Organizalional .science in a global context. The Academv ofManagement Review.16 2).262-290 1991)2. Churchill G.A.. A Paradigm los developing better measures ofmarkeliiig constructs.Journal of M arketing Research 16 .64 1979)3. DcVellis R.F., Scale development: Theory and applications,Newbury Park. CA, Sage Publications 1991)

4. Douglas S.P. and Craig C.S., International marketing research.Englewood Cliffs, NJ, Prentice-Hall 1983)5: Ford J.B., Cross-cultural scale development: An emic-eticbalancing act. Paper presented at the meeting of the Irish Academyof Management, Dublin, Ireland, September 2008)6. Hayton J.C . George G. and Zahra S.A., National cu lture andentrepreneurship: A review of behavioral research, EntrepreneurshipTheory and P ractice 26 4).33-52 2002)7. Hofstede G., Culture's consequences: International differences inwork-related values, Beverly Hill, CA, Sage Publica tions 1980)8. Hofstede G., Culture's consequences: International differences inwork related values,2' ed.. Thousand Oaks, CA, Sage 2001)9. House R.J., Hanges P.J.. Javidan M., Dorfman P.W., Gupta V.,and GLOBE Associates, Leadership, culture and organizations: Theglobe study of 62 societies. Thousand O aks, CA, Sage 2004)10. Lord R. et al Leadership and informational processing: Linkingperceptions and performance, Boston, Everyman 1991)11. Marcus B., Attitudes towards personnel selection metho ds: Apartial replication and extension in a German sample. AppliedPsychology: An International Review.5 2,515-532 2003)12. McArthur D.N., Construct equivalence in inicrnational businessresearch: The first and the last of it. The Journal of Business inquiry:Research Education an d Application 6(1), 28-38 2007)13. Mueller S.L. and Thomas A.S., Culture and entrepreneurialpotential: A nine country study of locus of control andinnovativeness. Journal of Business Venturing 16 .51-75 2000)14. Peter J.P., Reliability: A Review of Psychometric Basics andRecent Marketing Practices, Journal of Marketing Research 16, 6-17(1979)15. Peter J.P., Construct Validity: A Review of Basic Issues andMarketing Practices,Journal of Marketing Research 18 , 133 1981)16. Phillips J.M. and Gully S.M., Fairness reaction s to personnelselection techniques in Singapore and the US, International Journalof Human Resource Management 13 8), 1186-1205 2002)17. Scroggins W.A., Benson P.G.. Cross C. and Gilbreath B..Reactions to selection methods: An international comparison.International Journal of Management 25 2),203-216 2008)18. Thomas A.S. and Mueller S.L., A case for comparativeentrepreneurship: Assessing the relevance of culture. Journal ofInternational Business Studies 31 2),287-301 2000)19. Triandis H.C., Individualism and collectivism. Boulder, CO,Westview Press 1995).

(Received 15*January 201 0, accepted 15' March 2010) Advances In M anagement'? is blind peer reviewed international m onthly jour na lbeing published regularly since April2008.Ou r journ al is approved for inclusion inCabell's Directory and Research Papers in Economics (RePEc).

57


6/6

Copyright of Advances in Management is the property of Advances in Management and its content may not be

copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

Research Challenges in Cross-Cultural International Business Management

Documents

Transcript of Research Challenges in Cross-Cultural International Business Management