2006The effects of cultural diversity in virtual teams versus face-to-face teams

󰁢CSpringer2006

TheEffectsofCulturalDiversityinVirtualTeams

VersusFace-to-FaceTeams

D.SANDYSTAPLESANDLINAZHAO

Queen’sSchoolofBusiness,Queen’sUniversity,Kingston,Ontario,CanadaK7M6K5(E-mail:sstaples@business.queensu.ca)

Abstract

Diversityinthenationalbackgroundandcultureofteammembersiscommoninvirtualteams.Anexperimentalstudy,withshorttermteams,wasundertakentoexaminetheeffectofculturaldiversityonteameffectivenessandtoexamineifthiseffectchangesdependingiftheteamworkedface-to-face(F2F)orvirtually.Heterogeneousteamswerecreatedthathadgreaterdiversitythanhomogeneousteamsofindividualism/collectivismvalues,differentlanguagesspoken,countryofbirth,andnationality.TheteamsworkedonadesertsurvivaltaskeitherF2Forvirtually(viaaudioconferenceandelectronicchattools).Theoverallresultsindicatedthatheterogeneousteamswerelesssatisfiedandcohesiveandhadmoreconflictthanthehomogeneousteams,althoughtherewerenostatisticaldifferencesinteamperformancelevels.However,examiningjusttheheterogeneousteamsfoundthattheperformanceofthevirtualheterogeneousteamswassuperiortothatoftheF2Fheterogeneousteams.TheresultssupportCarteandChidambaram’s(2004)theorythatthereductivecapabilitiesofcollaborativetechnologiesarebeneficialfornewly-formeddiverseteams.

Keywords:teamdiversity,virtualteams,culturaldiversity,nationalculture,experiment,survivaltask,reductivecapabilitiesofcollaborativetechnologies

Introduction

Teamsoffercompaniespotentiallyeffectivewaystocombinethevariousskills,talentsandperspectivesofagroupofindividualstoachievecorporategoals.Withglobalization,transnationalteamsareareality(EarleyandMosakowski2000)andneedtobestudied(Hambricketal.1998).Theseteamsareusuallymadeupofmembersfromdifferentnationalbackgrounds,meaningthememberscomefromdifferentnationalcultures,possiblyspeakdifferentlanguages,andwereraisedindifferentcountriesthatmayhavedifferentvaluesystems.Today’selectroniccommunicationcapabilitiesmakesiteasierandcommonfortheseteammemberstoworktogetherwhilebeingphysicallylocatedindifferentcitiesorcountries.Thesegeographically-distributedteamsarecommonlyreferredtoasvirtualteams.Thedegreeofgeographicdispersionwithinavirtualteamcanvarywidelyfromhavingonememberlocatedinadifferentlocationthantherestoftheteamtohavingeachmemberlocatedinadifferentcountry.Virtualteamsthatspanmultiplecountriescreatethestrongpossibilitythatmembersoftheteamwillhavediversenationalbackgrounds(Evaristo2003;Powelletal.2004).Understandingthepossibleimpactsofthisdiversityonteamperformanceisimportantfortoday’sorganizations.

390

STAPLESANDZHAO

Thepurposeofthisstudyistoexaminetheeffectofculturaldiversityonteamprocessesandperformanceandtoexamineifthiseffectchangesdependingonthecommunicationmodeused.Intheirtheorypaper,CarteandChidambaram(2004)proposedthatthereductivecapabilitiesofcollaborativetechnologies(e.g.,electronictoolssuchasemail,groupsupportsystems,computerconferencing)canreducethenegativeeffectsofdiversityearlyinthelifeofadiverseteam.Iftheeffectofculturaldiversityisdifferentforteamscommunicatingelectronicallyversusthosecommunicatingface-to-face(F2F),thismayhaveimportantimplicationsforthedesignofvirtualteams(i.e.thoseusinginformationandcommunicationtechnologies(ICT)astheirprimarycommunicationmedia)andtheirorganizations.Forexample,asCarteandChidambaramsuggested,diversevirtualteamsmaybebetteroffnotmeetingF2Funtilrelationshipshavebeendeveloped.Minimizingthesalienceofsurface-leveldiversitybyavoidingF2Fmeetingsearlyinthelifeofteammayreducethepotentialnegativeimpactofthisdiversity.

AnexperimentalmethodwasusedtoexaminetheeffectofculturaldiversityonF2Fandvirtualshort-term,projectteams.Inthisexperiment,virtualteamscommunicatedviaaudioconference(telephoneconferencing)andasynchronouschatsystem.Anaudioconfer-encephonesystemwaschosenasthemaincommunicationtoolforthevirtualteamssincestudiesofseveralon-goingindustry-basedvirtualteams(Staplesetal.2004)foundthatthemostcommonwayteamsmetwereviaaudioconferencing,withotherelectronictoolsusedasneededduringtheaudioconferencemeeting(e.g.,electronicwhite-boards,synchronouschats,instantmessaging).Choosingaudioconferenceandsynchronouschatasthecom-municationchannelsforthevirtualteamsalsoaddressesBaltesetal.(2002)callformorerelevantresearchusingindustry-adoptedcommunicationtools,suchasaudioconferencing.Thispaperisorganizedasfollows.Inthenextsection,thetheoreticalbackgroundandrelevantliteraturearepresentedandthehypothesesforthestudyaredeveloped.Themethod-ologyusedtotestthehypothesesisthenpresented,followedbyadescriptionoftheresults.Thefinalsectiondiscussesthefindingsandlimitations,andofferssuggestionsforfutureresearch.

TheoreticalBackgroundandDevelopmentofHypotheses

AsreviewedbyMillikenandMartins(1996),differentaspectsofdiversityinteamshavebeeninvestigatedincludingobservabledifferencessuchasrace,ethnicity,gender,andage,andunobservabledifferencessuchasskills,informationandknowledge,values,cognitiveprocesses,andexperience.Varietyinovertcharacteristics(readilyobservabletraits)istypicallyreferredtoassurface-leveldiversity,whilevarietyinunobservablecharacteristicsisreferredtoasdeep-leveldiversity(CarteandChidambaram2004).

Therecanbebothpositiveandnegativeaspectsofteamdiversity.Valueindiversitycomesfromincreasedcreativity,innovationandflexibility(Jehn,NorthcraftandNeale1999;LauandMurninghan1998;McLeodetal.1996).Avarietyofperspectivesandexperiences(i.e.,deep-leveldiversity)canbringmoreinformationandideasintotheteam,stimulatethinking,andcanbringdifferentnetworksofcontactsandresourcestotheteam.Whenthedeep-leveldiversityisrelevanttothetaskfacingtheteam,higher-qualityoutcomesshouldresult;however,ifthevarietyisnotrelevanttothetask,thenthereisnobasisforexpecting

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS391

thediversitytoenhancetheteam’sactivity.Therefore,thebeneficialeffectofdeep-leveldiversitywithinateamiscontingentontherelevanceofthediversitytotheteam’stask(McLeodetal.1996).

Negativeaspectsofteamdiversityincludecommunicationdifficulties,misunderstand-ings,decreasedcohesionandincreasedconflict.Theseprocesslossesresultindecreasedperformanceandsatisfaction(Hambricketal.1998;LauandMurninghan1998;WilliamsandO’Reilly1998).Socialidentitytheory,socialcategorizationtheoryandthesimilar-ity/attractionparadigmsuggeststhatthenegativeeffectsassociatedwithdiversityareduetothecreationofin-groupsandout-groups(CarteandChidambaram2004;SalkandBrannen2000).Peopleimplicitlycategorizethemselvesintosubgroupsaccordingtosalientcuesandidentifymorecloselywithpeopletheyperceiveasbeingsimilartothemselves.Theydothistoachieveandmaintainpositiveself-identity.Asin-andout-groupcharac-teristicsbecomesalientwithinsubgroups,individualsbecomemorebiasedtowardstheirsubgroup.Emotionalattachmentstothesubgroupsbecomepotentialsourcesforinterper-sonalandrelationshipconflictwithmembersofothersubgroups.Relationshipconflictreducessatisfactionandteamperformance.Lowerpersonalattractiontogroupmembers,higherturnover,andpoorercommunicationalsoresultsinreducedsocialintegrationandcohesion.Ifgroupmembersarediverseonmultipleattributesthataligntogether,strongfaultlinescandevelopwhichcreatefurthersubgroupproblemsandteamdevelopmentprob-lems(LauandMurninghan1998).

Theeffectofthedifferenttypesofdiversityonteamoutcomesissignificantlyaffectedbytime(CarteandChidambaram2004).Whengroupsnewlyform,theyhavelittleinformationabouteachotherthanobservablecharacteristics.Memberstrytomakesenseofeachotherandtheirtask.Ifsurface-leveldiversityispresent,membersmayusesalientcharacteristicstoimplicitlycategorizethemselvesintosubgroups,creatingthenegativeeffectsdescribedabove(LauandMurninghan1998).Althoughsurface-leveltraitsareimmediatelyapparentuponteamformation,deep-leveltraitsbecomesalientastheteammembersinteractovertime.Thepotentiallypositiveeffectsofdeep-leveldiversitytaketimetoemerge.Therefore,intheshort-term,diverseteamstypicallyperformworsethanhomogeneousteamsduetotheearlyimpactsofsurface-leveldiversity.Empiricalresearchhasfoundthispatterninthatteamdiversitybenefitsseemtobeobtainedafterateamhasinteractedforsometimeandrelationaltieshavebeendeveloped(CarteandChidambaram2004;McLeodetal.1996;Watsonetal.1993).

Whiletherearemanydifferenttypesofdiversity,thisstudyfocusesondifferencesinthenationalandculturalbackgroundofteammembers.Culturalbackgroundhaselementsofbothsurface-levelcharacteristicsanddeep-levelcharacteristics.Surfacelevelcharacteristicsthatcanvarydependingonwhereonewasbornand/ornationalityincluderaceandethniccharacteristicsandnativelanguage.Deep-levelcharacteristicsthatwillvaryforpeoplefromdifferentcountriesareculturalvalues.Thenationalcultureonegrowsupandlivesininfluencesthinking,expectations,andbehavior(Evaristo2003;Hambricketal.1998).Thisstudyfocusesonsurface-levelculturaldiversityspecificallyintermsofnationality,countryofbirthandnativelanguage,andononedeep-levelaspectofculturaldiversity–individualism/collectivismvalues.Thisaspectofnationalcultureandthereasonforincludingitisexplainednext.

392

STAPLESANDZHAO

Nationalcultureisacomplexconstructtodefineandstudy.Acommondefinitionofnationalculture(Evaristo2003)isofferedbyHofstede(1980).Hedefinesculture“asacollectivephenomenon,becauseitisatleastpartlysharedwithpeoplewholiveorlivedwithinthesamesocialenvironmentwhereitwaslearned.Itisthecollec-tiveprogrammingofthemindthatdistinguishesthemembersofonegrouporcate-goryofpeoplefromanother.”Hofstede’sresearchdefinesnationalcultureintermsoffivevaluedimensions:individualism/collectivism(personalinterestsversusgroupinter-ests),powerdistance(acceptanceofinequality),uncertaintyavoidance(dislikeforam-biguity),masculinity/femininity(assertivenessandfocusonworkgoalsversuspersonalandfamilygoals),andtimehorizon(sometimescalledConfuciandynamism–short-termversuslong-termorientation)(formoreinformationonthedimensionsandhowtheywerederived,seeAndersonandHiltz2001,Bond1988,Evaristo2003orHofstede1983).

Whilerecognizingtheimportanceofthefivedimensions,thisstudyfocusesontheindi-vidualism/collectivismdimensionduetoitspotentialrelevancetothephenomenonbeingstudy–teamperformance.Individualism/collectivismindicateshowanindividualputshis/herinterestsaheadoftheteam’sinterests.Researchhasfoundthatpeoplefromcol-lectivistculturalbackgroundsaremorewillingtohelppeople,makepersonalsacrificesandaremorecooperativethanpeoplefromindividualistculturalbackgrounds(McLeodetal.1996).Therefore,individualismvaluespotentiallyaffectcommunicationandcoordi-nationpatternsamongindividualsworkinginteamsandtheirexpectations(Earley19).Inhighindividualisticcultures,peoplerelyontheuseofwordstoconveymeaningwhereasinlowindividualistic/highcollectivisticcultures,toneofvoice,timing,facialexpressionsandbehaviorarealsoimportantpartsofthecommunication(AndersonandHiltz2001).Theindividualism/collectivismdimensionhasalsobeenusedinpreviousresearchtorep-resentrespondents’nationalculture(e.g.,KessapidouandVarsakelis2002;StedhamandYamamura2004),andSondergaard(1994)suggestsitisthemostvalidatedofHofestede’sfivedimensions.

Itisexpectedthatduetothein-groupandout-groupproblemsdescribedabove,surface-leveldiversityregardingnationalbackgroundwillleadtolowercohesionandgreaterconflictintheearlylifeofateam.Thisinturnleadstolowerteamperformanceandsatisfaction.Teamsthatarehighlydiverseonindividualism/collectivismvalueswillhavepeopleontheteamthathavedifferentexpectationsandvaluesregardingcommunicationsandinteractionpatterns.Somepeoplewillbewillingtocooperateandsacrificefortheoverallteam,andsomewillnot.Thisalsocouldcreateconflictandlowercohesion,contributingtolowerteamoutcomes.Differencesinnativelanguagescanalsocontributetocommunicationdifficultieswithinteams(McDonoughetal.1999).Iftheculturaldiversityisnotbeneficialtothetasktheteamisperforming,thentherewouldbenooffsettingincreaseinperformanceovertimeandthediverseteamswouldbeexpectedtoperformworsethanhomogeneousteams.Thus,inthiscontext,ourfirsthypothesisis:

Hypothesis1–Higherlevelsofculturaldiversitywillbeassociatedwithlowercohesionandhigherconflict,andlowerteamoutcomes(i.e.,teamperformanceandsatisfaction).

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS393

CarteandChidambaram(2004)proposethatcommunicationtechnologieshavebundlesofcapabilitiesandthesecapabilitiescanbecategorizedastwotypes:reductiveoradditive.Reductivecapabilitiesreduceaspectsofcommunicationandspeechpatternsthatwouldbepresentintraditionalface-to-facecommunication.Reductivecapabilitiesincludevisualanonymity(identificationislimited),equalityofparticipation(normalturntakingmaybereduced),andasynchronouscommunication(immediatefeedbackislimited).Additivecapabilitiesenhancenormalcommunicationexchangesandincludecoordinationsupport(trackingresourcesandprojectprogress),electronictrails(creatingrecordsandretriev-inginformation)andenhancedfunctions(decisionmakingtools,filetransfersandrichmessaging).

CarteandChidambaram(2004)furthersuggestthatthebundlesofcapabilitiesaremostusefulatdifferentstagesofadiverseteam’sdevelopment.Specifically,reductivecapabil-itiesarevaluableearlyinthelifeofadiverseteam,whereastheadditivecapabilitieswilladdvaluelaterinateam’slife(i.e.,afterasharedteamidentityisestablished),byprovidingsupportfordecision-makingandcoordination.Thisstudyisparticularlyfocusedonthepo-tentialbenefitsofreductivecapabilitiesinthefunctioningofanewly-formeddiverseteam.Thekeyreductivecapabilityisvisualanonymitysincethisreducestheimmediatesaliencyofsurface-leveldiversity.Teammemberscanassessinputsfromothersandformopinionsbasedonmerit,withoutsurface-leveldiversityaffectingjudgments.Decreasingperceivedsurface-leveldiversitypotentiallyreducestheteammember’scategorizationprocesses,therebyreducingtheformationofperceivedin-groupandout-groups.Lessdisintegrationoftheteamintosubgroupsshouldimproveteaminteractionprocessesbyreducinginterper-sonaldisagreementsandconflict,resultinginhighercohesion,teamperformanceandteamsatisfaction.Enhancedequalityofparticipationcanincreaseparticipation,allowminorityopinionstobeheard,andfosterasenseofbelongingtoagroup.Asynchronouscommu-nicationcanalsobenefitanewly-formedteamsincememberswillhavetimetoconsiderhowtheyshouldsaythings,potentiallyreducingquickreactions,miscommunicationandattributionerrors(CarteandChidambaram2004).

Asmentionedintheintroduction,tomirrorvirtualteampractice,thevirtualteamsinthisstudycommunicatedwithtelephoneconferenceandelectronicchattools,withthepar-ticipantschoosinghowfrequentlytouseoneorbothofthetools.Thereductivecapabilitiesofthesecommunicationmediawouldnotbeattheextremeendsofthepotentialtoreducesalienceofdiversity(CarteandChidambaram2004);however,somereductivecapabilitiesarecertainlypresentinthisbundleofcommunicationmedia.Forexample,althoughcom-municatingwithmediaricherthanapuretext-basedsystem,ourvirtualteamsdidnothavevisualcontactandtheycommunicatedusingdisguisednames.Therefore,visualanonymitywaspresent.However,telephoneusedoesallowonetoheartoneandvoiceexpression(suchthattheamountofverbalcuesandabilitytonoticedifferentlanguageaccentsarerelativelyhighcomparedtoanothermediumlikeanelectronictext-basedmessagesystem(Baltesetal.2002)).Telephoneconferenceishighonsynchronicitysincetheconversationshappeninreal-timeandthechattoolusedwasasynchronoustool.Therefore,theasyn-chronouscommunicationreductivecapabilitypotentialbenefitswereverylimitedinthisstudy.Equalityofparticipationthroughtheelectronicchattoolandthetelephonewouldpotentiallybehigherthanface-to-facecommunication,althoughnotashighasinapure

394

STAPLESANDZHAO

text-basedchatsystem.Overall,consistentwithCarteandChidambaram’s(2004)propo-sitions,weexpectthereductivecapabilitiesinthecommunicationmediausedbyvirtualteamstoreducethesalienceofsurface-leveldiversity.Inaddition,wealsoexpectthattheeliminationofvisualcueswillreducethevisibilityofdifferentcommunicationandinter-actionstyles,andthevisibilityofnegativereactionstothisdiversity,reducingthenegativeimpactonateamhavingmemberswithdifferentindividualistic/collectivisticexpectationsandvalues.Hence,weproposethatthenegativeeffectsofculturaldiversity(surface-leveldiversityandindividualismvaluediversity)invirtualteamswouldbereducedbythelackoftraditionalface-to-facecommunication.Inteamswithlowdiversity(i.e.,culturallyhomo-geneous),wedonotexpecttoseedifferencesbetweentheface-to-faceandvirtualteams.Thus:

Hypothesis2–Communicationmodemoderatestherelationshipsbetweenculturaldiversityandteamprocessesandoutcomes.Specifically,culturallyheterogeneousvirtualteamsusingcommunicationtechnologieswithreduc-tivecapabilitieswillhavehighercohesiveness,lessconflict,bettertaskperformance,andhighersatisfactionthanculturallyheterogeneousF2Fteams.

Themethodologyusedtotestthesehypothesesisdescribednext.

MethodologyParticipants

Seventy-nineteamsparticipatedinthe2×2experiment(seeTable1).Teamsizewasdesignedtobe5people,althoughsomelastminuteno-showsresultedin4-personteams(whichwasdeemedtobeacceptableintermsofstimulatingenoughinteraction;threepersonteamswerenotallowed).Intotal,380peopleparticipated.Foradequatepower,aminimumof15teamspercell(ThompsonandCoovert2003)wasrequiredandexceededforallcells.

Participantswereuniversitystudents(60%undergraduate;40%graduatestudents)withafairlyevengendersplit(i.e.,58%female).Diversityofstudentmixanddiversityofgender

Table1.Experimentdesignandsampleinformation.

Culturaldiversity

Homogeneous(low)

Communicationmode

F2F

21teams

12teams∗5person+9teams∗4person(n=96)20teams

17teams∗5person

+3teams∗4person(n=97)

19teams

18teams∗5person

+1team∗4person(n=94)19teams

17teams∗5person

+2teams∗4person(n=93)Heterogeneous(mixed–high)

Virtual

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS395

Table2.Backgroundofsubjects.CountryofbirthCanadaorU.S.A.AsiaEuropeAfrica

SouthAmericaMiddleEastMexico

CentralAmericaAustralia

Numberofpeople1951261711119821380

Percentage(%)51.333.24.52.92.92.42.10.50.3

wassimilarinthefourcells(i.e.,therewerenostatisticaldifferencesonBlauheterogeneityindexes:F(3,75)=0.9,p>.10;F(3,75)=0.2,p>.10,respectively).Thetwolargestgroups(seeTable2)intermsofcountryofbirthwerepeopleborninCanadaortheU.S.A.(51%)andinAsia(33%),providinggoodvarianceonculturalvalues(Hofstede1983)andsurface-leveldiversity.

Participationofsubjectswasvoluntaryandwasnotlinkedtoanycoursesorcoursecredit.Subjectseachreceived$15forparticipating.Asateamperformanceincentive,thetopteamsineachtreatmentcellwerealsogivena$20perpersonbonus.Subjectsweredrawnfromacrossmultiplefacultiesofalargeuniversitycampusandwerecombinedintoteamswithmembersfromdifferentareasasmuchaspossibletominimizepreviousknowledgeoffellowteammembers.Totestforprevioushistory,wedidasksubjectsinapost-questionnaireiftheyknewanyoftheirteammatespriortothestartoftheexperiment(i.e.knewtheminanyway–wedidn’taskiftheyworkedwitheachotherbefore).Seventypercentknewnobodyontheteamand24%knewonlyoneperson,implyingthattheteamdidhaveverylowpriorhistorysince94%knewnoneoronlyonememberoftheteampriortothestudy.

Decisiontask

Adecision-makingtaskthatrequiredinteractionandcommunicationandthathadanexpertsolutionavailable(tocreateameasureoftheteam’sperformance)wasneeded.Wealsowantedataskwheretheculturaldiversitywasunlikelytobeofbenefitsothatthepotentialpositiveinfluencesofdifferentviewsdidnotconfoundthepossiblenegativeeffectsofdiversityoninteractionsandteamprocesses.JohnsonandJohnson’s(1994)desertsurvivaltaskwaschosensinceitwasascenariothatfewpeoplewouldhaverelevantexperienceinandculturalbackgroundwasnotlikelytobebeneficial.Inthistask,participantsfirstreadashortdocumentthatplacesthemintoanairplanecrashscenarioinadesertwiththerestoftheirteammates.Therearetwelveitemsthatparticipantshavetorankinorderoftheirimportanceforsurvival.Thisisdonefirstindividuallyandthenasateam.Thistaskhasbeenusedfrequentlyinsmallgroupresearch(e.g.,BottgerandYetton1987;Haslametal.

396

STAPLESANDZHAO

1998;RogelbergandO’Connor1998;Straus1996)andhassimilaritiestoproblemsdealtwithbytemporaryteamsonthejob(ThompsonandCoovert2003).ItismostlyaQuadrantIItaskinMcGrath’s(1984)grouptaskcircumplex,sinceitisaninterdependentintellectivetask(PotterandBalthazard2002).Itrequiresteamstosolveaproblemthathasacorrectanswer(i.e.,anexpert’sanswer).Italsohasaspectsofajudgment/decisionmakingtasksinceateammembercannotprovethecorrectnessofhis/heranswerandhastopersuadeteammates,andhasaspectsofanegotiation/cognitive-conflicttasksinceteammateshavetodiscussandresolvedifferingopinionsregardingsurvivalstrategiesandtherankingoftheitems(ThompsonandCoovert2003).

Designandprocedure

Fourexperimentalconditionscorrespondingtotwomediatypes(i.e.,communicationmode)andtwodegreesofculturaldiversitywerecreated(seeTable1).Theface-to-face(F2F)teamsmetinaroomandworkedonthetaskaroundatablewithoutanycomputertools.ThevirtualteamsdidnotmeetF2F.Theyworkedinseparaterooms(withinthesamebuilding)sotheycouldnotseeeachotherastheyinteracted.TheycommunicatedasateamviatelephoneconferenceandasynchronouselectronicchatsystemonPC’s(implementedthroughLotusNotes–seeFigure1).Itwasuptheteam’smemberstodecidehowtheyusedthemediachoices.Someteamsusedbothextensively,whereasotherstendedtouseonemediummorethantheother,withtelephoneusuallybeingthemediumofchoiceinthesituation.

Theculturaldiversitytreatmentwascreatedusinginformationgatheredfromsub-jectsinapre-questionnairetheycompletedbeforetheywereassignedtoexperimen-talteams.Thisallowedustocollectinformationusedtocreatehomogeneousandheterogeneousgroups.Twooftheindicatorsofculturaldiversitywereusedtocre-atetheseteams.Ascorefortheindividualism/collectivismindexofHofstede’sVal-uesSurveyModule1994(VSM94)wasused(usingHofstede’sformula),alongwithinformationonthesubject’sfirstlanguage.Varietyofsubjects’languagehasbeenpreviouslyusedasanindicatorofdiversityofnationalbackground(e.g.,Ander-sonandHiltz2001;MortensenandHinds2001)andtheindividualism/collectivismdimensionhasalsobeenusedinpreviousresearchtorepresentrespondents’na-tionalculture(e.g.,KessapidouandVarsakelis2002;StedhamandYamamura2004).Creatingteamsthathaddissimilarindividualism/collectivismscores(i.e.,thehighdiversity/heterogeneousteams)placedpeopleonthesameteamthatpoten-tiallyhaddifferentexpectationsandvaluesregardingcommunicationsandinter-actionpatterns,which,aspreviouslyexplained,couldaffectteamprocessesandoutcomes.

Thesubjectswereassignedintoteamsoffivesuchthatthehomogeneousteamshadmemberswithallthesamefirstlanguage(i.e.,nodiversityonfirstlanguagecharacteristics)andsimilarvaluesontheirindividualscoresontheindividualismindex.Specifically,theteammembers’scoresoftheindividualismindexhadastandarddeviationoflessthantwenty(therangeobtainedfortheVSMindividualismscoreswas−45to225soastandard

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS397

Figure1.Asamplescreenviewofthesynchronouschatsystem.

deviationof20indicatesfairlysimilarindividualscores).Fortheculturallyheterogeneousteams,thereweretwoormorefirstlanguagesamongtheteammembers(i.e.,diversitywithintheteamonfirstlanguagebackground)andtheirscoresoftheindividualismindexmeasurehadastandarddeviationofmorethanforty(i.e.muchmorevarianceinthescoresthaninthehomogeneousteams).Theeffectivenessofthisprocedurewascheckeduponcompletionoftheexperiments.Theaveragestandarddeviationforthehomogeneousteamsandheterogeneousteamswere16.6(standarddeviationof6.7),and55.5(standarddeviationof12.7),respectively.Thesemeansarestatisticallysignificantlydifferent(t77=16.8,p<.001),indicatingthatthecreationofthediverseandsimilarteamsonthisdimensionwassuccessful.Note,thistreatmentwasnotattemptingtocreateteamsthatwerehighorlowontheindividualismindex;thepurposewastocreateteamswithmembersthathadsimilarvalues(i.e.lowdiversity)andteamswithmembersthathaddifferentvalues(i.e.,highdiversity).Consistentwiththis,theaverageindividualismindexscoresforthehomogeneous

398

STAPLESANDZHAO

andheterogeneousteamswererelativelysimilar(97.2and87.6,respectively;t77=1.6,p>.05).Themeanteamvalues(rangingfrom39to165)indicatethat,onaverage,theteamsheldscoresclosertotheindividualismendofthescale(i.e.,Hofstede’sVSMmanualsuggeststhat100indicatesindividualism,whereasscoresapproachingzeroindicatescollectivism).

Thediversityofteammembers’reportednationalityandcountryofbirth(collectedinthepre-questionnaire)wereexaminedbycreatingBlau’s(1977)indexofheterogene-ity(D)forthesetwoindicatorsofnationalculture.Blau’snationalityindexvaluesforthediversetreatmentteamsandthehomogenoustreatmentteamswerestatisticallydif-ferentat0.53and0.16,respectively((t77=8.48,p<0.001).Blau’scountryofbirthindexvaluesforthediverseandhomogenousteamswerealsostatisticallydifferentat0.63and0.21,respectively((t77=9.,p<0.001).Thespecificdiversityofcountryofbirth,combinedintoregions(seeTable2),wasalsoexamined.Forthehomogeneoustreatmentteams,54%ofallteamshadallmembersborninthesameregion,37%oftheteamshadmembersfromtworegions,7%werefromthreeregions,and2%oftheteamshadmembersfromfourregions.Fortheheterogeneoustreatmentteams,noteamshadallteammembersbornfromthesameregion,32%hadpeoplebornintwooftheregions,50%hadpeopleborninthreeregions,and18%hadpeopleborninfouroftheregions.Thisanalysissupportsthatthetreatmentdidcreateteamswithdifferentlevelsofdiversity,withrespecttotheindicatorsofnationalbackgroundandculturethatweremeasured.

Experimentalprocedure

ParticipantsfirstcompletedaquestionnairethatcollectedinformationontheirbackgroundandtheirculturalvaluesaspertheVSMquestions(Hofstede1994).Basedonthatinforma-tion,thehomogeneousandheterogeneousteamswerecreatedandscheduledforthetask.Uponarriving,membersoftheF2Fteamswereledtooneroomandaskedtositaroundata-ble.Theexperimentstartedwithparticipantsbeinggiventhedescriptionofthetaskdescrip-tionandthenbeinggiventenminutestoindividuallyreadandrankthesurvivalitems.Theythenworkedtogetherforuptoforty-fiveminutestodiscussandagreeonarankingthatwassubmittedastheirteam’srecommendation.Uponarrival,membersofthevirtualteamswereindividuallygivenashorttrainingsession(lessthanfiveminutessincethesystemwasquitesimple)forthesynchronouschattoolandthentheywaitedintheirseparateofficesuntilallteammemberswerereadytobegin.Whentheywereallready,teammemberswerealsogiventenminutestoworkindividually.Attheendofthisperiod,thecoordinatorphonedallteammemberstoestablishtheconferencecall.Theelectronicchatsystemwasalsoactivatedatthesametime.Theteamdiscussedtheirviewsforamaximumof45minutesandagreeduponasolutionthatwassubmittedtothecoordinatorelectronically.Apost-questionnairewascom-pletedbyallsubjectsaftertheirteam’ssolutionhadbeensubmitted.Thepost-questionnairecontainedmeasuresforthedependentvariables,exceptfortheperformanceconstruct(whichwasobtainedbycomparingtheteam’sanswerwiththeexpert’sanswer,asperthedeserttaskmanual).

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS399

Constructmeasurement

Thepre-questionnairegathereddemographicdataandmeasuredHofstede’sVSMfiveculturaldimensions(Hofstede1994).Thepost-questionnairemeasuredtheindividual’ssatisfactionwithteamprocessusingGreenandTaber’s(1980)scale(Cronbach’salpha=.77).Cohesion(Cronbach’salpha=0.79)andconflict(Cronbach’salpha=0.65)weremeasuredusingLind’s(1999)scales,whichwerebasedonVanDeVenandFerry’s(1980)measures.Teamperformancewasassessedbycomparingtheteam’srecommendationwiththeexpertrankingofthesurvivalitems.Theunitofanalysiswasattheteamlevelsotheindividualresponsesforsatisfactionwithteamprocess,cohesionandconflictwereaggregatedtotheteamlevel.Averagerwg(Jamesetal.1984)valueswere0.88,0.85and0.78respectively,indicatingitwasvalidtoperformtheaggregation(i.e.,greaterthan0.70).Analysistodeterminestatisticallysignificantdifferencesbetweenthetreat-mentsgroupswasdonewithMANOVAanalysis(H1)andt-tests(H2–duetothere-ducedsamplesize).One-tailp-valueswereused,consistentwiththedirectionalityofthehypotheses.

Results

TheconstructscoresforthedependentvariablesarepresentedinTables3through6.Hypothesis1wastestedwithMANOVAprocedures.StatisticallysignificantdifferencesbetweencellsEandFwereexaminedforallfourdependentvariables.TheOmnibustestwassignificant(F(4,74)=2.80;p=0.03).UnivariateF-testsshowedthattheteamattitudesandconflictwereallstatisticallysignificant(SatisfactionwithTeamProcessF(1,77)=9.78;p=0.001;CohesionF(1,77)=7.88;p=0.003;ConflictF(1,77)=3.32;p=0.036).Teamperformancewasnotstatisticallysignificant(F(1,77)=0.05;p=0.42).Examinationoftheconstructscoresshowedthatcohesionandsatisfactionwerehigherinthehomogeneousteamsandconflictwaslower.Thereforehypothesis1waspartiallysupportedinthatthediverseteamshadpoorerteamprocesses(cohesionandconflict)andlowersatisfaction;however,theperformanceoftheseteamswasnotstatisticallylessthanthehomogeneousteams.

Hypothesis2wasalsopartiallysupported.Examinationofonlytheculturallydiverse(i.e.,heterogeneous)teams,foundthattherewasastatisticallysignificantdifferenceintheteamperformance(cellDversusB:t(36)=1.71,p=0.048).Thecellmeans(BandD)showthatteamperformancescoresweresuperiorforthevirtualheterogeneousteamscomparedtotheF2Fheterogeneousteams,indicatingthevirtualteamrankingswereclosertotheexpertranking(i.e.,theyperformedbetter).TherewerenostatisticallysignificantdifferencesintheperformanceoftheF2Fhomogeneousteamsandthevirtualhomogeneousteams.TherewerenostatisticallysignificantdifferencesinanyoftheteamattitudesorconflictlevelsacrosstheF2FandvirtualhomogeneousteamsoracrosstheF2Fandvirtualheterogeneousteams.

400

STAPLESANDZHAO

Table3.Satisfactionwithteamprocessscores(alowscoreindicatesamorepositiveviewofteamprocess;aggregatedforeachteam).

Diversity

Homogeneous(low)

F2F

Heterogeneous(mixed–high)

Virtual

(Combined)

Communicationmode

CellACellBMean=1.68Mean=1.88SD=0.29SD=0.39N=21N=19CellCCellDMean=1.55Mean=1.80SD=0.27SD=0.33N=20N=19CellECellFMean=1.61Mean=1.84SD=0.29SD=0.35N=41N=38

Table4.Cohesionscores(higherscores=morecohesiveteam;aggregatedforeachteam).

Diversity

Homogeneous(low)

Communicationmode

F2F

CellA

Mean=4.39SD=0.37N=21CellC

Mean=4.42SD=0.26N=20CellE

Mean=4.40SD=0.32N=41

CellB

Mean=4.17SD=0.37N=19CellD

Mean=4.20SD=0.37N=19CellF

Mean=4.19SD=0.36N=38

Heterogeneous(mixed–high)

Virtual

(Combined)

Discussion

Inarecentreviewofvirtualteamresearch,Powelletal.(2004)suggestedthatdiversitymaybelessapparentinvirtualteams,potentiallyreducingtheprocesslossescausedbyculturalheterogeneity.CarteandChidambaram(2004)wentconsiderablybeyondthistosuggesthowdifferenttypesofdiversityaffectteamfunctioning,proposehowdifferenttypesofelectroniccommunicationcanaffecttheimpactofdiversity,andproposehowtimechangestheeffects.ThemaincontributionofourpaperistotestasubsetofCarteandChidambaram’spropositions.Whetherornotthereductivecapabilitiesofcommunication

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS401

Table5.Conflictscores(higher=lessteamconflict:aggregatedforeachteam).

Diversity

Homogeneous(low)

F2F

Heterogeneous(mixed–high)

Virtual

(Combined)

Communicationmode

CellACellBMean=4.35Mean=4.14SD=0.43SD=0.48N=21N=19CellCCellDMean=4.43Mean=4.30SD=0.35SD=0.34N=20N=19CellECellFMean=4.38Mean=4.22SD=0.39SD=0.42N=41N=38

Table6.Teamperformancescores(lowerscore=higherperformanceintermsofmatchingexpertssolution).

Diversity

Homogeneous(low)

Communicationmode

F2F

CellA

Mean=42.52SD=12.07N=21CellC

Mean=43.00SD=9.84N=20CellE

Mean=42.76SD=10.90N=41

CellB

Mean=46.00SD=9.12N=19CellD

Mean=40.53SD=10.58N=19CellF

Mean=43.26SD=10.13N=38

Heterogeneous(mixed–high)

Virtual

(Combined)

mediacouldreducethenegativeeffectsofnationalbackgroundandculturaldiversityinnewly-formedteamswasexamined.

Thisidea(Hypothesis2)waspartiallysupportedindicatingthatthereductivecapabilitieswithcommunicatingviatelephoneandelectronicchatdidreducethenegativeimpactofteamdiversity.AlthoughperformancewashigherinthevirtualheterogeneousteamcomparedtotheF2Fheterogeneousteam,therewerenostatisticaldifferencesinthesatisfaction,cohesionorlevelofconflict(althoughallhadslightlybettermeanvaluesforthevirtualteams).Post-hocanalysiswasdonetoexaminetherobustnessofthefindings.TheteamsweresplitintothosethatwerehighontheBlau’sheterogeneityindexfornationality(i.e.,highlyheterogeneous=highdiversity)andteamsthatwerelow(i.e.,homogeneous).Inthe

402

STAPLESANDZHAO

heterogeneousteams,thevirtualteamshadstatisticallybetterperformanceandstatisticallylessconflictthanintheheterogeneousF2Fteams.Therewerenosignificantdifferencesinthehomogeneousteams.ThesefindingsareconsistentwiththefindingsofHypothesis2andaddsupporttotheideathatreductivecapabilitiesmayalsoreduceconflict.

Theseresultshaveimportantimplicationsforpractitioners.Conventionalwisdomisthatvirtualteamsshouldhaveaprojectkick-offmeetingthatisheldF2F.Thishelpsestablishsocialbondsandrelationships.However,assuggestedbyCarteandChidambaram(2004)andsupportedbytheresultsofthisstudy,thispracticeshouldonlybefollowediftheteamsarehomogeneous.Ifteamsarediverse,especiallyonsurface-levelelements,thenrichmediameetings,suchasF2F,shouldbeavoideduntilateamidentityhasbeenestablished.Teamsshouldcommunicateusingcollaborativetechnologiesthathavereductivecapabilities.Inthisway,thecreationofsubgroupsthathurtteamprocessesandoutcomesisminimized.Ifthesurface-leveldiversityislowintheteam,thenthepracticeofearlyF2Fmeetingislikelyadvantageous.

Partialsupportwasalsofoundforhypothesis1inthatdiverseteamsoverall(i.e.,teamsinbothcommunicationtreatmentmodescombined)hadmorenegativeattitudestowardtheirteam(satisfactionandcohesion)andmoreconflict.Thisisconsistentwithprevioustheoryandresearchregardingthenegativeeffectofculturaldiversityonteaminteractions(e.g.,Johanssonetal.1999;KayworthandLeidner2000;MaznevskiandChudoba2001;SarkerandSahay2002;vanRyssenandGodar2000).Interactingwithpeoplewithsimilarculturalvalues,nativelanguageandnationalbackgroundleadstoacohesiveteamandlessconflictpresumablybecausesubgroupfracturesdonotform.

Ideasforfutureresearchandlimitations

Therearemanypossibleextensionsforthiswork.Wesimplyexaminedtwolevelsofdiver-sity–lowandrelativelyhigh.Examiningdifferentdegreesofheterogeneitycouldbevalu-abletoexamineifandwhenfaultlinesdevelop(LauandMurninghan1998).Homogeneousgroupsaresimilarsofaultlinesdonotdevelop.Inhighlyheterogeneousteams,fewcom-monalitiesexistforthebasisofcreatingin-groupsandfaultlines(EarleyandMosakowski2000).However,in-groupsshouldbestrongundermoderatediversitysincesometeammemberswouldsharecharacteristicsthatcouldbethebasisforsubgroupidentity.Specif-icallycreatingconditionsthatvarythelevelofin-groupcreationwouldbevaluabletoseeiftheimpactofvirtualcommunicationchangeswiththelevelofin-groupcreation.

Diversityofnationalbackgroundandnationalcultureismulti-facetedandisachal-lengetoassessadequately.Wecreatedteamsthatwerediverseonfouraspects–coun-tryofbirth,nationality,nativelanguage,andtheindividualism/collectivismdimensionofHofstede’smeasureofnationalculture.Althoughwefeelthattheseprovidedreasonableindicationsofteam-levelculturaldiversity,thereareotheraspectsofnationalculturethatcouldbeexamined(e.g.,theotherfourHofstededimensions)andtheycouldbeexam-inedsingularlyortogethertoseeifcomplexinteractionsarepresent.Forexample,powerdistancecouldbeimportanttothewayteammemberinteract,ifthereisdiversityofsta-tuspresentintheteam.Ethnicitycouldalsobeexaminedsinceitmaybeamoredirect

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS403

assessmentofsurface-leveldiversity,whichisthebasisforthecreationofsubgroupsearlyinthelifeofateam.Alimitationofourworkisthatwedidnotdirectlymeasureeth-nicity;however,peoplewiththesamebirthcountryandlanguagearemorelikelytobeofthesameethnicbackground(althoughnotalwayssincemanycountrieshavemulti-pleethnicgroups).Researchdeterminingwhichaspectsofculturaldiversityarethemostimportanttoteamfunctioningwouldalsobevaluable.Ifoneortwoaspectswereiden-tifiedasbeingcritical,companiescouldusethisinformationtocreateadiagnostictoolusedtoidentifygoodcandidatesforworkinginvirtualteamsand/ortoidentifytrainingneeds.

Wedeliberatelystudiedateamthathadashortlife(aboutonehour)inordertoexamineifthepotentialforreductivecapabilitiesofelectronicmediawereofbenefitintheearlylifeofateam,asproposedbyCarteandChidambaram(2004).Whilewefeelthetimeframewechosewaslikelyshortenoughsuchthatasenseofteamidentitydidnotdevelop,futureresearchshouldconductlongitudinalstudies,measuringteamidentity,teamprocessesandperformanceatmultipletimepoints.Also,addingintheadditivecapabilitiesofadditionalmediaatdifferenttimeswouldhelpidentifytheoptimumtimetodoso.ThiswouldalsoallowamorecompletetestingofCarteandChidambaram’stheorysinceourstudyonlydealswithearlyteamlife.

Thetypeoftaskcouldbevariedaswell.Futureresearchcouldexamineataskwhereculturaldiversityisrelevanttoteamperformance.Thiswouldallowtheneteffecttobedeterminedandseewhenthepositivebenefitsofdeep-leveldiversityovercomethenegativeimpactsofsurface-leveldiversity,andifthesechangesareconsistentforteamsthatusedifferentcommunicationmodes.

OurstudyexaminedonebundleofelectroniccommunicationcapabilitiesforateamcommunicatingsynchronouslyandcomparedthistoteamscommunicatingF2F.Thetech-nologywechosewasbasedonwhatwehaveseeninvirtualteampractice;however,therearetechnologiesthathavestrongerreductivecapabilities.Researchthatvariesthereductivestrengthwouldbevaluablesincethiswouldhelpidentifytheoptimumsetoftechnologiesforuseearlyinateam’slife.Forexample,eventhoughcommoninpractice,itcouldbethatitisbettertointroducetelephoneconferencingatalaterstageinateam’sdevelopment.Therearealsoresearchopportunitiestovarydifferentotheraspectsofvirtuality(timeandspace)toexamineifthefindingsareconsistentfordifferenttypesofvirtualteams.

Conclusion

Diversityinnationalbackgroundandcultureiscommonintransnationalandvirtualteams.Understandingthepotentialadvantagesanddisadvantagesofthisdiversityisimportantfororganizations.Ourstudyfocusedonthepotentialnegativeaspectsofdiversityinnationalbackgroundandcultureearlyinthelifeofateamandfoundthatnegativeimpactsappearedtobereducedbyusingappropriatecommunicationmedia.ThisfindingsupportsCarteandChidambaram’s(2004)theorythatreductivecapabilitiesincollaborativetechnologiescanreducethesalienceofsurface-leveldiversity.Doingthisearlyinthelifeofadiverseteamisbeneficialsinceitallowsateamidentitytoformandreducesthetendencyofdiverseteams

404

STAPLESANDZHAO

tobreakintosubgroups.Muchmoreworkneedstobedonetounderstanddiversityinteamsandhowtomanageiteffectively.

Acknowledgements

FundingfromtheSocialScienceandHumanitiesResearchCouncilofCanadaandtheResearchProgramoftheSchoolofBusinessatQueen’sUniversityisgratefullyacknowledged.ThisstudybenefitedsignificantlyfromtheresearchassistanceofSayyidaJafferandadviceofJaneWebster.AnearlierversionofthepaperwaspresentedattheAcademyofManagement2004AnnualMeetingheldinNewOrleans.Wethanktheau-diencefortheirattentionandsuggestions.WealsothanktheeditorandthreeanonymousreviewersofthespecialissueofGDNfortheirsuggestionswhichledtoasignificantlyimprovedpaper.

References

Anderson,W.N.andS.R.Hiltz.(2001).“CulturallyHeterogeneousvs.CulturallyHomogeneousGroupsinDis-tributedGroupSupportSystems:EffectsonGroupProcessandConsensus,”Proceedingsofthe34thHawaiiInternationalConferenceonSystemSciences.

Baltes,B.B.,M.W.,Dickson,M.P.,Sherman,C.C.,Bauer,andJ.S.LaGanke.(2002).“Computer-MediatedCom-municationandGroupDecisionMaking:AMeta-Analysis,”OrganizationalBehaviorandHumanDecisionProcesses87(1),156–179.

Blau,P.M.(1977).InequalityandHeterogeneity:APrimitiveTheoryofSocialStructure,NewYork,FreePress.Bond,M.H.(1988).“FindingUniversalDimensionsofVariationinMulticulturalStudiesofValues:TheRokeachandChineseValueSurveys,”JournalofPersonalityandSocialPsychology55(6),1009–1015.

Bottger,P.C.andP.W.Yetton.(1987).“ImprovingGroupPerformancebyTraininginIndividualProblemSolving,”JournalofAppliedPsychology72,651–657.

Carte,T.andL.Chidambaram.(2004).“ACapabilities-BasedTheoryofTechnologyDeploymentinDiverseTeams:LeapfroggingthePitfallsofDiversityandLeveragingitsPotentialWithCollaborativeTechnology,”JournaloftheAssociationforInformationSystems5(11–12),448–471.

Earley,C.(19).“SocialLoafingandCollectivism:AComparisonoftheUnitedStatesandthePeople’sRepublicofChina,”AdministrativeScienceQuarterly34,565–581.

Earley,P.C.andE.Mosakowski.(2000).“CreatingHybridTeamCultures:AnEmpiricalTestofTransnationalTeamFunctioning,”AcademyofManagementJournal43(1),26–49.

Evaristo,R.(2003).“TheManagementofDistributedProjectsAcrossCultures,”JournalofGlobalInformationManagement11(4),58–70.

Green,S.G.andT.D.Taber.(1980).“TheEffectsofThreeSocialDecisionSchemesonDecisionGroupProcess,”OrganizationalBehaviorandHumanPerformance25,97–106.

Hambrick,D.C.,S.CanneyDavison,S.A.Snell,andC.C.Snow.(1998).“WhenGroupsConsistofMultipleNationalities:TowardsaNewUnderstandingoftheImplications,”OrganizationStudies19(2),181–205.Haslam,S.A.,C.McGarty,P.M.Brown,R.A.Eggins,B.E.Morrison,andK.J.Reynolds.(1998).“InspectingtheEmperor’sClothes:EvidencethatRandomSelectionofLeaderscanEnhanceGroupPerformance,”GroupDynamics2(3),168–184.

Hofstede,G.(1980).Culture’sConsequences:InternationalDifferencesinWork-RelatedValues.BeverlyHills,SagePublications.

Hofstede,G.(1983).“TheCulturalRelativityofOrganizationalPracticesandTheories,”JournalofInternationalBusinessStudies14,75–.

THEEFFECTSOFCULTURALDIVERSITYINVIRTUALTEAMSVERSUSFACE-TO-FACETEAMS405

Hofstede,G.(1994).VSM94:ValueSurveyModule1994Manual,IRIC,Tilberg.Accessibleathttp://feweb.uvt.nl/center/hofstede/VSM.html(August2005).

James,L.R.,R.G.Demaree,andG.Wolf.(1984).“EstimatingWithin-GroupInterraterReliabilitywithandWithoutResponseBias,”JournalofAppliedPsychology69,85–98.Janis,I.L.VictimsofGroupthinkBoston,MA:HoughtonMifflin,1982.

Jehn,K.A.G.B.Northcraft,andM.A.Neale.(1999).“WhyDifferencesMakeaDifference:AFieldStudyofDiversity,Conflict,andPerformanceinWorkgroups,”AdministrativeScienceQuarterly44,741–763.

Johansson,C.,Y.Dittrich,andA.Juustila.(1999).“SoftwareEngineeringAcrossBoundaries:StudentProjectinDistributedCollaboration,”IEEETransactionsonProfessionalCommunication42,286–296.

Johnson,D.W.andF.P.Johnson.(1994).Joiningtogether:GroupTheoryandGroupSkills(5thed.),Boston,Allyn&Bacon.

Kayworth,T.andD.Leidner.(2000).“TheGlobalVirtualManager:APrescriptionforSuccess,”EuropeanManagementJournal18,183–194.

Kessapidou,S.andN.C.Varsakelis.(2002).“TheImpactofNationalCultureonInternationalBusinessPerfor-mance:TheCaseofForeignFirmsinGreece,”EuropeanBusinessReview14(4),268–275.

Lau,D.C.andJ.K.Murninghan.(1998).“DemographicDiversityandFaultlines:TheCompositionalDynamicsofOrganizationalGroups,”AcademyofManagementReview23(2),325–340.

Maznevski,M.andK.Chudoba.(2001).“BridgingSpaceOverTime:GlobalVirtualTeamDynamicsandEffec-tiveness,”OrganizationScience11,473–492.

McDonough,E.F.,K.B.Kahn,andA.Griffin.(1999).“ManagingCommunicationinGlobalProductDevelopmentTeams,”IEEETransactionsonEngineeringManagement46(4),375–386.

McGrath,J.E.(1984).Groups:InteractionandPerformance,EnglewoodCliffs,N.J.,Prentice-Hall.

McLeod,P.L.,S.A.Lobel,andT.H.Cox.(1996).“EthnicDiversityandCreativityinSmallGroups,”SmallGroupResearch27(2),248–2.

Milliken,F.J.andL.L.Martins.(1996).“SearchingforCommonThreads:UnderstandingtheMultipleEffectsofDiversityinOrganizationalGroups,”AcademyofManagementReview21(2),402–433.

Mortensen,M.andP.Hinds.(2001).“ConflictandSharedIdentityinGeographicallyDistributedTeams,”Inter-nationalJournalofConflictManagement12(3),212–238.

Potter,R.E.andP.A.Balthazard.(2002).“VirtualTeamInteractionStyles:AssessmentandEffects,”InternationalJournalofHuman-ComputerStudies56,423–443.

Powell,A.,G.Piccoli,andB.Ives.(2004).“VirtualTeams:AReviewofCurrentLiteratureandDirectionsforFutureResearch,”DataBaseforAdvancesinInformationSystems35(1),6–36.

Rogelberg,S.G.andM.S.O’Connor.(1998).“ExtendingtheStepladderTechnique:AnExaminationofSelf-PacedStepladderGroups,”GroupDynamics:Theory,ResearchandPractice2,82–91.

Salk,J.E.andM.Y.Brannen.(2000).“NationalCulture,Networks,andIndividualInfluenceinaMultinationalManagementTeam,”AcademyofManagementJournal43(2),191–202.

Sarker,S.andS.Sahay.(2002).“InformationSystemsDevelopmentbyUS-NorwegianVirtualTeams:ImplicationsofTimeandSpace,”ProceedingsoftheThirty-FifthAnnualHawaiiInternationalConferenceonSystemSciences.

Sondergaard,M.(1994).“ResearchNote:Hofstede’sConsequences:AStudyofReviews,CitationsandReplica-tions,”OrganizationalStudies5(3),447–456.

Staples,D.S.,I.K.Wong,andA.F.Cameron.(2004).“BestPracticesforVirtualTeamEffectiveness,”in:DavidPauleen(ed.),VirtualTeams:Projects,ProtocolsandProcesses,IdeaGroupPublishing,HersheyPA,pp.160–185.

Stedham,Y.E.andJ.H.Yamamura.(2004).“MeasuringNationalCulture:DoesGenderMatter?”WomeninManagementReview19(5/6),233–243.

Straus,S.G.(1996).“Gettingaclue:TheEffectsofCommunicationMediaandInformationDistributiononParticipationandPerformanceinComputer-MediatedandFace-to-FaceGroups,”SmallGroupResearch27,115–142.

Thompson,L.F.andM.D.Coovert.(2003).“Teamworkonline:TheEffectsofComputerConferencingonPerceivedConfusion,Satisfaction,andPostdiscussionAccuracy,”GroupDynamics:Theory,ResearchandPractice7(2),135–151.

406

STAPLESANDZHAO

VandeVen,A.H.andD.L.Ferry.(1980).MeasuringandAssessingOrganizations.NewYork:JohnWiley&Sons.

VanRyssen,S.andS.H.Godar.(2000).“GoingInternationalWithoutGoingInternational:MultinationalVirtualTeams,”JournalofInternationalManagement6,49–60.

Watson,W.E.,K.Kumar,andMichaelsen.(1993).“CulturalDiversity’sImpactonInteractionProcessandPer-formance:ComparingHomogeneousandDiverseTaskGroups,”AcademyofManagementJournal36(3),590–602.

Williams,K.Y.andC.A.O’Reilly.(1998).“DemographyandDiversityinOrganizations:AReviewof40YearsofResearch,”inB.M.StawandL.L.Cummings(eds.),ResearchinOrganizationalBehavior(20),77–140.