麥肯錫為500強(qiáng)企業(yè)做的咨詢報(bào)告9

DoesNetworkingwithColleaguesMatterinEnhancingJobPerformance?

IreneHau-siuChow

DepartmentofManagement

TheChineseUniversityofHongKong

Shatin,N.T.

HongKong

Tel:(852)2609-7798

Fax:(852)2603-6840

Email:irene@.hk

&

IgnaceNg

FacultyofBusiness

UniversityofVictoria.

POBox1700,StnCSC

Victoria,BC

Canada

V8W2Y2

2

DoesNetworkingwithColleaguesMatterinEnhancingJobPerformance?

Thepersonalnetworksofmanagershavereceivedwidespreadattentioninthemanagementliteratureinrecentyearsbecausethesenetworksrepresentaformofsocialcapitalforachievingvariousgoalsineconomicandsociallife(Lin,2001).Havinganextensivenetworkisakeyelementtocareersuccess(Seibertetal.,2001).Otheradvantagesofhavinganextensivenetworkincludereceivingjoboffers(Granovetter,1995;Mouw,2002),gainingaccesstodecisionsinhigherauthority(Podolmy&Byron,1997),upwardmobility,andinformationexchange(Brass,1985).Intheentrepreneurshipliterature,thepersonalnetworkoftheentrepreneurisviewedasthemostimportantresourceoftheorganization(Johannisson,1990)asitisthroughthisnetworkthatthefirmacquiresthenecessaryscarceresourcesforitsgrowthandsurvival.

TheabovebenefitsofpersonalnetworksarebynomeanslimitedtotheWesternworld.InChinaandotherAsiansocieties,thesepersonalnetworksintheformofguanxirelationshipsalsoplayacriticalroleinthesuccessofindividualsandfirms.WhilethereexistimportantdifferencesbetweenguanxirelationshipsandWesterntypepersonalrelationships(Farhetal.,1998),theyarealsosimilarinthatbothguanxiresearchandpersonalnetworkresearchtreattheirrespectiverelationshipsasasourceofsustainablecompetitiveadvantage(Tsang,1998;OstgaardandBirley,1994).ConsistentwiththefindingsoftheirWesterncounterpart,guanxistudieshavedemonstratedthatpersonalrelationshipshavehelpedfirmsoperatinginChinaachievegreatersuccess(AbramsonandAi,1999;Luo,1997).Inspiteoftheseadvancesontheorganizational

3

impactofAsianpersonalnetworks,however,verylittleisknownastohowthesenetworksaffect

theindividualnetwork-holder(or‘ego’).AnexceptionhereisthestudybyBian(1997)whofoundthatindividualsinurbanChinadobenefitfromtheirpersonalnetworkswhenjobhunting.

Thestudyalsodemonstratedthatthecompositionofone ’snetwork,asmeasuredbystrongan

weaktiesinthiscase,hasdifferentialimpactsonthekindofjobsthattheegogets.What

remainsuntested,however,istheextenttowhichthisparticular,andothernetworkcharacteristicsaffecttheChinesejobholder.

Therearemanywaysinwhichindividualpersonalnetworksdifferfromeachother.Networkcompositionshavebeenconceptualizedinmanyways,varyingintermsofsize,gendercomposition,densityandotherdifferentiatingnetworkcharacteristics.SinceGranovetterseminalworkonstrongandweakties,thestrengthofnetworktiesisonenetworkcharacteristicthathasreceivedconsiderableattentionintheliterature(Granovetter,1995;Bian,1997).Whetherweaktiesaremoreimportantthanstrongtiesdependsonthenatureofinformationsharedbetweenthenetworkassociates(Bian,1997;Hansen,1999).Networkcentralityhasalsobeenthesubjectofseveralstudiesandtheevidenceshowsthatcentralityaffectspower(Brass,1984),decisionmaking(Friedkin,1993),andinnovation(Ibarra,1993).

Othernetworkcharacteristics,however,havebeensubjecttolessscrutiny.Crossetal.(2002),forexample,arguedthategoswhosenetworkconsistsofindividualsexternaltotheorganization(boundaryspanningtypenetwork)arelesslikelytoberecognizedandrewardedbytheirorganizations.Thislackofrecognitionshouldinturnleadtolowerperceivedjobperformance.Ontheotherhand,aboundary-spanningtypenetworkismorelikelytogeneratediverse

4

informationandsimilartothestrengthofweaktiesargument,thisshouldleadtohigherjobperformance.Whetherthereisempiricalsupportforthislineofreasoning,however,remainstobeseen.Similarly,whileitseemslogicalthatanetworkofhigherrankedcolleaguesshouldbeofgreatervaluetotheegothananetworkoflowerrankedcolleagues,wearenotawareofanyresearchthathastestedthisproposition.Thus,anobjectiveofthisstudyistoexaminehowthese

networkcharacteristicsaffecttheperceivedjobperformanceofChinesemanagers.Tothisend,weproposeamodeldrawnfrompaststudiesofnetworkoutcomesinwhichjobperformanceisaffectedbythosenetworkcharacteristicsaffectinginformationaccessibilityanddiversity.

Researchersingenderdifferencesinpersonalnetworkshaveshownthatthesedifferencesremainevenaftercontrollingforstructuralfactors(Moore,1990).Othershavenotedthatforsimilarlyrankedindividuals,maleshavebeenmoreabletoconverttheirpositionalauthorityintogreaternetworkaccessthantheirfemalecounterparts(Ibarra,1992).Theconsensusinthenetworkliterature,therefore,isthatmalesandfemalesdifferintermsofboththeirnetworkcompositionandnetworkaccess.Whetherthebenefitsofpersonalnetwork,ifany,varybetweenmalesandfemaleshasyettobeascertained.Inaddition,tothebestofourknowledge,wearenotawareofanyresearchthathasexaminedwhetherdifferencesinnetworkingbenefitsareaffectedbytherankof,andjobtypeheldby,theego.Thus,asecondobjectiveofthisstudyistoinvestigatewhethertheperformanceandnetworkcharacteristicslinkagesaremoderatedbythegender,jobandpositionalauthorityoftheego.

TheoryandHypotheses

5

Thebenefitsderivedfromone’spersonalnetworkarebaseduponthesocialexchangetheory(Blau,1964).Accordingtothistheoreticalframework,whenanindividualvoluntarilyprovides

abenefittoanother,thisactioninvokesasenseofobligationontherecipienttoreciprocateatalaterunspecifieddate.WhilethissocialexchangeframeworkisapplicabletobothWesternandChinesepersonalrelationships,itshouldbepointedoutthattheanticipatedrepaymentofthe

obligationisbasedontheequityruleforWesternersandtheneedsrulefortheChinese(Hwang,1987).Byadoptingtheruleofreciprocityappropriateforagivenculture,therecipientdevelopsaprocessofreciprocationandmutualexchangerelationshipswiththeirbenefactors.Fromthese

exchanges,stableandongoingpersonalrelationshipsaredevelopedandcanbecalleduponwhenneeded.

Studieslinkingsocialnetworkstojobperformance(Baldwin,1997;Brass,1981;Sparrowetal.,2001)positthatnetworksprovidetheegoaccesstoinformationwhichcancomeintheformofadvice,assistance,andproblemsolvinghelp.Overtime,thishelpstheegoacquirethenecessaryknowledgeandcompetenciestobesuccessfulatwork(Sparrowetal.,2001).Ourinteresthereistoidentifythosenetworkcharacteristicsthatcanleadtogreaterinformationaccessibility.

Whilenetworkcentralityisonecharacteristicoftenassociatedwithaccesstoinformation(Brass,1984;Ibarra,1993),othernetworkpropertiescanalsoleadtogreateraccessibility.

OstgaardandBirley(1994)arguedthatnetworksizeisakeymeasureofresourceandinformationaccessibility.Thatis,thepotentialamountoforganizationalinformationandresourcesthatonecanaccesstodependstoalargeextentonthenumberofcoworkersthattheegotalkstoregardingadviceandwhatisgoingonintheorganization.Themorecoworkers,

6

adjustedforthetotalnumberofcolleaguesinthedepartment,onehasdirectconnectionwiththroughone’spersonalnetwork,thegreateristheaccesstotheorganizationalinformation.Asidefromthesedirectnetworklinks,onecanalsoobtaintherequiredresourcesbynetworkingwithindividualswhoarewellconnected.Throughthem,onecanaccesstoawiderangeofindividuals.AsdemonstratedinCrossetal.(2002),amanagercangetaccesstomany

employeesthroughacentralconnectororinformationbrokerwithouthavingtocontacttheseemployeesdirectly.Thus,accessibilitytoinformationisnotonlyamatterofhowmanypeopleoneknowsbutalsoamatterofhowwellconnectedthesepeopleare.Giventhelinkbetween

networkcentralityandpositionalauthority(LincolnandMiller,1979),itthereforefollowsthatanetworkconsistingofhighrankedcolleaguesprovidesmoreaccesstoinformationthananetworkoflowerrankedcolleagues.

Asidefrominformationaccessibility,wearguethatdiverseinformationcanalsoimpactuponjobperformance.Theperformance-informationdiversitylinkhasnotreceivedmuchattentioninthenetworkliterature.Thisisasomewhatsurprisingomissionbecauseitisgenerallyacceptedinthemanagementliteraturediverseviewpointsandlessgroupthinkarepositivelyrelatedtoperformance(Errol,1989;Kochan,etal.2001;Reagans&Zuckerman,2001).Itthereforefollowsthatanetworkwhichprovidesthenetworkholderwithdiverseinformationislikelytobemoreperformanceenhancingthananetworkcontainingsimilarinformation.

Granovetter(1973)positsthatweaktiesserveasabridgebetweendifferentgroupsofnetworks.Bynetworkingwiththeseties,oneobtainsinformationandresourcesfromothernetworksthatis

notavailableinone ’simmediatenetworkcircle.Inotherwords,informationobtainedfrom

7

weaktiestendstobediverseinnaturewhereasinformationfromstrongtiesislikelytoberedundant.Thus,inpersonalnetworkswheretheegohasstrongtieswiththenetworkmembers(oralters),theinformationbaseismorelimitedascomparedtoanetworkwheretheegohasonlyweaktieswiththenetworkmembers.Asidefromthestrengthofties,informationdiversityisalsoinfluencedbytheextenttowhichthenetworksmembersarebeyondtheorganizational

boundaries.Agreaterproportionofnetworkmembersbeingnon-colleagueswouldsignifythattheegoisaboundaryspanner(Crossetal.,2002),abletosecureinformationbeyondtheconfinesoftheorganization.Conversely,morecolleaguesasnetworkingalterswouldindicatea

narrowerinformationbase.

Theaboveparagraphsthereforesuggestthatinformationdiversityisnegativelyrelatedto

proportionofcolleaguesinone’snetworkandtotheproportionofstrongtieswiththecoworkersinthenetwork.Informationaccessibilityishoweverpositivelyrelatedtonetworksizeandthe

managerialranksofcolleaguesinone ’snetwork.Giventhatjobperformanceispositively

relatedtoinformationaccessibilityanddiversity,itfollowsthat:

Hypothesis1:Jobperformanceispositivelyrelatedtonetworksizeandthemanagerialranksof

colleaguesinone’snetwork,andnegativelyrelatedtotheproportionofcolleaguesinthenetworkandthestrengthoftieswiththem.

Whilethereiscurrentlynoevidenceindicatingwhetherthebenefitsofnetworkingaffectmalesandfemalesdifferentially,onecaninferfromthementoringliteraturethatfromagivennetworkcomposition,femaleswillreceivelessbenefitsthanmales.Acommonlyviewheldinthis

8

literatureisthatcross-gendermentoringislesslikelytodeveloptoitsfullpotentialbecauseofthefearofsexualintimacyandofdamagingrumors(Feldman,1999).Thisinabilitytodevelopadeeperrelationshipmakesitmoredifficultforthementortosharecareer-enhancingadviceandinformationwiththesubordinate.Theinformationexchangedinsucharelationshipthereforetendstobegenericandoflesservaluetotherecipient.Extendingthisreasoningtonetworking,

crossgenderrelationshipsthereforeyieldlessvaluableworkrelatedinformationthansamegenderrelationship.

Existingresearchingenderdifferencesinnetworkshasshownthatforinstrumentalpurposes,womenprefertonetworkwithmalecolleagues(Ibarra,1992and1995).Thelatterhavehigherpositionalauthorityandgreaternetworkcentrality(Lincolnetal.,1979).Theyarealsoperceivedtobemorecompetentthanfemales(Bergeretal.,1977;Ely,1995).Theseargumentsandsupportingempiricalevidence(Ibarra,1992)thereforeindicateforinstrumentalties,womenaremorelikelytobeinvolvedincross-gendernetworkrelationships.Thistogetherwiththenotionthatcrossgenderrelationshipsyieldinformationoflessqualityimpliesthatinformationaccessibilityplaysareducedeffectonjobperformanceforfemales.Further,becauseofthelowerqualityofinformationreceived,anarrowerinformationbasewouldbemoredetrimentaltothejobperformanceoffemales.Thus,

Hypothesis2:Forfemaleegos,informationaccessibilityasmeasuredbynetworksizeandranksofcolleagueshasalesserimpactonjobperformance.Anarrowerinformationbase,asmeasuredbytheproportionofcolleaguesinthenetworkandstrongtieswithcoworkers,hasagreaternegativeimpactonperformance.

9

Asidefromthemoderatingeffectofgender,thelinkbetweenone

’snetworkcompositionan

performanceisalsoaffectedbyone ’smanagerialrank.Thetheoreticalunderpinningforthisis

basedonthepropositionthathigherlevelmanagersdodifferentthingsthanlowerlevelmanagers.Inastudyofwhatmanagersdo,Luthansetal.(1985)shadowed200managersovera

2-weekperiod.Theirsurveyfoundthatthemanagers’timeisspenton4setsofactivitieshumanresourcesmanagement(suchascoaching),traditionalactivities(suchasplanning),communicatingactivities,andnetworkingactivities.Atypicalmanageristhereforeinvolvedin

bothinternalandboundaryspanningactivities.Whathappenstothemanager

’sfocusas

movesuptheorganizationalhierarchyremainsunclear.Itishoweverreasonabletoassumethathigherlevelmanagersaremoreinvolvedinthedevelopmentoflong-termstrategiesandalongtheway,theyoftenbecometheinitiatorandchampionofchange.AspointedoutbyFarkasandWetlaufer(1996),designinglong-termstrategiesoftenrequirestheleadertospendadisproportionateamountoftimedealingwithmattersexternaltotheorganization.Becauseoftheincreasingimportanceofboundaryspanningrolesforhigherlevelmanagers,itbecomes

morecriticalforthesemanagerstoacquiretheirinformationfromabroaderbase.Inotherwords,informationdiversitytakesonamorecrucialroleintheperformanceofhigherlevelmanagers.Further,asthechangechampion,higherlevelmanagersneedtobeabletogainaccesstomorepeopleintheorganizationiftheyaretoconvincetheemployeesoftheneedsandbenefitsofchange.Thus,

Hypothesis3:Forhigherrankedmanagers,anarrowerinformationbaseasmeasuredbytheproportionofcolleaguesinthenetworkandstrongtieswithcolleagueswillhavealarger

10

negativeimpactonjobperformance.Conversely,greateraccessibilityasmeasuredbynetworksizeandtherankofcolleagueswillhavealargerpositiveimpactontheperformanceofhigherrankedmanagers.

Inadditiontomanagerialranks,thespecificfunctionalareaofthemanageralsoinfluences

whetherthatmanagertakesonamoreinternalorboundaryspanningrole.Whilethereislittleempiricalevidencetolinkspecificfunctionalareastotheinternal-externalrolesofamanager,itcanbearguedthatthoseinthemarketingandsalesareaaremorelikelytospendtheiractivities

externally,dealingwithcustomers,suppliersandotherexternalstakeholderswhencomparedtomanagersfromotherfunctionalareas.Similartotheargumentspresentedforhigherrankedmanagers,diverseinformationhasthereforeagreaterimpactontheperformanceofmarketingandsalesmanagers.Thus,

Hypothesis4:Formanagersinthemarketingfield,anarrowerinformationbaseasmeasuredbytheproportionofcolleaguesinthenetworkandstrongtieswithcolleagueswillhavealargernegativeimpactontheirjobperformance.

Data

Thesampleforthisstudyisbasedon121respondentswhoarefull-timeemployeesandenrolledinapart-timeMBAprograminBeijing,China.Therespondentsare,onaverage,30yearsold,and70%aremarried.Intermsofgenderdistribution,74.4%oftherespondentsaremales.Onaverage,theyhavehad7.7yearsofworkexperience,andhavebeenintheirorganizationsfor3.7

11

yearsandintheirpresentpositionfor2years.Individualsworkedinavarietyofjobsacrossaheterogeneouscross-sectionofoccupations.Theadvantageofhavingsuchaheterogeneoussampleisthattheparticipantsarenotuniformlyinfluencedbythecontextualconstraintsofanysingleorganization(Diefendorff,Brown,Kamin,&Lord,2002).

Personalnetworkinformationwasobtainedbyaskingrespondentstonameindividualsintheadvice,support,influenceandfriendshipnetworkties.Examplesofnamegeneratingquestions

are

“whomwouldyouapproachforimportantdecision?

”,

“Forneededadvice?

”,

“Cou

timeofcrisis? ”,and “Socializewithafterwork? ”Therespondentswereaskedtogenerate

5individualsforeachquestion,foramaximumnetworksizeof20associates.However,becausethesameassociatecanbenamedforseveralties,thenumberofnon-overlappingassociatesinthenetworkisactuallysmallerthanthemaximum,withanaveragenetworksizeof13associates.

Theself-perceivedperformancemeasuretogetherwiththedemographiccharacteristicsoftherespondent,andthesocialnetworkdatawerecollectedintwodifferenttimeperiodswithanintervalofthreemonths.Splittingthedatacollectionintotwotimeperiodswasaimedatreducingtheburdenofacompletingalongquestionnaire.PodsakoffandOrgan(1986)furtherarguedthatsplittingthedatacollectionovertimelowerstheeffectsofcommonmethodvariance,thusenhancingthequalityoftheresponses.

ThevariablesusedinthisstudyarepresentedinTable1.Thedependentvariable,perceivedjobperformance,wasadoptedfromBabinandBoles(1998)It.consistedof7self-reporteditemson

a7-pointLikertscale,rangingfrom “1”forstronglydisagreeto “7”forstronglydisa

12

itemsincludeI

“amatopperformer

”,

“havehigherproductivitythanothers

”,

“kn

tasks/duties

”,

“getbetterawards/bonusthanmost.

”Thealphacoefficientfortheseitemswa

Informationaccessibilityismeasuredintwoways.First,theextenttowhichanindividualisdirectlyconnectedtoothersintheorganizationismeasuredbythenumberofcolleaguesinone

networkadjustedforthenumberofemployeesworkinginthatperson’sdepartment(Direct)Second,informationaccessibilityiscapturedbytheextenttowhichanindividualisconnectedtothosewhoarethemselveswell-connected.Thisisoperationalizedbythenumberofhigher

rankedcolleaguesinone(Indirect).

’snetworkdividedbythenumberofcolleaguesinthenetwork

Informationdiversityisalsocapturedintwoways.First,thenumberofcolleagueswithwhomtheegohasstrongtieswithdividedbythenumberofcolleaguesinthenetworkisameasureofthestrengthoftiesbetweentheegoandtheassociates(Strong).Thisvariableisinverselyrelatedinformationdiversity.Second,thenumberofcolleaguesdividedbythenumberofindividualsin

one’snetworkmeasurestheinternalorientationofthenetworkholder(Internal).Ahighervalueforthisvariablesignifiesthatthenetworkismoreinternallyoriented.Theinformationgeneratedfromsuchanetworkisthereforelessdiverse.

Inouranalysis,weincluded5controlvariablescapturingthedemographicandworkcharacteristicsofthenetworkholder.Thedemographiccharacteristicsincludetheageandgender(Female)oftherespondent.Theworkcharacteristicsconsistofthenumberofyearstherespondenthasbeeninthecurrentposition(Position),andwhethertherespondentworksinsales

13

andmarketing(Marketing).Themanagerialrankoftherespondent(Rank)ismeasuredbythe

numberofmanagerialranksbelowtherespondentdividedbythetotalnumberofranksintheorganization.

’srankminusthenumberofranksabove,

Results

Totestthe4hypothesespresentedinthetheoreticalsection,werunaseriesofOLSregressions,withperceivedperformanceasthedependentvariable.Theindependentvariablesconsistedof

thecontrolandnetworkvariables,andanumberofinteractiontermsinvolvingthenetworkvariablesandthecharacteristicsoftherespondents.Theseinteractiontermsareusedtotesthypotheses2to4.TheresultsoftheregressionresultsarepresentedinTable2.

Theadditionofthenetworkvariablestothecontrolvariables(Model2ofTable2)didnotaddanysignificantexplanatorypowertovariationsinperformance.Noneofthenetworkvariableswasstatisticallysignificant.Hypothesis1isthereforenotsupported.Thatis,thecompositionof

one’snetworkdoesnotaffectone’sjobperformance.Whetherthisconclusionholdsregardlessofthedemographicandworkcharacteristicsoftherespondentrestsuponthetestresultsoftheremaininghypotheses.

Hypothesis2whichstatesthatcomparedtomales,femalesarelesslikelybenefitfromtheirnetwork.Empirically,thishypothesispredictsthattheinteractiontermsbetweenthe

variableandinformationaccessibility(asmeasuredbythe “direct ”and “indirect ”

shouldbenegativewhiletheinteractionwith(thelackof)informationdiversityshouldhavea

14

lessernegativeimpactonfemalejobperformance.TheresultsfortheinteractiontermspresentedinModel3indicatethatwhilemalesandfemalesaresimilarintermsofthereturnsfromnetworkdiversity,femalesarelesslikelytobenefitfromtheirdirectandindirectnetworkmembers.Hypothesis2isthereforepartiallysupported.

TheregressionresultoftheinteractiontermsinModel4impliesthatthelinkbetween(thelack

of)informationdiversity(asmeasuredbythe “internal ”and “strong”variables)andp

isaffectedbythemanagerialrankofthenetworkholder.Hypothesis3predictsthatthe

regressioncoefficientsforboththeinteractiontermsbetweenmanagerialrankandthe

and“strong”variablesshouldbenegative.Theresultsfortheinteractionbetweentheand“strong”variableshoweversuggestsotherwise.Forthisreason,then,itisarguedthathypothesis3isonlypartiallysupported.

Hypothesis4statesthatinformationdiversityismorebeneficialtothoseinthemarketingfieldswhentothoseinotherfields.Thehypothesisthereforepredictsthattheinteractionterms

betweenthe “marketing”variableandthenetworkingvariables “internal ”and “stro

haveanegativeimpactonjobperformance.Theregressionresults,presentedinModel5,showthatthecoefficientsontheinteractiontermsarenegativeaspredictedbutonlyoneisstatisticallysignificant.Forthisreason,wearguethattheevidenceonlypartiallysupportshypothesis4.

DiscussionandConclusion

15

Thisstudyexaminedtheextenttowhichdifferentformsofnetworkingwithcolleaguesaffecttheperceivedperformanceofthenetworkholder.Theresults,basedonasampleofmanagersinBeijing,showthatthe4networkmeasuresconsideredinthestudydidnothaveanyeffectonperformance.Thereareseveralreasonsastowhythismaybethecase.Itisgenerallyaccepted

thatinthecaseofChina,andformostAsiansocieties,socialnetworksareguidedbyguanxirelationships(Bian,1997;Hwang,1987)andthattheunderlyingprinciplesofguanxiaredifferentfromthoseofWesterntyperelationships(Farhetal.,1998).Thus,whatmattersin

Chinaisguanxi,andnotnetwork,relationships.Sinceourmeasuresofnetworkingreflectmoreofthewesternrelationships,itcanbearguedthatthesemeasureshavenoeffectontheperformanceofChinesemanagers.

Ournetworkmeasuresareproxiesfortheextenttowhichthenetworkholderhasaccesstoinformation,andto(thelackof)diversityininformation.Whileaccessibilityto,anddiversityin,informationisdesirable,whatisimportanttojobperformanceisthequalityofinformationreceived.Intheabsenceofanyevidencelinkinginformationaccessibilityanddiversitytoinformationquality,onecanthereforearguethatforthisreason,thenetworkingmeasuresusedinthepresentstudyhavenoeffectonjobperformance.

Arelatedexplanationforthenon-significanceofournetworkmeasuresisthepropositionthatitisnotinformationaccessibilityanddiversitypersethatmatters.Ratheritishowthisinformationbaseisbeingputtousebythenetworkholder.Itthereforefollowsthatthelinkagebetweenperformanceandinformationisaffectedbythecharacteristicsofthenetworkholder.

16

Thepartialempiricalsupportforhypotheses2to4lendssupporttothislastexplanationofthenon-significanceofthenetworkmeasures.

Theempiricalresultsonthemoderatingeffectofgenderontheperformance-informationlinkshowthatcomparedtomales,femaleswithgreateraccesstoinformation,whetherdirectlyor

indirectly,arelesseffective.Thisisconsistentwiththepredictionofhypothesis2whichpositedthatwomenhaveaccesstolesserqualityinformationbecausetheyhavetorelyonmalesforinstrumentalties.Analternativeexplanationisthatbyseekingmoreinformation,oneis

indirectlytellingothersofone ’signorance(BorgattiandCross,2003).Womenseekingmore

informationarethereforeunwittinglyreinforcingthenegativestereotypethattheyareless

competentintaskrelatedactivities(Bergeretal.,1977). Giventhatmentorspreferprot

whoareyoungerversionsofthemselves(RagginsandCotton,1999),theywillbelesswillingto

takethosewomenperceivedtobelesscompetentastheirprot égés.Withoutamentor,and

togetherwiththeirreducedreputation,andpossiblyself-esteem,thesewomenthereforeendupwithalowerjobperformance.

Ourresultsalsoshowthathigherrankedmanagerswhosenetworkisinternallyoriented(thatis,agreaterproportionoftheirnetworkassociatesisfromwithin,asopposedtooutside,theorganization)havelowerjobperformance.Thisisconsistentwiththepredictionofhypothesis3inthathigherrankedmanagersinlightoftheirboundaryspanningrolesneedtohaveamoreexternallyorientednetworktobeeffective.Contrarytothestrengthofweaktiesargument,however,ourresultsshowthatforhigherrankedmanagers,networkingwithstrongtiesarepositivelyrelatedtojobperformance.Aplausibleexplanationforthisfindingisbasedon

17

Hansen’s(1999)studywhichfoundthatforthetransferoftacitandcomplexknowledgeacrossdepartments,strongtiesarem