Rudman Corinne A MossRacusin Julie E Phelan Sanne Nauts Rutgers University New Brunswick USA Yale University USA Langer Research Associates USA Behavioural Science Institute Radboud University Nijmegen Netherlands abstract article info Article hi ID: 22413
Download Pdf The PPT/PDF document "Status incongruity and backlash effects ..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Statusincongruityandbacklasheffects:DefendingthegenderhierarchymotivatesprejudiceagainstfemaleleadersLaurieA.Rudman,CorinneA.Moss-Racusin,JulieE.Phelan,SanneNautsRutgersUniversityNewBrunswick,USAYaleUniversity,USALangerResearchAssociates,USABehaviouralScienceInstitute,RadboudUniversityNijmegen,NetherlandsarticleinfoArticlehistory:Received10April2011Revised30August2011Availableonline15October2011 BacklasheffectSexdiscriminationGenderstereotypeGenderprejudiceSystemjusticationtheoryImpressionmanagement JournalofExperimentalSocialPsychology48(2012)165 Correspondingauthorat:DepartmentofPsychology,TillettHall,Rutgers,theStateUniversityofNewJersey,53AvenueE,Piscataway,NJ08854-8040,USA.E-mailaddress:(L.A.Rudman). seefrontmatter©2011ElsevierInc.Allrightsreserved. ContentslistsavailableatSciVerseScienceDirectJournalofExperimentalSocialPsychologyjournalhomepage:www.elsevier.com/locate/jesp thatwomenarepenalizedspecicallyforstatusviolationsbecausedefendingthegenderhierarchyisakeymotiveforbacklash.Finally,wepresentfourexperimentsthatprovidesupportforthestatusincon-gruityhypothesis.GenderstereotypesasgenderrulesGenderstereotypesarenotmerelydescriptivebutalsoprescriptive,consistingofrulesconcerninghowmenandwomenshouldbehaveEagly&Karau,2002;Fiske,1998;Fiske&Stevens,1993).Agenticattri-butes(e.g.,assertive,competitive,andindependent)areprescriptiveformen,whereascommunalattributes(e.g.,warm,kind,andsupportive)areprescriptiveforwomenplausiblybecauseagencyisrequiredforleadershipandcareersuccess,whereascommunalityisrequiredforcar-ingforthewelfareofothers(Burgess&Borgida,1999;Eagly,1987;Prentice&Carranza,2002;Williams&Best,1990Genderrulesalsoconsistofcharacteristics,rulescon-cerninghowmenandwomenshouldnotbehave.Proscriptionsarerelativelynegativequalitiesthatareprohibitedforonlyonegen-der.Forexample,dominantmasculinetraits(e.g.,controllingandarrogant)areproscribedforwomenbuttoleratedformen,andweakfemininetraits(e.g.,weakandnaive)areproscribedformenbuttoleratedforwomen(Prentice&Carranza,2002).Fromtheseex-amples,itseemslikelythatproscriptionsservetoreinforcethegen-derhierarchy(e.g.,thatmenoughtnottobeweakbecauseitislowinstatusandwomenoughtnottobedominantbecauseitishighinstatus).Therefore,aconsiderationofhowgenderrulesalignwithsta-tuscharacteristicsshouldhelptoexplainwhyagenticfemaleleadersareperceivedashighlycompetentbutsociallyunattractive(i.e.,whytheyarerespectedbutnotliked;e.g.,Heilmanetal.,2004;Rudman&Phelan,2008).Giventhatwomenencounterbacklashwhentheyhavemerelydemonstratedagency(andnotnegativetraits),whatisitaboutcompetent,ambitiouswomenthatputsthematriskforsocialrejection?BecauseprejudiceagainstagenticwomenfullyaccountsforwhytheyarelesslikelytobehiredasleadersthanagenticmenRudman&Glick,1999;2001),itisimperativetoanswerthisques-tiontoilluminatebarrierstogenderequality.Genderstereotypesandstatusrststepwastoexaminetheextenttowhichgenderrules(prescriptionsandproscriptions)arealignedwithstatuscharacteris-tics.Todoso,Study1consistedofalarge,onlinesurveydemonstrat-ingthatmaleagencyprescriptionswerealignedwithhighstatus,andmaleweaknessproscriptionswerealignedwithlowstatus.Thus,whatmenshouldbeishighinstatusandwhattheyshouldnotbeislowinstatus,andbothrulesreinforcethegenderhierarchy.Ofmoreimportance,femaledominanceproscriptionswerelinkedtohighstatus,whereasfemalecommunalityprescriptionsconsistedoftraitsthatwere,onaverage,statusneutral.Thus,whatwomennotbeishighinstatus,whereaswhattheyshouldbedoesnotneces-sarilyreinforcethegenderhierarchy.Althoughthendingthatwomen'sprescriptionsarestatusneutralmightseemsurprising,pastresearchhasshownthatagencyislinkedtohighstatusmoresothancommunalityislinkedtolowstatus(Conway,Pizzamiglio,&Mount,1996).Moreover,agencyandcommunalitytendtobeposi-tivelyassociatedinperceptionsofindividuals(e.g.,Henik&Tzelgov,1985;Judd,James-Hawkins,Yzerbyt,&Kashima,2005;Rosenberg,Nelson,&Vivekananthan,1968),andinperceptionsofingroupmem-bers(Fiske,Cuddy,Glick,&Xu,2002),suggestingthattheyarenotbi-polarqualities,orlikelytobeoppositionalvis-à-visstatus.TheresultsofStudy1aredescribedindetailbelow,buthereweemphasizethattheyconrmedthatagencyisalignedwithhighsta-tus,butthatcommunalityisnotnecessarilylowinstatus(etal.,1996;Hoffman&Hurst,1990).Instead,womenareproscribedagainsthighstatus,dominancedisplays(e.g.,dominant,controlling,andarrogant)traitsreservedforleadersandmen.Theseprovideakeytothepuzzleofwhichstereotypesarelikelytobecul-pableinbacklashagainstagenticwomen,andwhy.Thatis,theybegintopointtowardstatusincongruitiesasacriticalcomponentofbacklash.ThestatusincongruityhypothesisResearchersdistinguishbetweenstatusbasedonachievement(earningyourwaytothetop)andascribedstatusbasedonpersonalcharacteristics(e.g.,sex,race,andage;Berger,Webster,Ridgeway,&Rosenholtz,1986;Ridgeway,2001;Ridgeway&Bourg,2004).Be-causeoftheirgender,womenareautomaticallylinkedtolowstatusRudman&Kilianski,2000somuchsothatwhentheyenteranoc-cupationinlargenumbers,itsprestigecandropsignicantly(Nieva&Gutek,1981;Touhey,1974).Thestatusincongruityhypothesis(SIH)proposesthatwomenwhopossessorpursuepoweraredefactosta-tusincongruent,butparticularlywhentheirbehaviorviolatesstatusexpectations.Giventhatagencyishighinstatus,femaleagencyisdis-crepantwithwomen'slowascribedstatus,andthisstatusincongruityelicitsbacklash.Byexhibitingmasculinecompetencies,agenticwomenunderminethepresumeddifferencesbetweenthegenders,anddiscreditthesysteminwhichmenhavemoreaccesstopowerandresourcesforostensiblylegitimatereasons.Thatis,agenticwomenshouldincurpenaltiesbecausetheythreatenthegenderhier-archy.Asaresult,women'sperceivedstatusviolations(asopposedtoanytypeofgenderroleviolation)shouldaccountforbacklasheffects.ThedominancepenaltyAccordingtoStudy1,femaledominanceproscriptionsarethegenderrulesforwomenthataremoststronglyalignedwithstatusandthus,reinforcethegenderhierarchy.Therefore,agenticfemaleleadersshouldbeviewedasextremeondominanttraits(e.g.,domi-nant,controlling,andarrogant),comparedwithagenticmaleleaders.Termedthedominancepenalty,weexpectedtheseratingstofullyac-countforwhyagenticwomenarelikedlessthanagenticmen(i.e.,prejudice).Ifso,thendefendingthegenderstatusquowillberevealedasakeymotiveforbacklash.Interestingly,Eagly,Makhijani,andKlonsky(1992)invokedadominancepenaltyintheirmeta-analysisinvestigatingevaluationsofmaleandfemaleleaders(p.5):Inadditiontoatendencyforwomeninleadershiprolestobeevalu-atednegatively,theirbehaviormayberegardedasmoreextremethanthatoftheirmalecounterpartsthatis,asmoredominantandcontrolling,and,ingeneral,asembodyingahigherlevelofproto-typicalleadershipqualities.Thisperceptionoffemaleleadersasmoreextremethantheirmalecounterpartswouldbelikelytooccurtotheextentthatfemaleleadershipbehaviorsarequitediscrepantfrompeople'sstereotypesaboutwomenandarethereforeperceptuallycontrastedfromthesestereotypes(seeManis,Nelson,&Shedler,ThisanalysisforeshadowsakeypredictionoftheSIH,withtwoimportantcaveats.First,leadershipaloneisnotsufcienttoprovokebacklashbecausefemaleleaderswholowertheirstatus(e.g.,byadoptinginclusivestyles)areevaluatedfavorably(Eaglyetal.,1992;Rudman&Glick,2001).Bycontrast,femaleagencyintheab-senceofleadershipcanelicitbacklash(forareview,see),suggestingthatstatusincongruencydrivesbacklashmoresothanoccupationalroleincongruency.Second,theSIHproposesamo-tivational,ratherthanacognitive,explanationforthedominancepenalty(protectingthegenderhierarchy),whereasEaglyetal.proposedthatperceptualcontrasteffectswouldyieldex-tremelyhighratingsonprototypicalleadershipqualitiessuchasdominanceandagency.Specically,behaviorsthatarediscrepantwiththestereotypeshouldbeperceptuallycontrastedL.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 [awayfromthestereotype]Manisetal.,1988,p.28).Indeed,ifagenticfemaleleadersareviewedashighlydiscrepantwomen,contrasteffectsshouldalsoyieldextremelylowratingsoncommunalityandweakness(maleproscriptions).However,theSIHpredictsthatratingsofagency,communality,andweaknesswillnotaccountforprejudiceagainstfe-maleleaders,therebyrulingoutacognitiveexplanationforbacklash.Instead,itpredictsthatonlythedominancepenaltywillaccountfor(i.e.,mediate)prejudiceandthus,iskeytounderstandingwhyleader-shiprolesaremorechallengingforwomenthanmen.SystemjusticationmotivesforbacklashButhowdoesthedominancepenaltyrevealthemotiveforback-lash?AccordingtotheSIH,statusincongruentwomenjeopardizeex-tantnormsforpower,forcingareconsiderationofthegenderstatusquo(Ridgeway,2001;Ridgeway&Bourg,2004).Becausemotivestole-gitimizeexistingsocialstructuresarepervasiveandoftennonconsciousJost&Banaji,1994;Jost,Banaji,&Nosek,2004;Rudman,Feinberg,&Fairchild,2002),peoplemayinstinctivelyresistpowerfulwomenRudman&Kilianski,2000).Thus,motivestodefendthegenderhierar-chymayresultinperceptionsthatagenticfemaleleadersaretoopowerful,asreectedbythedominancepenalty,whichisthenusedtojustifyprejudice.Consistentwiththisview,epithetsforpower-fulwomenoftencastthemasdestroyersofmalevirility(e.g.,ball-breakerandcastratingbitch),signalingtheextenttowhichwomenareexpectedtoyieldtomeneconomically,politically,andsexuallyKanter,1977).Women'sascribedstatus(aswomen)isincongruentwiththeirachievedstatus(asleaders),aswellaswiththerequisiteagencydemandedofleaders.Bycontrast,agenticmaleleadersarenotatriskforstatusviolationsbecausetheirascribedstatus(asmen),achievedstatus(asleaders),andtherequisiteagencydemandedofleadersarecongruent.Inessence,agenticmaleleaderssupportthegen-derhierarchy,whereasagenticfemaleleaderssubvertit.Judgingonlythelatterastoopowerful(i.e.,dominantandcontrolling)isexpectedtosignifyasystem-justifyingmotiveforrejectingwomenwhodisruptthestatusquo.Moreover,pinpointingthedominancepenaltyasthesolemediatorofprejudice(rulingoutextremeratingsonagency,com-munality,andweakness)willsupportourclaimthatitisusedtojustifybacklashandthereby,todefendmalehegemony.Thepresentresearchsoughttotestthiskeymotiveforbacklashinmultipleways,asoutlinedbelow.OverviewoftheresearchandhypothesesStudy1examinedtheoverlapbetweengenderstereotypesandstatustolaythefoundationfortestingpredictionsderivedfromtheSIH.Studies24usedahiringparadigm,inwhichagenticapplicantsviedforleadershiprolesthatvariedwithrespecttogendertypicality(Englishprofessor,pretestedasgenderneutral,andmarketingman-ager,pretestedasmaledominated;seeFootnotes1and3).Weexpectedagenticmaleandfemaleapplicantstobeviewedassimilar-lyqualiedforthejob,becausestrongindividuatinginformationde-featswomen'srsthurdle(Eagly&Karau,2002).Nonetheless,agenticwomenshouldencounterprejudice,whichshouldfullyac-countforhiringdiscrimination(Rudman&Glick,1999,2001),reingfemaleleaders'secondhurdle(i.e.,backlash;Eagly&Karau,TotesttheSIH,applicantsinStudies24wereratedonthefoursetsofgenderrules(i.e.,maleandfemaleprescriptionsandproscriptions).Weexpectedonlyfemaledominanceproscriptionswould(1)distin-guishbetweenagenticmenandwomen,and(2)accountforprejudiceagainstagenticwomen.Theseresultswouldshow,remarkably,thathigh-poweredtraitsthatareviewedasmoretypicalofmen(indeed,arereservedforthem)caninfactbeattributedmoretowomenwhenwomenarestatusincongruent(i.e.,agenticleaders).Astheonlysetofrulesconsistingofhighstatustraitsprohibitedforwomen,femaledominanceproscriptionsshouldbeusedtojustifybacklashandthus,todefendthegenderhierarchy.Studies35soughtadditionalevidenceforthesystem-justifyingmotiveunderlyingbacklash.InStudy3,weexpectedthatperceiverswhoendorsedthegenderhierarchyasjustandfair(i.e.,gendersystem-justiJost&Kay,2005)wouldbemostlikelytoadminis-terthedominancepenalty.Bycontrast,gendersystem-justishouldnotpenalizewomenasinsufcientlycommunalbecausethisruleisnotalignedwithstatus.InStudy4,wemanipulatedsystemthreattoprovideacausaltestoftheSIH'sclaimthatbacklashisusedtoprotectthegenderhierarchy(Kayetal.,2009).InStudy5,randomlyassignedmaleandfemaleconfederatestoaleader-shippositionwhilemanipulatingtheiragency.Becauseagencyisnec-essaryforleadershipsuccess,weexpectedparticipants(assignedtobesubordinates)tosabotageonlyfemaleleaderswhowerehigh(notlow)onagency.Thisisbecauseonlywomenwhoeffectivelychallengethegenderhierarchy(i.e.,whoexhibitmasculinecompe-tencies)shouldbeatriskforbacklash.Inconcert,theseresultswouldbolstertheSIH'sclaimthatdefendingthegenderstatusquoisakeymotiveforbacklash.Study1Study1examinedtheextenttowhichprescriptiveandproscrip-tivegenderstereotypesarealignedwithstatus.Todoso,wecompiled64traitsandadministeredoneofsixsurveystoparticipantsinordertodetermine(a)howstereotypicalthetraitswereformenorwomen,(b)theextenttowhicheachtraitfellonaprescriptive-proscriptivecontinuum(i.e.,reectedgenderrules)formenorwomen,and(c)theextenttowhicheachtraitwascharacteristicofhighstatusorlowstatuspeople.Byisolatingeachtypeofjudgment,wesoughttoreducecontexteffectsanddemandinordertoobtainrelativelypureestimatesofthedegreetowhichgenderrulesoverlapwithParticipantsVolunteers(=832;406men,415women,and11whodidnotindicatetheirgender)participatedinexchangeforpartialfulofanIntroductoryPsychologycourserequirement.Ofthese,366(44%)wereWhite,259(31%)wereAsian,77(9%)wereLatino,63(8%)wereBlack,and67(8%)reportedanotherethnicity.MaterialsandprocedureWebeganwiththetraitsshownbyPrenticeandCarranza(2002)toreectgenderrules.Afteraddingtraitsoftenusedinbacklashre-search(e.g.,performswellunderpressureissensitivetoothers')andeliminatingsynonyms,weobtained64traitsusedtocre-atesixseparatesurveys.Thetwogenderrulessurveysaskedpartici-pantstoindicatehowdesirableitisinAmericansocietyforawoman[man]topossesseachofthefollowingcharacteristicsonascalefrom1(notatalldesirable)to9(verydesirable).Thetwosurveysaskedparticipantstoindicatehowcommonortyp-icalyouthinkeachofthefollowingcharacteristicsisinsomeonewhohashigh[low]statusonascalefrom1(notatalltypical)to9(.Finally,thetwogendertypicalitysurveysaskedparticipantsindicatehowcommonortypicalyouthinkeachofthefollowingcharacteristicsisinwomen[men]inAmericansocietyonascalefrom1(notatalltypical)to9(verytypicalAllmeasureswereadministeredonline.Afterreadingtheconsentform,participantswererandomlyassignedtocompleteoneofthesixsurveys.Aftercompletingthesurvey,participantsweredebriefedandL.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 ResultsanddiscussionPrescriptivestereotypesWedenedmaleprescriptionsastraitsthatwereratedabove6onthe1(notatalldesirable)to9(verydesirable)scaleformen,andthatwhencomparedtothedesirabilityratingforwomen,alsohadagenderdifferenceeffectsizegreaterorequalto=.40.ThetophalfofTable1showsthese16traits,rankedbyprescriptivescores.Ascanbeseen,men'sprescriptionsstronglyreectagency(=.75,range=.40to1.12).Wedenedfemaleprescriptionsastraitsthatwereratedabove6forwomen,andthatalsohadagenderdifferenceeffectsizelessthanorequalto.40.These16traitsareshowninthebottomhalfofTable1,rankedbyprescriptivescores.Ascanbeseen,thesetraitspri-marilyreectcommunality(.76,range:1.12toTable1'sfourthcolumnshowstheeffectsizefortypicality(posi-sindicatetraitsratedasmorestereotypicalformen;negativesindicatetraitsratedasmorestereotypicalforwomen).Ascanbeseen,men'sagencyprescriptionsweregenerallyratedasstereotypi-callymale(=.53),andwomen'scommunalityprescrip-tionswereallratedasstereotypicallyfemale(ThelastcolumnofTable1provideseffectsizesforthedifferencebe-tweenhighstatusandlowstatusratings.Positivescoresreecttraitslinkedtohighmorethanlowstatuspeople;negativescoresreectthereverse.Table1revealsthatallofmen'sprescriptionswerehighstatustraits(status=1.26,range=.43to1.60).Bycontrast,women'sprescriptionsreectedamix(status=.08,range=.80to1.20).Fouroftheirprescriptionswerelowstatustraits(emotional,warm,interestedinchildren,andhumble).However,fourtraitswereratedhighinstatus(cheerful,enthusiastic,excitable,andattendstoappearance).Theremainingeighttraitswereneutralinstatus(sensitivetoothers,goodlistener,cooperative,friendly,supportive,po-lite,helpful,andlikeable).Theseresultssuggestthatmenshouldenacthighstatus,agentictraitswhereaswomenshouldenactcommunaltraitsthat,onaverage,arestatusneutral.ProscriptivestereotypesWedenedmaleproscriptionsastraitsthatwereratedbelow4ondesirabilityformen,andthatalsohadagenderdifferenceeffectsizegreaterthanorequalto=.40.ThetophalfofTable2these10traits,rankedbyproscriptivescores.Ascanbeseen,men'sproscriptionsreectvulnerabletraitsassociatedwithweak-ness(=.87,range=.56to1.12).Theseattributesareperceivednotonlyasfeminine(1.05),butalsoaslowinsta-tus(.76,range=1.32to.05);onlytwowerestatusneutral(melodramaticandmoody),andnonewerehighinstatus.Similarly,wedenedfemaleproscriptionsastraitsthatwereratedbelow4forwomen,andthatalsohadagenderdifferenceeffectsizelessthanorequalto.40.ThebottomhalfofTable2showsthese13traits,rankedbyproscriptivescores.Ascanbeseen,thesetraitspri-marilyconsistofdominance-relatedattributesthatareprohibitedforwomen(.68,range=1.03to.41),andperceivedasmoretypicalofmen(typicality=.53),withtheexceptionofdemandingandself-centered,whichweregenderneutral.Femaledominancepro-scriptionswereprimarilyhighinstatus(=.73,range=.47to1.42).Onlytwowerelowstatustraits(rebelliousandangry),andonlyonewasstatusneutral(cynical).Thesendingssuggestthatwomencannotenactstatus-enhancingtraitsthatarereservedformenandleaders(e.g.,dominant,controlling,andarrogant),whereasmencannotenactstatus-attenuatingtraitsthatareacceptableforwomen(e.g.,emotional,naive,andweak).Insummary,Study1revealedcorrespondencebetweendomi-nanceproscriptions(whatwomenshouldnotbe)andhighstatus Table1Prescriptivetraitsformenandwomen(Study1).TraitPrescriptiveMaleFemaleTypicalityStatusdMMddMen'sprescriptionsCareeroriented1.127.745.74.491.57Leadershipability1.097.865.89.791.45Aggressive1.036.163.911.36.43Assertive1.017.265.20.781.39Independent.987.675.57.651.23Businesssense.977.395.76.861.60Ambitious.957.956.28.351.37Hardworking.808.086.78.13.91Workswellunderpressure.747.396.05.651.26Self-starter.586.265.13.531.18Intelligent.557.676.78.041.08Analytical.486.715.83.021.29Highself-esteem.487.296.56.741.59Persuasive.416.445.74.041.30Competitive.407.675.281.081.43Competent.407.176.48.141.12Women'sprescriptions1.123.876.511.036.077.99Interestedinchildren1.005.927.82Sensitivetoothers1.005.487.52Goodlistener.896.147.821.18.07Cheerful.876.067.58.83.57Enthusiastic.835.717.06.87.91.765.696.95.91.42.756.297.50.45.11.716.777.76.70.09Supportive.696.697.75.73.06.576.907.74.62.04.565.776.80Attendstoappearance.466.687.51.341.20.456.957.58.67.25Likeable.427.007.74.47.15.Positivescoresreectstrongerprescriptionsortypicalityformenthanwomen,orstrongertypicalityforhighthanlowstatuspeople.Negativescoresreectthereverse.Conventionalsmall,medium,andlargeeffectsizesforare.20,.50,and.80,respectively(Cohen,1988). Table2Proscriptivetraitsformenandwomen(Study1).TraitProscriptiveMaleFemaleTypicalityStatusdMMddMen'sproscriptionsEmotional1.123.876.51Naive1.032.354.55Weak.971.853.96Insecure.912.294.08Gullible.892.804.67Melodramatic.882.874.80Uncertain.802.784.13Moody.782.674.44.71.05Indecisive.742.814.31Superstitious.563.164.15Women'sproscriptions1.036.163.91.431.36.985.373.29.891.21Dominating.945.743.54.971.42.764.612.931.111.08.695.023.64.66.655.162.96.151.24.654.412.96.64.59.654.143.82.71.615.203.88.421.33.554.743.63.42.65Coldtowardothers.513.432.49.38.35.414.203.21.141.05.413.943.27.28.12.Positivescoresreectstrongerproscriptionsortypicalityformenthanwomen,orstrongertypicalityforhighthanlowstatuspeople.Negativescoresreectthereverse.Conventionalsmall,medium,andlargeeffectsizesforare.20,.50,and.80,respectively(Cohen,1988).L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 traits,whereascommunalityprescriptionsdonotnecessarilyrein-forcethestatusquo.ThesendingslaythegroundworkfortheSIH'spredictionthatthedominancepenalty(i.e.,extremeratingsonstatus-enhancingtraitsthataretypicalofleadersandmen,butpro-hibitedforwomen)willbeusedtojustifyprejudiceagainstagenticfemaleleaders.ThegoalofStudy2wastotestthisprediction.Study2ParticipantsreadrecommendationlettersforcandidateseligibleforpromotiontoEnglishprofessoratYaleUniversity.Thejobwaspre-testedtobegenderneutral.Candidateswerelaudedasinternationallyrecognizedauthorsandliterarycriticswhowerehighlyintelligent(e.g.,winnersoftheMacArthurGeniusAward).However,theauthorfurtherdescribedthecandidate'sstyleasaliterarycriticassomewhatcontro-versialbecauseitwaseitheragentic(brutallyhonest)orcommunal(overlypolite).Inbothcases,theauthordefendedthecandidate'sstyle,claimingthatagentictargetswerehighlycriticalinordertomaintainthehighstandardsoftheeld,whereascommunaltargetswereexcessivelydiplomaticinordertoprotectauthors'fragileegos.Consistentwithbacklasheffects,theagenticwomanshouldberatedassimilarlycompetent,butloweronlikingandhiringdimensionsthantheagenticman,andprejudiceshouldmediatehiringdiscrimina-tion(Rudman&Glick,1999,2001).TosupporttheSIH,weexpectedtheagenticwomantoberatedhigheronfemaledominanceproscriptionsthanacomparableman,andthatthedominancepenaltywouldmedi-atetargetgenderdifferencesinliking(i.e.,prejudiceagainstfemaleagency).Toruleoutperceptualcontrasteffects,noothergenderrulesshouldsignicantlydifferforagenticwomenandmen.ParticipantsVolunteers(=178;114women)participatedinexchangeforcredittowardtheirIntroductoryPsychologyresearchparticipationrequirement.Ofthese,103(58%)wereWhite,45(25%)wereAsian,12(7%)wereHis-panic,8(4%)wereBlack,and10(6%)reportedanotherethnicidentity.Candidatestylemanipulation.Tomanipulatethecandidates'critiquingstyle,fourrecommendationletterswerecreated,foreitherEdwardorEmilyMullen.Theletterswereostensiblywrittenbythe(male)ChairoftheEnglishDepartmentatCornellUniversitywhostronglyrecom-mendedthecandidatebepromotedtofullprofessoratYale.Inallfourletters,thecandidatewaslaudedassupremelycompetent,withapres-tigiousacademicbackground(e.g.,aPhDfromHarvard)andnumerousaccomplishments(e.g.,over40publications,vewellreceivednovels,andrecipientoftheMacArthurGeniusAward).Thecandidatewasalsodescribedasaworld-renownedliterarycritic.AppendixAcontainstheonlysectionoftheletterthatdifferedbycondition.Inthecommunalcondition,thecandidatewasdescribedasatactfulreviewerwhopro-tectedwriters'egos,nomatterhowpoorthework.Intheagenticcondi-tion,thecandidatewasdescribedasbrutallyhonestwhentheworkfellshort.Ineachcase,theauthorstatedthattheapplicant'sreviewingstylewasconsideredcontroversialbysome,butthatinhisopinionitwasnecessary(eithertomaintainhighstandardsortoprotectfragileegos).Eachletterconcludedthatthecandidatewasageniuswhowashighlydeservingofthepromotion,andthatifYaleislookingtopro-motethebestliterarymindsoftheera,thenyouneedlooknofurther.Candidateratings.Participantsrespondedtoalldependentmeasuresonscalesrangingfrom1(notatall)to6(verymuch).TwoitemsDidtheapplicantstrikeyouascompetent?Howlikelyisitthattheapplicanthasthenecessaryskillsforthisjob?)wereaver-agedtoformthecompetenceindex,(274)=.31,.001.Threeitemswereaveragedtoformthelikingindex:Howmuchdidyouliketheapplicant?Isthispersonsomeoneyouwanttogettoknowbetter?;andWouldtheapplicantbepopularwithcol-=.80).Threeitemswereaveragedtoformthehireabil-ityindex:Wouldyouchoosetointerviewthecandidate?youpersonallypromotethecandidate?;andHowlikelyisitthatthecandidatewillbepromoted?=.89).Genderrules.FourgenderstereotypeindexeswerederivedfromStudy1.Maleagencyprescriptionswere:career-orientedleadershipindependentbusinesssensehighself-esteem,and=.77;=1.38).Femalecommunalityprescriptionswere:sensitivetotheneedsofothersinterestedinchildrenenthusiastic,and=.90;=.03).Maleweaknessproscriptionswere:,and=.76;1.01).Femaledominanceproscrip-tionswere:intimidatingcontrollingcoldtowardothers,and=.90;=.87).ProcedureParticipantsweretoldthatwewereinterestedintheirresponsestoacandidatebeingevaluatedforpromotion.Aftersigningaconsentform,participantswereescortedtoseparatecubiclesandrandomlyassignedtooneofthefourletters,whichwerewrittenonCornellsta-tionaryandaddressedtotheDeanofFacultyofArtsandSciencesatYale.Afterreadingtheletter,theexperimenterstartedaprogramtoadministerthemeasuresintheorderdescribedabove.Theprogramrandomlypresenteditemswithineachmeasure.Finally,participantscompleteddemographicitems(age,gender,andrace).Theyweresubsequentlydebriefedandaccredited.ResultsanddiscussionCandidateratingsTable3showsthecandidates'ratingsasafunctionoftargetgen-derandcritiquingstyle.Resultsof2(targetgender)×2(style:agen-tic,communal)×2(participantgender)ANOVAsoneachmeasurerevealednosignicantparticipantgendereffects(consistentwithpreviousresearch;Rudman&Phelan,2008),alls(1,170)Becausetherewerenosignicanteffectsforcompetence,alls�.11,thefourlettersweresimilarlysuccessfulincon-veyingthecandidates'intelligence.Thelikingandhireabilityindexeseachshowedmaineffectsfortar-getgenderandstyle,qualiedbyTargetGender×Styleinteractions;boths(1,170)�3.92,.05.Table3showsnosignicanttargetgen-derdifferencesforcommunaltargets.Bycontrast,theagenticwomanwasjudgedaslesslikable,(82)=4.47,.0001,=.88,andlesshir-able,(82)=3.41,.01,=.72,comparedwiththeagenticman.GenderrulesPlannedcomparisonswereconductedforthestereotypeindexes.Table3(bottomhalf)showsnotargetgenderdifferencesinthecom-munalconditionforanyofthestereotypingindexes,all StudentsatRadboudUniversityNijmegenrespondedtotwoitems,Whenconsid-eringEnglishprofessorsatYaleUniversity,whatpercentageofthemarelikelytobewomen[men]?Onaverage,percentageswere51%and49%,respectively.Thediffer-encebetweenthesewasnotreliablydifferentfromzero, ANOVAsrevealedlargemaineffectsfortargetstyleondominanceandcommunal-ity,boths(1,170)�100.00,suchthatagentictargetswereratedasmoredominant,butlesscommunal,comparedwithcommunaltargets.Theywerealsoratedasmoreagenticthancommunaltargets,(1,170)=5.30,.05.Fortheweakindex,therewasamaineffectofparticipantgender,(1,170)=4.41,.05,suchthatmenper-ceivedtargetstobeweakerthandidwomen.Becausetheseeffectswerenotoftheoret-icalinterest,wereportdetailsonlyfortheplannedcomparisons.L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 .Aspredicted,theagenticwomanwasratedashigheronfemaledominanceproscriptionsthantheagenticman,(82)=2.31,.37.Noothertargetgenderdifferencesemergedintheagenticcondition,all.TheseresultssupporttheSIH'spredic-tionthatagenticwomensufferadominancepenalty,notacommu-nalitydecit,andfurther,theyarenotratedasextremelyhighonagencyorlowonweakness,rulingoutcontrasteffects.AccountingforbacklashagainstfemaleleadersPriorbacklashresearchhasfoundthathiringdiscriminationwasfullymediatedbygreaterlikingforagenticmenthanwomen(i.e.,Rudman&Glick,1999,2001).Totestthisprediction,weusedthePRODCLINprogramtocomputecondenceintervalsbasedonanasymmetricaldistributionofthemediated(indirect)effectMacKinnon,Fritz,Williams,&Lockwood,2007Table4'sModel1showsthepredictedresults.Theeffectoftargetgenderonhireabilitywasreducedtononsignicanceafteraccountingforliking,andthe95%condenceintervalsforthemediatedeffectdidnotincludezero,resultinginfullandreliablemediation.Newtothepresentresearch,Table4'sModel2teststhedomi-nancepenaltyasamediatoroftargetlikingdifferences.Theeffectoftargetgenderonlikingremainedreliable,butwassignicantlyre-ducedaftercontrollingforthedominanceindex(i.e.,the95%condenceintervalsforthemediatedeffectdidnotincludezero).TheseresultssupportpartialbutsignicantmediationforModel2.Finally,Model3showsthatcommunalityratingsdidnothaveasimilarmedi-ationeffect;theeffectoftargetgenderwasnotreduced(infact,itslightlyincreased),andthe95%condenceintervalsforthemediatedeffectincludedzero.Insum,Study2supportstheSIH'sclaimthatbacklashdependsonstatusviolations(i.e.,aconictbetweenwomen'sascribedandachievedstatusthatisexacerbatedbyhighstatusdisplays).Havinganoutstandinglistofachievementsdidnotprotectawomanfrombacklashorthedominancepenaltywhenshedisplayedhighstatusbehaviorsreservedformen,evenwhentheyarenecessaryforeffec-tiveleadership(i.e.,frankassessmenttomaintainhighstandards).However,itdidprotectacommunalwomanfrombothhurdles,sug-gestingthatwomencanbeexceptionallyintelligentandaccom-plishedwithoutriskingbacklash,providedtheyattenuatetheirstatuswithextremediplomacy.Acontrasteffectwasnotresponsibleforourndingsbecausecommunality,agency,andweaknessratingsweresimilarforwomenandmen.Instead,thedominancepenalty'semergenceasakeyfactorinbacklashsupportstheSIH'sclaimthatagenticwomenaredemonizedtosubordinatethem,thusprotectingthegenderhierarchy.Study3furthertestedtheSIHinajobinterviewcontextmorecomparabletopastresearch,usingliveconfederates.Inaddition,weinvestigatedgendersystem-justicationbeliefsasamod-eratorofthedominancepenalty.Study3InStudy3,confederatesrespondedtoscriptedquestionswhilebeinginterviewedforamarketingmanagerjobpre-testedtobemale-dominated.Participantsconductedphoneinterviewswithagentic(i.e.,self-promotingandambitious)targetstrainedtorespondsimilarlyusingscriptedanswers.WeexpectedtoreplicateStudy2'sbacklashpattern,wherebytargetswereviewedassimilarlycompetent,butagenticwomenencounteredprejudice(mediatedbythedominancepenalty)andhiringdiscrimination(mediatedbyprejudice).TheSIHpositsthatakeymotiveforbacklashistodefendthegen-derstatusquo,whichcausespeopletoexaggeratethedominanceofagenticwomen,ratherthandenythemcommunality(becauseonlythedominanceruleisalignedwithhighstatus).Ifthatargumentiscorrect,thenpeoplewhoendorsethegenderstatusquo(i.e.,whoaregendersystem-justiers)shouldbeparticularlylikelytoadminis-terthedominancepenalty,andtoshowprejudiceandhiringdiscrim-ination.Bycontrast,gendersystem-justicationbeliefsshouldnotmoderatecommunalityratingsbecausewomen'scommunalpre-scriptionsarestatusneutral.ParticipantsVolunteers(=74;38women)participatedinexchangeforcred-ittowardtheirIntroductoryPsychologyresearchparticipation ThisapproachavoidsrelyingonSobel'sZ,whichisantestforthesig-canceoftheaxb(mediated)effect(Baron&Kenny,1986)thatassumesthesam-plingdistributionofaxbinthepopulationisnormal(Shrout&Bolger,2002).Itisalsosuperiortobootstrapping,whichcanresultininatedTypeIerrorrates(upto10%;MacKinnonetal.,2007),whereascomputingasymmetricalcondenceintervalsoffersthebestbalanceoflowTypeIerrorswithincreasedpower(seealsoLockwood,&Williams,2004 Table4Mediationanalysesforagenticcondition(Study2).Path/effectBSE95%CondenceintervalsModel1(targetgender(targetgenderhire).33.21.11(mediationeffect)Model2(targetgender(targetgenderdominant).19(mediationeffect)Model3(targetgender(targetgendercommunal).05.08like).55(mediationeffect).03.05.058,.120.Targetgenderwascoded0().Estimatesareunstandardized.denceintervalsforarebasedonanasymmetricaldistribution.Intervalsthatdonotincludezerosupportrejectingthenullhypothesisthat=0. StudentsatRadboudUniversityNijmegenrespondedtotwoitems,Whenconsid-eringmarketingmanagers,whatpercentageofthemarelikelytobewomen[men]?Onaverage,percentageswere37%and63%,respectively.Thedifferencebetweenthesewasreliablydifferentfromzero,(40)=8.61, Table3EvaluationsofcandidatesforfullpromotiontoEnglishProfessor(Study2).CommunaltargetsAgentictargetsMaleFemaleMaleFemaleSDdCompetence5.35.785.26.74.125.34.645.17.84.23Liking4.181.454.161.06.013.781.342.531.16.88Hireability5.031.114.91.96.114.90.894.111.22.72StereotypesAgentic5.74.665.75.79.015.91.636.03.58Communal4.62.974.541.13.073.52.923.62.70Weak2.27.602.32.74.062.13.772.09.90.05Dominant3.151.003.341.16.134.941.245.48.Meanswithdifferentsubscriptsdifferbytargetgenderwithineachcondition.05).Effectsizes(Cohen's)representtargetgendersexdifferences.Positiveeffectsizesfavormalecandidates;negativeeffectsizesfavorfemalecandidates.Conventionalsmall,medium,andlargeeffectsizesforare.20,.50,and.80,respectively(Cohen,1988).L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 requirement.Ofthese,29(39%)wereWhite,20(27%)wereAsian,8(11%)wereHispanic,5(7%)wereBlack,and12(16%)reportedan-otherethnicidentity.Jobdescriptionandconfederates.Themarketingmanagerpositionwasdescribedinacompanyadvertisementasfollows:Youareresponsi-blefortheformulationandexecutionofamarketingstrategy,togeth-erwithyourteamofeightexperiencedmarketers.Youcoordinatemarketanalysesaimedatidentifyingconsumerneeds,andintroducenewproductsandservicestostrengthenourpositioninthemarket.Thequalicationsincluded,Youhaveamastersdegreeinmarketing;youliketotaketheinitiative;youhaveexcellentanalyticandcom-municationskills;andyoucanmanageandinspireateam.ates(twomen,threewomen)weresolicitedthroughadvertisementsandpaid$10.00anhour.Becausetheinterviewswereconductedbyphone,confederatesdidnothavetomemorizetheirscripts.However,theyweretrainedtopresentthemselvessimilarly.Applicantscript.Theconfederates'scriptcontainedagenticresponsestoninequestions,duringwhichapplicantsstressedtheirpriormana-gerialsuccessandtheirabilitytoinitiateprojects,leadateam,andworkwellunderpressure.AppendixBprovidesmoredetail.Applicantratings.WeusedStudy2'slikingindex(=.81).Becausethejobwasdescribedasrequiringbothagenticandcommunalqual-ities,wedesignedthecompetenceindextoreectthesedemands(seealsoRudman&Glick,2001).Participantsrespondedtofouritemsonascalerangingfrom1(notatall)to6(verymuch)thataskedwhethertheapplicanthadstronganalyticskills,wasaself-starter,agoodlistener,andwaslikelytogetaheadintheircareer.Twoitemsaskedforestimatesofthepercentageofmarketingprob-lemstheapplicantwouldbeabletosolveindependently,andtheper-centageofsubordinateswhowouldfeelcomfortableseekinghelpfromtheapplicant(alsoscaledon6-pointscales).Thesesixitemswereaveragedtoformthecompetenceindex(=.72).Usingthesame6-pointscale,hireabilitywasassessedwiththreeitemsinwhichparticipantsindicatedtheirlikelihoodofchoosingtointerviewthecandidatefortheactualjob,thattheywouldpersonallyhiretheapplicant,andthattheapplicantwouldbehired(=.89).Genderrules.WeadoptedStudy2'sstereotypictraitindexes.Reliabil-ityestimatesrangedfrom=.70(maleagencyprescriptions)to=.82(femaledominanceproscriptions).Gender-relatedsystemjustiSupportforthegenderhierarchywasassessedusingtheGenderSystemJusticationBeliefsscaleJost&Kay,2005).Participantsindicatedtheiragreementwitheightitems(e.g.,Ingeneral,relationsbetweenmenandwomenarejustandfair,Societyissetupsothatmenandwomenusuallygetwhattheydeserve),onascalerangingfrom1tronglydisagree)to9(stronglyagree).Responseswereaveragedsothathighscoresreectendorsingthegenderhierarchy(=.75).Participantsweretoldthattheywouldbehelpingarecentrecipi-entofaMaster'sdegreeinMarketingrehearseforanupcomingphoneinterview.Theyreadthejobdescriptionandreceivedtheirquestionsonindexcards(inrandomorder).Afterbrieymeetingtheconfederate,participantswereescortedtoseparatecubicleswheretheyconductedthephoneinterview.Confederatesansweredusingscriptedresponses.Aftertheinterview,participantscompletedtheirevaluations(competence,liking,hireability,andstereotypictraits)usingacomputerprogram.Gendersystem-justicationbeliefswereassessedweeksinadvanceaspartofadepartmentalpretest.Aftercompletingdemographicitems(genderandrace),participantsweresubsequentlydebriefedandaccredited.ResultsanddiscussionApplicantratingsPreliminaryanalysesusingtheveindividualconfederatesasamod-eratorrevealednegligibleeffectsonalldependentvariables,all64)1.29,.Wethereforecollapsedacrossthisvariableandsubmittedthecompetence,liking,andhireabilityindexestoseparate2(targetgender)×2(participantgender)ANOVAs.AsinStudy2,nosignieffectsinvolvingparticipantgenderemerged,alls(1,70)1.26,Table5depictsthecandidates'evaluationsbytargetgender.Re-sultsrevealedthatagenticmenandwomenwereviewedassimilarlycompetent,alls(1,70)s�.21.Theexpectedbacklasheffectsemerged:menwerelikedmorethanwomen,(1,70)=4.05,=.43),andweremorelikelytobehiredthanwomen,(1,70)=.05(=.51).Table5alsodepictsthestereotypictraitratings,bytargetgender.Thelastrowrevealsthedominancepenaltyfor(71)=2.03,.05(.46).AsinStudy2,resultsfortheagency,communal,andweakindexesshowednotargetgenderdifferences,allAccountingforbacklashtowardfemaleleadersTable6providesaconceptualreplicationofStudy2.Models1and2revealthattheeffectsoftargetgenderonhireabilityandlikingwerereducedtononsignicanceafteraccountingforliking(Model1)ordominance(Model2),andthe95%condenceintervalsforthemedi-atedeffectsdidnotincludezero,resultinginfullandreliablemedia-tionineachcase.Further,Model3showsthatcommunalityratingsdidnotmediatetheeffectoftargetgenderonliking;althoughthisef-fectwasslightlyreduced,the95%condenceintervalsforthemediat-edeffectincludedzero.GendersystemjustiTheSIHclaimsthatdefendingthegenderhierarchyisakeymo-tiveforbacklash.Ifso,thengendersystem-justiersshouldbeespe-ciallylikelytoadministerthedominancepenalty(butnotacommunalitydecit)whenratingagenticwomen.Totestthesepre-dictions,westandardizedallvariablesandthenhierarchicallyregressedfemaledominanceproscriptionsontargetgender,gendersystem-justicationbeliefs(GSJB),andtheirinteraction.Wecon-trolledforparticipantgendertobecautious,althoughresultswerevirtuallyidenticalwhetherornotwedidso,andnosexdifferencesemergedonGSJBscores,(58)=1.58,s=5.72and5.40formenandwomen).Beyondtheknowneffectfortargetgender,theexpectedTargetGender×GSJBinteractionwassigni.01.Simpleeffectsshowedthatforfemaletargets,GSJBwasposi-tivelycorrelatedwithperceiveddominance,(29)=.46,=.01.For Table5Evaluationsofagenticmaleandfemalemarketingmanagerapplicants(Study3).MaletargetsFemaletargetsMSDMSDCompetence4.88.454.75.61.23Liking4.89.654.53.84.43Hireability5.43.605.04.82.51StereotypesAgentic5.25.545.27.65Communal4.16.604.15.71.02Weak1.78.811.78.64.00Dominant3.03.763.44.Meanswithdifferentsubscriptsdifferbytargetgenderwithineachcondition.05).Effectsizes(Cohen's)representtargetgenderdifferences.Positiveeffectsizesfavormaleapplicants;negativeeffectsizesfavorfemaleapplicants.Conventionalsmall,medium,andlargeeffectsizesforare.20,.50,and.80,respectively(Cohen,1988L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 maletargets,thisrelationshipwasunreliablynegative,(27)=.Further,peoplewhoendorsedthegenderstatusquo(i.e.,whoscoredabovethemedianontheGSJB)ratedagenticwomenassignif-icantlymoredominantthanagenticmen,(28)=2.12,=.79).Forlowgendersystem-justiers(i.e.,whoscoredatorbelowthemedianontheGSJB),therewasnosignicanttargetgen-derdifference,(28)=1.62,=.12(=.52).Communalityratingswerethensubmittedtotheidenticalanaly-sis.Asexpected,resultswerenegligible,alls�.34.Thesendingsbolsterourclaimthatwomen'sprescriptionsdonotrein-forcethegenderhierarchy;iftheydid,gendersystem-justishouldhavepenalizedagenticwomenwithacommunalitydeResultsforthehiringindexshowedtheknowntargetgendereffectandareliableTargetGender×GSJBinteraction,.30,.05.ThecorrelationbetweenGSJBandhireabilitywasnegativeforagenticwomen,(31)=.38,.05,butweaklypositiveforagenticmen,(29)=.14,.Forthelikingindex,theTargetGender×GSJBinter-actionwasmarginallysignicant,.23,.08.Simpleslopesrevealedanegativebutunreliablerelationshipforagenticwomen,(31)=.18,=.32,butamarginallypositiverelationshipforagenticmen,(29)=.35,=.06.Finally,peoplewhoendorsedthegendersta-tusquoratedagenticmenasmorehirableandlikablethanagenticwomen,boths(28)�2.46,.05(s=1.13and1.04,respectively).Forlowgendersystem-justiers,theetargetgenderdifferenceswerenegligible,boths(28)1.00,s=.01and.06forhiringandliking).ReplicatingStudy2,Study3showedthatagenticfemaleleaderssufferedhiringdiscriminationthatwasfullymediatedbyprejudice,andprejudicethatwasfullyexplainedbythedominancepenalty.NewtoStudy3,gender-systemjustierswereespeciallylikelytoad-ministerthedominancepenalty(butnotacommunalitydecit).Fur-ther,gender-systemjusticationbeliefsmoderatedperceptionsofhiringandliking.Ineachcase,participantswhodefendedthegenderstatusquoweremorelikelytopenalizeagenticwomen,comparedwithagenticmen.Study4Study3'sresultssupporttheSIH'sclaimthatgendersystemjusti-cationmotivesplayaroleinbacklash,buttheydonotprovideacausaltest.Priorresearchhasprovideddirectsupportforsystem-justifyingmotivesbyexposingCanadianstoinformationthattheirnationisontheriseorindecline(Kayetal.,2009).Whenundersys-temthreat,evenwomenrespondedbydefendinggenderstereotypes(e.g.,thatmenarebetteratbusinessthanwomen);theywerealsolikelytodevalueafemalebusinessstudent(Kayetal.,2009).InStudy4,wemanipulatedsystemthreatbyexposingparticipantstonewsarticlesreportingthatAmericawaseitherindecline,ontherise,ornothingwassaid.Toinducebacklash,weusedStudy2'srec-ommendationlettersforagenticmaleandfemalecandidates.Study4'sdesignisa3(systemthreat:high,low,control)×2(targetgender)×2(participantgender)betweensubjectsfactorial.AsinStudy2,weexpectedtheagenticwomantoberatedasmoredominant,butlesslikableandhirable,comparedwiththeagenticman.However,ifbacklashservesasystem-justifyingfunction,thenallthreepenaltiesforfemaleagencyshouldbestrongestinthehighsystemthreatcondition,comparedwiththelowthreatconditionandwithcontrols.ParticipantsVolunteers(=142,90women)participatedinexchangeforpar-tialfulllmentoftheirIntroductoryPsychologyresearchrequire-ment.Ofthese,44%wereWhite,25%wereAsian,11%wereBlack,6%wereHispanic,and14%reportedanotherethnicidentity.MaterialsandprocedureSystemthreatmanipulation.Undertheguiseofpilotingmaterialsforanupcomingproject,someparticipantsreadanewsarticledescribingAmerica'sstatus.Inthehighthreatcondition,theyreadanarticleti-AmericainDeclinethatcontainedthefollowing:Thesedays,manypeopleintheUnitedStatesfeeldisappointedwiththenation'scondition.Whetheritstemsfromtheeconomicmelt-downandpersistenthighratesofunemployment,fatiguefromingprotractedwarsintheMiddleEastthathavecostAmericadearlyinbloodandtreasure,orgeneralanxietiesregardingglobalandtech-nologicalchangesthatthegovernmentseemsunabletoleveragetotheiradvantage,Americansaredeeplydissatised.Manycitizensfeelthatthecountryhasreachedalowpointintermsofsocial,eco-nomic,andpoliticalfactors.ItseemsthatmanycountriesintheworldareenjoyingbettereconomicandpoliticalconditionsthantheU.S.Inrecentnationwidepolls,moreAmericansthaneverbeforeexpressedawillingnesstoleavetheUnitedStatesandemigratetoothernations.Inthelowthreatcondition,theyreadanarticletitled,AmericaontheRisethatcontainedthefollowing:Thesedays,despitethedifcultiesthenationisfacing,manypeopleintheUnitedStatesfeelsaferandmoresecurethaninthepast.Whetheritstemsfromthenation'srelativelyfastrecoveryfromaglobaleco-nomiccrisis,ortheforeseeableendtoprotractedwarsintheMiddleEast,orageneralfaiththatthegovernmentwillbeabletoleverageglobalandtechnologicalchangestotheiradvantage,Americansare,onthewhole,satised.Manycitizensfeelthatthecountryisrelative-lystableintermsofsocial,economic,andpoliticalfactors.Itseemsthatcomparedwithmanycountriesintheworldthesocial,economic,andpoliticalconditionsintheU.S.arerelativelygood.Inrecentna-tionwidepolls,veryfewAmericansexpressedawillingnesstoleavetheUnitedStatesandemigratetoothernations.Followingthethreatmanipulation,participantswereencouragedtospend3minelaboratingonwhattheyhadreadandwhythey Table6Mediationanalysesformarketingmanagerapplicants(Study3).Path/effectBSE95%CondenceintervalsModel1(targetgender(targetgenderhire).58.15.10(mediationeffect)Model2(targetgender(targetgenderdominant).20.18.11(mediationeffect)Model3(targetgender(targetgender.01.10like).80(mediationeffect).01.08.166,.149.Targetgenderwascoded0().Estimatesareunstandardized.denceintervalsforarebasedonanasymmetricaldistribution.Intervalsthatdonotincludezerosupportrejectingthenullhypothesisthat=0.L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 thoughtthearticle'sopinionwasjustied.Controlparticipants(whodidnotreadanewsarticle)spent3minwritingabouttheirexperiencesthedaybefore.Tobolsterthecoverstory,participantswhoreadnewsarticlesreportedtheextenttowhichtheyfoundthenewsarticleclearlywritten,understandable,interesting,andcompelling.Becausenodiffer-enceswerefoundbetweenhighandlowthreatconditions,alls(89)1.33,s�.19,theseresultsarenotfurtherdiscussed.Candidateratings.WeusedStudy2'scompetence,liking,andhireabil-ityindexes(alls�.81).Becausetheresultsforcompetencedidnotdifferasafunctionoftargetgenderorthreatcondition,alls�.15,wereportonlythendingsforlikingandhireability.Wealsoemployedthefoursetsofgenderrules(alls�.75).Becausetheresultsforagencyandweaknessdidnotdifferasafunctionoftar-getgenderorthreatcondition,alls�.17,wereportonlyndingsfordominanceandcommunality.Participantsbelievedtheywouldbecompletingtwostudies,oneconcerningpilotingnewsarticlesforuseinanupcomingproject,andtheotherconcerningtheirresponsestoacandidatebeingevalu-atedforpromotion(toheightenthecoverstory,participantscon-sentedseparatelytothetwostudies).Participantswererandomlyassignedtothreatconditionbythecomputerprogram.Afterreadingthenewsarticle(ornot),theycompletedthewritingtask.Theywerethenassignedbythecomputerprogramtoreadoneofthetworec-ommendationletters,ostensiblywrittenbyaCornellEnglishprofes-sor,foreitheramaleorfemalecandidateupforpromotion(asinStudy2,exceptthatalltargetswereagentic).Afterparticipantsreadtheletter,theprogramadministeredthemeasuresexactlyasinStudy2(itemswithineachmeasurewererandomlypresented).Fi-nally,participantscompleteddemographicitems(age,gender,andrace).Theyweresubsequentlydebriefedandaccredited.ResultsanddiscussionDependentmeasuresweresubmittedtoseparate3(systemthreat:high,low,control)×2(targetgender)×2(participantgender)CandidateevaluationsReplicatingbacklasheffects,resultsforlikingandhireabilityrevealedmaineffectsfortargetgender,boths(1,130)�33.06,.001,suchthatthemaletargetwaslikedmorethanthefemaletar-get(s=4.16vs.3.11,=.87),andviewedasmorehirablethanthefemaletarget(s=5.55vs.4.30,=.95).Ofmoreimportance,resultsforbothmeasuresshowedtheexpectedSystemThreat×TargetGenderinteraction,boths(2,130)�3.23,.05.Fig.1showstheresults.Aspredicted,thefemaletargetwaslikedlessinthehighthreatconditioncomparedwiththelowthreatcondition,(42)=2.90,=.70,andcomparedwithcontrols,(44)=2.53,=.66.Similarly,thefemaletargetwasviewedassignicantlylesshirableinthehighthreatconditioncomparedwiththelowthreat(42)=4.39,=.95,andmarginally,comparedwithcontrols,(44)=1.80,=.08,=.41.Nosignicantdifferencesforlikingemergedbetweenthelowthreatandcontrolconditions,.However,lowthreatparticipantsweremorewillingtohirethefemaletarget,comparedwithcontrols,(44)=2.16,.05,=.53.Forthemaletarget,likingandhireabilitydidnotdifferasafunctionofthreat,all1.04,�s.38.GenderrulesResultsforthedominanceindexrevealedatargetgendermainef-(1,130)=9.30,.01,suchthattheagenticwomanwasviewedasmoredominantthantheagenticman(s=5.50vs.4.36,=.68).Inaddition,theexpectedSystemThreat×TargetGenderin-teractionwassigni(2,130)=3.20,Fig.2showsthere-sults.Aspredicted,thefemaletargetwasjudgedtobemoredominantinthehighthreatconditioncomparedwiththelowthreatcondition,(42)=3.03,.01,=.79,andmarginally,comparedwithcontrols,(44)=1.68,=.50.Nosignicantdifferencesemergedbe-tweenthelowthreatandcontrolconditions,(44)1.00,.Themaletargetwasviewedassimilarlydominantinallthreatconditions,all1.00,Bycontrast,wefoundnosignicanttargetgendereffectforthecommunalityindex,(1,130).Onaverage,maleandfemaletargetswereviewedassimilarlylowoncommunality(s=2.98vs.2.97).ThisreplicatesthenulleffectsfoundinStudies2and3.Unex-pectedly,amarginallysignicantSystemThreat×TargetGenderin-teractionemerged,(2,130)=2.93,.06.Innoconditiondidwendsignicanttargetgenderdifferences,all1.41,ps�.15.Instead,Fig.2showsthatthemaletargetwasviewedasmarginallymorecommunalinthehigh,comparedwiththelow,threatcondition,(45)=1.88,=.54.Nootherreliabledifferencesemergedforthemaletarget,boths�.10.Finally,thefemaletarget'scommunalitydidnotdifferacrossthreatconditions,alls�.21.Insummary,Study4bolsterstheSIH'sclaimthatbacklashservesasystem-justifyingpurpose.Asexpected,thefemaletargetwasjudgedasparticularlylowonlikingandhireability,andespeciallyhighondominance,whenparticipantswerethreatenedwithAmerica'sde-cline.Bycontrast,themaletargetwasviewedsimilarlyregardlessofthethreatmanipulation(withtheexceptionofamarginally 1 MaleTargetFemaleTargetMaleTargetFemaleTarget Liking Hireability System Threat Fig.1.Likingandhireabilityasafunctionoftargetgenderandsystemthreat(Study4). 1237 MaleTargetFemaleTargetMaleTargetFemaleTarget High Dominance Communality Fig.2.Dominanceandcommunalityasafunctionoftargetgenderandsystemthreat(Study4).L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 cantdifferenceinhiscommunalityratings).Asinpriorstudies,thefemaletargetwasviewedashigherondominancethanthemaletarget,butnotloweroncommunality,suggestingthatbacklashstemsfromperceivedstatusviolations.Giventhatpeopleundersystemthreattendtodefendtheirworldviews,whichincludegenderstatusdifferences(Kayetal.,2009),andbecausefemaleagencywasespe-ciallyrejectedbypeopleundersystemthreat,Study4providesdirectevidencethatbacklashfunctionstopreservethegenderhierarchy.Study5Thusfar,wehavetestedtheSIHusingagentictargetscompetingforleadershiprolesasituationthatevokesthedominancepenaltyforwomen.InStudy5,wesoughttoremovethedominancepenaltybymakingconfederatesaccidentalleaders.Wealsomanipulatedagencytoinstantiatefemaletargets'statusincongruity.Wereliedonhighorlowleadershipabilitytooperationalizeagencybecauseitisalignedwithhighstatus(=1.45,seeTable1).TheSIHarguesthatagenticwomenelicitbacklashbecausetheyareviewedasstatus-incongruent(theirgender'slowstatusconictswiththeirhighstatusbehavior).Agenticfemaleleadersalsodirectlychallengethegenderhierarchy:womenwhohavethetraitsrequiredforsuc-cessfulleadershiparepreciselythosewhocanbringaboutsocialchange.Bycontrast,lowagencyfemaleleadersarelesslikelytosuc-ceed,representinglessofathreattothestatusquo(indeed,iftheiranticipatedfailureoccurreditwouldreinforceratherthanchallengegenderstereotypesthatwomenmakepoorleaders,ultimatelysup-portingthestatusquo).Asaresult,evenintheabsenceofthedomi-nancepenalty,perceiversshouldexhibitbacklashagainstagenticfemaleleaderstodefendthegenderhierarchy.Afteracomputerprogramrandomlyassignedmaleandfemaleconfederatestoaleadershiprole(andallparticipantstoasubordi-naterole),participantswereaffordedtheopportunitytosabotageconfederates(i.e.,underminetheirabilitytosucceed).Becausesabo-tagereectsactiveharm,itisacompellingformofbacklash(Phelan&Rudman,2010;Rudman&Fairchild,2004).TheSIHpredictsthatonlyhighagencyfemaleleaderswillbesabotagedbecausetheyeffectivelythreatenthegenderhierarchy.Bycontrast,lowagencyfemaleleadersshouldbejudgedaslessqualiedthancomparablemaleleaders.Asarule,womenneedtobeexceptionallyagenticbeforebeingviewedasmen'sequals(e.g.,Biernat&Fuegen,2001;Eagly&Karau,2002;Foschi,2000;Glicketal.,1988).Theproblemforwomenisthatmen'sequalsyieldsbacklash.ParticipantsVolunteers(=237,123women)participatedinexchangeforpartialfulllmentoftheirIntroductoryPsychologyresearchrequire-ment.Ofthese,52%wereWhite,26%wereAsian,7%wereBlack,5%wereHispanic,and10%reportedanotherethnicidentity.AgencymanipulationTomanipulateconfederates'agency,participantscompletedacomputerizedleadershipaptitudetestinthebeliefthatconfeder-atesdidsoalso.Thetestconsistedof14itemswithhighfacevalidity.Sampleitemsincluded,WhenIaminchargeofagroup,thingsal-waysgosmoothly,Iusuallydonotneeddeadlinesandtimeta-blestobeproductive.Sevenindeterminateitemswereincludedsothatparticipantswouldndtheirownleadershipaptitudescorecred-ible(e.g.,Giventhechoice,Iwouldratherdosomethingphysicalthanreadabook).Allparticipantswereinformedbytheprogramthattheyscoredinthe77thpercentile,toheightenrespectforagenticconfederateleaders,whoostensiblyscoredatthe97thpercentile.Inthelowagencycondition,confederatesscoredatthe67thpercentile.Participantsalsocompletedabriefparticipantprothattheexperimenterexchangedsothateveryonewillknowsomethingabouttheirpartnerbeforetheystartedworkingtogether.Theproincludedgender,collegemajor,andleadershipaptitudescore.Themainpurposeoftheprolewastodelivertheagencymanipulation,butitalsounderscoredthatleaderswouldhavecontroloversubordi-nates'outcomes.Forthatpurpose,twoitemsasked,Ifyouareassignedtobetheboss,howcomfortablewouldyoufeelgradingothers'work?Ifyouareassignedtobetheboss,howcomfort-ablewouldyoufeelcriticizingothers?onscalesrangingfrom1(atall)to6(verymuchso).Allconfederateshadcircled6and5forthetwoitems,respectively.Atthetopoftheprole,theconfederate'stestscore(highorlow)appearedprominently.DependentmeasuresCompetence.Participantsindicatedwhethertheythoughttheotherparticipantwascompetent,intelligent,andwouldmakeagoodlead-er,onscalesrangingfrom1(stronglydisagree)to7(stronglyagreeTheseitemswereaveragedtoformthecompetenceindex(=.80).Likinganddominance.Participantsrespondedtothreeitemsonthesamescalesaveragedtoformthelikingindex(=.84).TheitemsIdonotlikemypartnerverymuch(reversecoded),wouldliketogettoknowmypartnerbetter,Ilookforwardtoworkingcloselywithmypartner.Theyalsoindicatedwhethertheythoughttheirpartnerwasdominant,controlling,andarrogant,aver-agedtoformthedominanceindex(=.80).Sabotagetask.Sabotagewasoperationalizedasinhibitingtargets'abili-tytosuccessfullyperformatask(Rudman&Fairchild,2004;Tesser&Smith,1980).Participantswereaskedtoprogramananagramtaskfortheleaderwhiletheexperimenterostensiblygavetheleaderfurtherin-structions.Participantsreceivedananagramanditssolutionandweretoldthattheconfederatehadtotypeinthecorrectanswerwithin30sinordertoscorepoints.Participantswereinstructedtochooseonlyonecluefromalistofthreepossiblecluestopresenttotheconfederateforeachof10anagrams.Theclueswerepretestedtovaryintheirhelp-fulnessfromlowtohigh.AsampleanagramwasCPESNRAA(thean-swerisPANCREAS).Thefollowingclueswereprovided:Itstartswiththeletter(unhelpful),It'sanorganinyourbody(medium),andIt'stheorganinyourbodythatstartswith(helpful).Theclues(whichwereunlabeledandpresentedinrandomorder)weresubse-quentlyscoredonascalefrom1(helpful)to3(unhelpful)andsummedtoformthesabotageindex(possiblerange1030,=.94).ProcedureUptofourparticipantswererecruitedforaWorkRolesandtoldasagroupthattheywouldberandomlypartneredandassignedtoeitheraleadershiporsubordinateroleforanupcomingtask.Atthispoint,theconfederate(maleorfemale)expressedapreferencefortheleadershiprolewithonestatement(IhopeIgettobetheleader),inresponsetowhichtheexperimenteremphasizedthatthecomputerwouldrandomlydeterminetheirrole.Iftheywere Toensurethatsaboteursknewtheywereharmingconfederates,weincludedthreeitemsonscalesrangingfrom1(stronglydisagree)to7(stronglyagree).SampleitemsIdonotexpecttheotherparticipantwillperformwellontheanagramtask,Itriedtobehelpfulaspossiblewhenprogrammingthetask(reversecoded).Weaveragedtheawarenessofsabotageindex(=.69)andfoundthatitcorrelatedrobustlywiththesabotagetask,(235)=.70,.001.Thus,participantswereawarethattheanagramtaskcouldbeusedtosabotagetargets(seealsoRudman&Fairchild,).Becausethendingsforthisindexmirroredthoseforthesabotagetask,wedonotdiscussitfurther.Weusedtwomaleandtwofemaleconfederates.Preliminaryanalysessuggestedwecouldcollapseacrossindividualtargetsforourfocalanalyses.L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 assignedtotheleadershiprole,theexperimenterwouldsetthemupwiththematerialsforthestudy.Meanwhile,subordinates(infact,allparticipants)wouldprogramataskfortheleaderdesignedtobechal-lenging(thesabotagetask).Participantsandconfederateswerethenescortedtoseparatecu-bicles,whereparticipantslledouttheproleandhandedittotheexperimenter,whothenstartedacomputerprogramtoadministerthemeasures.Participantsrstcompletedtheleadershipaptitudetestandreceivedtheirscore.Atthispoint,theexperimenterenteredwiththeconfederate'sprole,statingthattheconfederatehadbeenassignedastheparticipant'spartnerandpointingoutthatsheorhehadscoredatthe97thpercentileontheleadershipaptitudetest(highagencycondition),oratthe67thpercentile(lowagencycondition).Participantswerethenrandomlyassignedtothesubor-dinaterolebythecomputerprogram(withtheexperimenternotingthisinformation).Theexperimenterthenleftparticipantsalonetocompletethesabotagetask.Afterwards,theycompletedthesabotageawarenessindex,andcompetence,liking,anddominanceratings.Uponcompletion,theywerefullydebriefedandcompensated.ResultsanddiscussionDependentmeasuresweresubmittedtoseparate2(targetgender)×2(targetagency:high,low)×2(participantgender)ANOVAs.Thecompetenceindexshowedtheexpectedmaineffectfortarget(229)=27.13,.001,suchthathighagencytargetswereratedasmorecompetentandqualiedleaders(=4.94,=.93)thanlowagencytargets(=4.30,=1.06),=.62.Therewasalsoamaineffectfortargetgender,suchthatmenwereratedasmorecompetentthanwomen,(229)=5.31,=.30.How-ever,thiseffectwasqualiedbytheexpectedTargetGender×Agencyinteraction,(229)=6.05,Table7revealsthathighagencymenandwomenwereviewedsimilarly(=.11),butlowagencymenreceivedhigherscoresthancomparablewomen(=.46).Thisdemonstratestheneedforwomentoovercomethelackofbe-tweentheirgenderandleadershipbybeingexceptionallyqualiwomen'srsthurdle(Eagly&Karau,2002Resultsforthelikingindexrevealedamaineffectfortargetgen-(229)=5.26,.05.Onaverage,maleleaderswerelikedmorethanfemaleleaders(s=4.03vs.3.60,s=1.42and1.44),=.30.Noothereffectsemerged,alls�.18.Wepredictedthatwomenwouldbesparedthedominancepenal-tybecausetheywerenotactivelycompetingforleadership.Asexpected,theonlysignicanteffectfordominancewasthatmenwereratedhigherthanwomen,(229)=5.26,.05(s=4.32vs.s=1.55and1.60),=.31.Thus,participantslikelyreliedontheirgenderstereotypeswhenratingconfederates'dominanceSnodgrass&Rosenthal,1984;seeTable2Resultsrevealedamaineffectfortargetgender,(229)=5.00,.05,suchthatwomenweresabotagedmorethanmen(=.30).However,thiseffectwasqualiedbytheexpectedTargetGender×Agencyinteraction,(229)=4.59,.05.Table7revealsthathighagencywomenweresabotagedmorethancomparablemen,(115)=2.74,.01(=.51).Bycontrast,nodifferencesemergedamonglowagencytargets,(118)1.00,.Furtheranalysesshowedthathighagencyfemaleleadersweresabotagedmorethanlowagencyleaders,whethermale(=.51)orfemale(=.41),both�s2.08,.05.Finally,althoughwomenavoidedthedominancepenalty,per-ceiveddominancecorrespondedwithsabotage,butonlyforhighagencyfemaleleaders,(43)=.32,.05.Bycomparison,thisrela-tionshipwasweakforallothertargets,withsrangingfrom(forhighagencymen)to.12(forlowagencywomen),alls�.42.Insummary,Study5'sresultssupporttheSIHbecausewomenwhochallengedthestatusquobybeinghighlyqualiedleaderssuf-feredthemostsabotage,eventhoughtheywereviewedassimilarlycompetent,comparedwithhighlyqualiedmen.Further,althoughwesuccessfullyremovedthedominancepenalty,perceiveddomi-nancecorrelatedwithsabotageonlyforagenticwomen.Ofimpor-tance,lowagencyfemaleleaderswereratedaslessqualiedthancomparablemaleleaders,buttheywerenotatriskforsabotage,likelybecausetheyposednothreattothegenderhierarchy.Thus,womenoccupyingaleadershiproleisnotsufcienttoevokebacklash,butin-steadmustbecoupledwithagency(thehighstatusqualitiesexpectedofaleader)becausewomenwhoarelikelytosucceedasleadersarepreciselythosewhochallengethestatusquo.GeneraldiscussionThepresentresearchunderscoreswomen'sCatch-22:becausetheirgenderstatusislowerthanmen's,theymustenactagencytobeviewedasforleadershipyetiftheydoso,theyriskbacklash.Pastdemonstrationsofbacklashhaveleftopenthequestionofwhatmotivatesit(Rudman&Fairchild,2004;Rudman&Phelan,2008andwhichsetofgenderrulesareusedtojustifyit.Thestatusincon-gruityhypothesis(SIH)providesasignicantadvancebyspecifyingthatdefendingthegenderhierarchyisaprimarymotiveforbacklashand,asaresult,prejudiceagainstfemaleleadersstemsfromper-ceivedstatusviolations.TotesttheSIH,itwasnecessarytoestablishthestatusimplica-tionsofgenderrules.AfterStudy1determinedthatfemaledomi-nanceproscriptionsaremorealignedwithstatusthanfemalecommunalityprescriptions,theresultsoffourexperimentsshowedstrongsupportfortheSIH.Althoughagenticfemaleleaderswereviewedasmen'sequalsvis-à-viscompetence,theysufferedhiringdiscriminationthatwasmediatedbyprejudice,andprejudicewasmediatedbythedominancepenalty(extremeratingsonhighstatustraitsreservedforleadersandmen)inStudies24.Inaddition,Study3showedthatgendersystem-justierswereparticularlylikelytoengageinbacklash,includingthedominancepenalty,andStudy4showedthatprejudice,hiringdiscrimination,andthedominancepenaltywereexacerbatedwhenperceiverswereprimedwithasys-temthreat.Thesendingssupportourclaimthatakeyfunctionofbacklashistodefendthegenderstatusquo.InStudy5,femaleleaderswhowerehigh(butnotlow)onagencyweresabotagedmoresothanallothertargets,andalthoughtheyavoidedthedominancepenalty,perceiveddominancewassignicantlyrelatedtosabotageonlyinthiscondition.Whenagenticfemaleleadersarerejectedorsabotagedforbeingtoopowerful,thelogicalinferenceisthatbacklashservestodefendthegenderhierarchy.BeyondStudy4'sdirectevidenceforthismotive,furtherevidenceconcernsthefactthatwhenwomen Table7Reactionstoconfederateleadersbygenderandagency(Study5).HighagencyLowagencyMaleleadersFemaleleadersMaleleadersFemaleleadersMSDMSDMSDMSDCompetence4.98.964.87.864.501.094.02Sabotage13.004.3615.475.2912.994.4113.485.42.Meanswithdifferentsubscriptsdifferbetweenmaleandfemaleleaders(Competencewasscoredonascalerangingfrom1to7.Sabotagescorescouldrangefrom10to30.L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 leveledtheirstatusbymeansofextremediplomacy(Study2),orlowagency(Study5),theyavoidedbacklash,eventhoughtheyheldlead-ershippositions.Thus,theSIHarguesthatfemaleleadershiponitsownisnotsufcienttoprovokebacklash;whatcountsisfemaleagency,whichcanleadtotheperceptionthatfemaleleadersareratherthanentitledtowieldpower.TheSIHsetsahighbarbypredictingthatagenticwomenwillbeviewedasmoredominantthanmen,despiterobustgenderstereo-typesclaimingthereverse.Thepresentresearchsuggeststhatcon-trasteffectsarenotculpableinbacklash,giventhatnoothergenderrulesdistinguishedbetweentargetmenandwomen.Perceptualcon-trasteffectswouldresultinratingagenticwomenasexceptionallyhighonagency,andlowonweaknessandcommunality,whichwedidnotnd.Evengendersystem-justiersinStudy3eschewedrat-ingagenticwomenasloweronthanagenticmen,asdidparticipantsprimedwithasystemthreatinStudy4,whichsupportsStudy1'sndingthatcommunalityprescriptionsdonotstronglyre-inforcethegenderhierarchy.Intandemwiththedominancepenalty,thesenulleffectscontradictthenotionthatperceiverswilluseanymeansavailabletothemtojustifybacklashandinstead,supportdefendingthestatusquoasakeymotiveforrejectingagenticwomen.Asaresult,femaleleadersmayhavetowalkanelinebe-tweenpresentingthemselvesasqualiedforleadershipwhileside-steppingperceivedviolationsoffemaledominanceproscriptions.Asnoted,femaleleadersareviewedfavorablywhentheyavoidmaleleaders'privileges(e.g.,takingchargeandgivingdirections;Eaglyetal.,1992;Rudman&Glick,2001).Inessence,whenwomenleadbynotleading,theyaresparedbacklash.Althoughunfair,thisdilemmamayencouragewomentoadoptparticipatorystylesofleadershipthataremoreeffectivethanbeingdirective(Eagly&Carli,2007),whichmaybecauseforoptimism.However,ourconcernisthatstatusandcompetenceratingsaresohighlyrelated(2002;Fiskeetal.,2002)thatfemaleleaderswholowertheirstatusmayjeopardizetheirperceivedcompetence.Moreover,inordertobecomeleaders,womenmustrstdefeatthestereotypethattheytolead,andagencyisnecessarytothwartthathurdleRudman&Phelan,2008;cf.thepresentStudy5).Thatis,womenmustcombatsexismbasedonperceivedincompetence,buttheirhandsaretiedwhenfemaleagencyisreactedtonegativelytoopowerful(forawoman).Genderstatusrulesvirtuallyguaranteemen'sgreateraccesstopowerandresources,resultinginasystemthatrewardsmenforleadershipabilitieswhilepunishingcomparablewomen,therebyreinforcingtheperceivedconictbetweenawoman'sgenderandpower.Becausethisconictistherootoftheprejudicewomenface,theproblemisself-perpetuating.Attenuatingone'sstatustoavoidbacklashmightspareindividualwomenthedominancepenalty,butitdoeslittletoliftwomen'sstatusasawhichistheidealremedyforbacklash,andonethatshouldbekeptinsight.Clarifyingtheunderpinningsofbacklashhelpstoilluminateresid-ualbarrierstowomen'sadvancement(Catalyst,2010;Valian,1999Untilwomen'sculturalstatusisequaltomen's,perceiversarelikelytorejectpowerfulwomen,andunfairlyresorttothedominancepen-altytojustifydoingso.Forexample,theSIHmayexplainwhywomenwhoaremerelydescribedassuccessfulmanagersareviewedasinter-personallyabrasive(e.g.,hostileandmanipulative;Heilman,Block,&Martell,1995;Heilmanetal.,2004;Parks-Stammetal.,2008).Intheabsenceofstatus-attenuatingdisplays,perceiversmayassumesuc-cessfulwomenareextremeondominance.TheSIHalsoaccountsforwhypeoplewhoshowedautomaticprejudiceagainstfemaleau-guresdidsototheextenttheyshowedimplicitgender-statusassociations,ratherthanimplicitgender-stereotypeorgender-roleassociations(Rudman&Kilianski,2000).Byspecifyingwhichaspectsofgenderrolesareculpableinbacklash,aswellasakeymotiveforbacklash,theSIHreachesbeyondpasttheorizingEagly&Karau,2002)toexplainwhyagenticwomensufferbacklashevenwhentheyarenotleaders;forexample,whentheyaremerelydentorpresentthemselvesashavingexpertise(forare-view,seeRudman,1998),andwhyatypicalmenriskbacklash(e.g.,Heilman&Wallen,2010),coveredinmoredetailbelow.Finally,theSIHhelpstoexplainwhywomen,whoarelesslikelythanmentoendorsegenderstereotypes(e.g.,Twenge,1997),tendtoengageinbacklashtothesamedegreeasmen(forareview,seeRudman&Phelan,2008).Giventhatmenandwomenareintimatelyinterdependent,bothgendershaveastakeinthestatusquo(Glick&Fiske,1996,2001)andtherefore,mayreactnegativelytostatus-incongruentwomen.Further,whenwomendisplayagencyinpursuitofpower,theydisputethepresumeddivisionbetweenthesexes,ectedinthecommonvernacular,theoppositesex.Bothgendersaresteepedinthisdivisionfromchildhoodon,whichisreectedinthetraditionthatmenshouldbeleadersinthepublicsphere,where-aswomenarevaluedfordevotingthemselvestofamily(Eagly,1987;Eagly,Wood,&Diekman,2000).Womenwhochallengethisviewsubvertpowerfulculturalnormsdictatingwhatconstitutesanappro-priateandfulllinglife,aswellasdeeplyingrainedbeliefsaboutgen-derstatushierarchies.Asaresult,weshouldperhapsnotbesurprisedthatwomenandmenengageinbacklashsimilarly,andforthesameprimarymotive.LimitationsandfuturedirectionsGendersystem-justicationbeliefsmoderatedthedominancepenaltyinStudy3,butothervariablesshouldbeinvestigated.Forex-ample,theroleofessentialismmightbeexplored.Becauseagenticwomenchallengepresumedgenderdifferences,peoplewhobelievethatwomenarebynatureinherentlydifferentfrommenmayresistpowerfulwomenmoresothansocialconstructionists.Essentialistsaremorelikelytoendorsegenderstereotypes(e.g.,Brescoll&LaFrance,2004;Prentice&Miller,2007),butwhetherthiswouldin-uencereactionstofemaleagencyisaquestionworthyoffutureThepresentresearchfocusedonprejudicetowardfemaleleadersbecauseofitssignicanceforgenderequality.However,theSIHshouldalsoapplytomen.InStudy1,maleprescriptionsandproscrip-tionsalikewerestronglyalignedwithstatus(s=1.27andrespectively).Thus,toavoidstatusviolations,menshouldberequiredtoenactboththepositivetraitsassociatedwithleadersandtoavoidanysignsofweaknessassociatedwithlowstatus.Consistentwiththisview,modestmalejobapplicantswerejudgedtobelesslikablethanmodestfemaleapplicants,andthiseffectwasmediatedbybothmaleagencyprescriptionsandmaleweaknessproscriptionsMoss-Racusinetal.,2010).Nootherstereotypicrulesdistinguishedbetweenmodestmenandwomen(includingcommunalityitwasnotthatmodestmenweretoonice),andcompetenceratingsweresimilar.Instead,becausemodestyisastatus-attenuatingbehav-ior(Jones,1964),andbothsetsofmen'srulesarealignedwithstatus,modestmenwerelikelypenalizedforfailingtoupholdthegenderhi-erarchy.Similarly,menwhoweredescribedashighlysuccessfulinafemale-dominatedprofessionwereviewedaswimpy,weak,andspineless,whichcorrespondedtorejectingthemaspotentialsupervi-sors(Heilman&Wallen,2010).Finally,maleworkerswhorequestedafamilyleavetocareforanailingchildwerejudgedashighonweak-nessandlowonagency,andtheseperceptionsledtoeconomicpen-alties(e.g.,recommendationsfordemotionandtermination;&Mescher,inpress).AlthoughfurtherresearchisneededtotesttheSIHformen,theseinitialresultsprovidepromisingsupport.Agenerallimitationofbacklashresearchisthatithaslargelyfo-cusedonreactionstoatypicalWhitewomenandmen.Nonetheless,theSIHshouldhavebroaderapplications.Likegender,racialstereo-typesinarguablyservetojustifysocialhierarchies,withWhitesbeingviewedasmorecompetentandqualiedforleadership,relativetoracialminorities(forareview,seePratto&Pitpitan,2008).AsaL.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 result,thestrongestforWhitesmaybethattheyconformtostatusexpectancies,ratherthanstereotypicexpectancies(Phelan&Rudman,2010).Bycontrast,reactionstominoritymembersmaybeinformedbydominanceproscriptions,muchastheyareforWhitewomen.AlthoughthereisevidencesuggestingthisistrueforBlackmen(Livingston&Richardson,submittedforreview;Dovidio&Gaertner,1981;Livingston&Pearce,2009),todate,weknowofonlyonestudyinwhichprescriptivestereotypesforBlackmenandwomenwerecomparedandthenusedtoinformabacklashinvestiga-tion(Richardsonetal.,submittedforpublication).Theauthorsfoundthat(a)proscriptionsagainstdominancewerestrongerforBlackmenthanforBlackwomen,and(b)ahighlycompetentbutdominantBlackmansufferedmorebacklashthananidenticallydescribedBlackwoman.Thesendingssuggestthatgenderrulesmaydiffer(orevenbereversed)forBlacks,butmaysimilarlyinformbacklash.However,thisarenaremainswideopenforfurtherresearch.Asarststep,researchersshouldundergosystematicinvestigationsofprescriptiveandproscriptivestereotypesandtheirstatusimplica-tionsforthegroupstheyexamine.Becausegenderrulesareorganizingprinciplesthatlegitimizemen'sgreateraccesstopowerandresources(Ridgeway,2001),andindividualsaremotivatedtojustifytheexistingsocialstructure(etal.,2004;Rudmanetal.,2002),behaviorsthatchallengethegenderhierarchyarecensured.TheSIHproposesthatdefendingthegenderhierarchyisaprimarymotiveforbacklashandthus,womenarepe-nalizedforperceivedstatusviolations.Totheextentthatagenticwomenarejudgedasbrazen(ratherthanbold),theyareheldhostagetoasetofrulesthatperpetuatesinequalitybyreservingpoweranditsprivilegesformen.Becauseexhibitingmasculinetraitsisrequiredforadvancementinmosthighstatuscareers,rejectingagenticwomenreinforcesmalehegemony.Untilwomenareaffordedthesamelati-tudeasmentostriveforleadership,unequalopportunityanddispa-ratetreatmentintheworkplacewillpersist.Asaresult,apressingchallengefacinggenderequalityistheperceivedconictbetweenawoman'sgender,publicpositionsofinuence,andtheagenticquali-tiesnecessarytobejudgedworthyofwieldingpowerandcontrol.AcknowledgmentsThisresearchwaspartiallysupportedbygrantBCS-0443342fromNationalScienceFoundationtotherstauthorandanNSFGraduateResearchFellowshiptothesecondauthor.WethankRoosVonkandSusanFiskefortheirinvaluableassistancewithStudy3,whichwasconductedaspartofSanneNauts'internship,supervisedbyVonk,Rudman,andFiske.WealsothankNaliniAmbady,AliceEagly,SusanFiske,andPeterGlickfortheirhelpfulcommentsonearlierver-sionsofthispaper.CorrespondenceconcerningthisarticlemaybeaddressedtoLaurieA.Rudman,DepartmentofPsychology,TillettHall,RutgersUniversity,53AvenueE,Piscataway,NJ,08854-8040.Electronicmailmaybeaddressedtorudman@rci.rutgers.edu.AppendixAExcerptfromrecommendationletterforFullEnglishProfessor(femalecandidateversion).AllconditionsAlthoughProfessorMullenhasestablishedherselfasanaccom-plishednovelist,shemaybeevenbetterknownforherreputationasaliterarycritic.ProfessorMullenhasmadeanameforherselfasthemostrespectedopiniononcontemporaryction,andhercri-tiqueshaveappearedinthemostprestigiousnewspapersandjournalsthroughouttheworld.Herreviewshavesimplybecomethenalwordonwhatpeopleought(andoughtnot)tobereading.AgenticconditionProfessorMullenhasveryhighstandardsforcontemporaryction,andisnotafraidtoshareherbrutallyhonestopinionwhenhercol-leagues'workfallsshort.Althoughshecanbemercilessinherdamningcritiques,theabilitytoprovideunvarnished,criticalfeedbackisessentialtomovingourprofessionforward.Further,itshouldbenotedthatProfessorMullenisnotdiscriminatoryinherreviewing,treatingnew-comersandestablishedwritersaliketothesameviciousknife.Herneg-ativereviewofMargaretCrosley'sdreadfullastbookasthemostlazily-writtenbookIhaveeverreadhavingallthegraceandsubtletyofatelevisiondramaoftheweekisjustoneexampleofherwillingnesstocuttothebonewithoutregardforthefamouslyfragileartisticego.More-over,IwitnessedherpublicdebatewithPhilipBaldwinatthe2007ALAconvention,inwhichshecommentedthathis(terriblyboring)recentbook'scommercialsuccesswasremarkable,giventhatitwasterrible,bloated,boring,gratuitous,andshamelesslyuncontrolledandthatitcouldnotsurvivewithoutthegoodwillofitsreaders.Althoughsomethoughtshewentforthejugular,othersvieweditasatestimonytohercommitmenttomaintaininghighstandards.CommunalconditionProfessorMullenhasveryhighstandardsforcontemporarytion,andherreviewsareparticularlyartfulwhenhercolleague'sworkfallsshort.Althoughsheisattimesviewedasoverlypoliteandfriendlybecauseofherdiplomaticcritiques,theabilitytoprovidetactful,criticalfeedbackisessentialtomovingourprofessionforward.Further,itshouldbenotedthatProfessorMullenisnotdiscriminatoryinherreviewing,treatingnewcomersandestablishedwritersaliketothesamesensitivity.HerreviewofMargaretCrosley'sdreadfullastbookaswordy,predictable,andconventionalalthough,tobefair,somereadersmayappreciatethefamiliarityandaccessibilityoftheplotisjustoneexampleofherwillingnesstocuttothechasewhileprotectingthefamouslyfragileartisticego.Moreover,Iwit-nessedherpublicdebatewithPhilipBaldwinatthe2007ALAcon-vention,inwhichshecommentedthathis(terriblyboring)recentbook'scommercialsuccesswasremarkable,giventhatitwassojampackedwithverbiage,frustratinglycomplexcharacterdescrip-tions,andintricatelandscapedetailsandthatitwastrulyatesta-menttohisreaders'willpowerthattheywereabletoappreciatetheunderlyingstoryhiddenbeneathallthosewords.Althoughsomethoughtshewenttooeasyonhim,othersvieweditasatestimonytohercommitmenttoprotectingauthors'egos.AppendixBExcerptsfromapplicantscripts.Canyougiveanexampleofaprojectyoudidinyourformerpo-sitionatDWG,andwhatyourroleinthisprojectwas?AtDWG,Idevelopedanewwaytoconductmarketanalysisthatimprovedourresponseby20%.Insteadofconductingmarketanalysisbyphone,Iinitiatedaprojecttoswitchtomoderntechnologies,suchase-mailandusingFacebook.Wereachedamuchyoungergroupofcustomersanddramaticallyimprovedourmarketingplan.Canyounameyourtwomostimportantqualitiesandapointforimprovement?OneofmymostimportantqualitiesisthatIamgoodatanalyzingcomplexsituations.AtDWG,Iwasoftenconfrontedwithdifcultsit-uationsinwhichIhadtoincorporateperspectivesofdifferentpeopleL.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 withinthecompany,anddeterminewhatwasimportantandneededtobeaddressedimmediatelyandwhatcouldwait.IalsoknowwhatIwantandIliketomakequickdecisions,notdebateendlesslyaboutalltheoptionsthatareonthetable.Itisinefcienttokeeprepeatingar-gumentspeoplealreadyheardacoupleoftimes;ifyouhaveallthenecessaryinformation,atacertainpointyoujustneedtostoptalkinganddecidewhattodo.Apointofimprovement?Letmethink.UhIthinkapointofim-provementisthatIcanbealittleimpatientnowandthen.Makingde-cisionsquicklyisimportant,butyoushouldalsowaitforothersiftheyneedalittlemoretimetoreachacertainsolution.Underwhichconditionsdoyouworkbest?Itendtoperformbestinpressuresituations.IfIhavetodomanythingsinashortperiodoftime,IjusttrytoplanthingsreallywellandIalwaysgetalotofthingsdone.IfIhaveadeadline,IcanworkveryReferencesBaron,R.M.,&Kenny,D.A.(1986).Themoderatormediatorvariabledistinctioninso-cialpsychologicalresearch:Conceptual,strategicandstatisticalconsiderations.JournalofPersonalityandSocialPsychology,1173Berger,J.,Webster,M.,Jr.,Ridgeway,C.L.,&Rosenholtz,S.J.(1986).Statuscues,expecta-tions,andbehaviors.InE.Lawler(Ed.),AdvancesingroupprocessesVol.3.(pp.122)Greenwich,CT:JAIPress.Biernat,M.,&Fuegen,K.(2001).Shiftingstandardsandtheevaluationofcompetence:Complexityingender-basedjudgmentanddecisionmaking.JournalofSocialIssues,707Brescoll,V.,&LaFrance,M.(2004).Thecorrelatesandconsequencesofnewspaperre-portsofresearchonsexdifferences.PsychologicalScience(8),515Burgess,D.,&Borgida,E.(1999).Whowomenare,whowomenshouldbe:Descriptiveandprescriptivestereotypinginsexdiscrimination.Psychology,PublicPolicy,and,665Catalyst(2010).Statisticaloverviewofwomenintheworkplace.Availableonlineathttp://catalyst.org/publication/219/statistical-overview-of-women-in-the-workplaceCohen,J.(1988).Statisticalpowerforthebehavioralsciences.Hillsdale,NJ:Erlbaum.Conway,M.,Pizzamiglio,M.T.,&Mount,L.(1996).Status,communality,andagency:Implicationsforstereotypesofgenderandothergroups.JournalofPersonalityandSocialPsychology,25Dodge,K.A.,Gilroy,F.D.,&Fenzel,L.M.(1995).Requisitemanagementcharacteristicsrevisited:Twodecadeslater.JournalofSocialBehaviorandPersonality,253Dovidio,J.F.,&Gaertner,S.L.(1981).Theeffectsofrace,status,andabilityonhelpingSocialPsychologyQuarterly,192Eagly,A.H.(1987).Sexdifferencesinsocialbehavior:Asocial-roleinterpretation.Hillsdale,NJ.Eagly,A.H.,&Carli,L.L.(2007).Throughthelabyrinth:Thetruthabouthowwomenbe-comeleaders.Boston:HarvardBusinessSchoolPress.Eagly,A.H.,&Karau,S.J.(2002).RolecongruitytheoryofprejudicetowardfemalePsychologicalReview,573Eagly,A.H.,Makhijani,M.G.,&Klonsky,B.G.(1992).Genderandtheevaluationofleaders:Ameta-analysis.PsychologicalBulletinEagly,A.H.,Wood,W.,&Diekman,A.(2000).Socialroletheoryofsexdifferencesandsimilarities:Acurrentappraisal.InT.Eckes,&H.M.Trautner(Eds.),Thedevelop-mentalsocialpsychologyofgender(pp.123174).Mahwah,NJ:Erlbaum.Eckes,T.(2002).Paternalisticandenviousgenderstereotypes:Testingpredictionsfromthestereotypecontentmodel.SexRoles4),99Fiske,S.T.(1998).Stereotyping,prejudice,anddiscrimination.InD.T.Gilbert,S.T.Fiske,&G.Lindzey(Eds.),(4thed.).HandbookofsocialpsychologyVol.2.(pp.357411)NewYork:McGraw-Hill.Fiske,S.T.,Cuddy,A.J.C.,Glick,P.,&Xu,J.(2002).Amodelof(oftenmixed)stereotypecontent:CompetenceandwarmthrespectivelyfollowfromperceivedstatusandJournalofPersonalityandSocialPsychology,878Fiske,S.T.,&Stevens,L.E.(1993).What'ssospecialaboutsex?Genderstereotypinganddiscrimination.InS.Oskamp,&M.Costanzo(Eds.),Genderissuesincontempo-rarysociety(pp.173196).ThousandOaks,CA:Sage.Foschi,M.(2000).Doublestandardsforcompetence:Theoryandresearch.AnnualReviewofSociology,2142.Glick,P.,&Fiske,S.T.(1996).TheAmbivalentSexismInventory:Differentiatinghostileandbenevolentsexism.JournalofPersonalityandSocialPsychology,491Glick,P.,&Fiske,S.T.(2001).Ambivalentsexism.InM.P.Zanna(Ed.),Advancesinex-perimentalsocialpsychology.(pp.115188)ThousandOaks,CA:AcademicPress.Glick,P.,Zion,C.,&Nelson,C.(1988).Whatmediatessexdiscriminationinhiringde-JournalofPersonalityandSocialPsychology,178Heilman,M.E.(1983).Sexbiasinworksettings:Thelackoftmodel.ResearchinOrganizationalBehavior,269Heilman,M.E.,Block,C.J.,&Martell,R.F.(1995).Sexstereotypes:Dotheyinperceptionsofmanagers?JournalofSocialBehaviorandPersonality,237Heilman,M.E.,&Okimoto,T.G.(2007).Whyarewomenpenalizedforsuccessatmaletasks?:TheimpliedcommunalitydeJournalofAppliedPsychology,81Heilman,M.E.,&Wallen,A.S.(2010).Wimpyandundeservingofrespect:Penaltiesformen'sgender-inconsistentsuccess.JournalofExperimentalSocialPsychologyHeilman,M.E.,Wallen,A.S.,Fuchs,D.,&Tamkins,M.M.(2004).Penaltiesforsuccess:Reactionstowomenwhosucceedatmalegender-typedtasks.JournalofApplied,416Henik,A.,&Tzelgov,J.(1985).Controlofhaloerror:Amultipleregressionapproach.JournalofAppliedPsychology,577Hoffman,C.,&Hurst,N.(1990).Genderstereotypes:Perceptionorrationalization?JournalofPersonalityandSocialPsychology,197Jones,E.E.(1964).Ingratiation:Asocialpsychologicalanalysis.NewYork:Appleton-Jost,J.T.,&Banaji,M.R.(1994).Theroleofstereotypinginsystem-justicationandtheproductionoffalse-consciousness.BritishJournalofSocialPsychologyJost,J.T.,Banaji,M.R.,&Nosek,B.A.(2004).Adecadeofsystemjusticationtheory:Accumulatedevidenceofconsciousandunconsciousbolsteringofthestatusquo.PoliticalPsychology,881Jost,J.T.,&Kay,A.C.(2005).Exposuretobenevolentsexismandcomplementarygen-derstereotypes:ConsequencesforspecicanddiffuseformsofsystemjustiJournalofPersonalityandSocialPsychology,498Judd,C.M.,James-Hawkins,L.,Yzerbyt,V.,&Kashima,Y.(2005).Fundamentaldi-mensionsofsocialjudgment:Understandingtherelationsbetweenjudgmentsofcompetenceandwarmth.JournalofPersonalityandSocialPsychologyKanter,R.M.(1977).Menandwomenofthecorporation.NewYork:BasicBooks.Kay,A.C.,Gaucher,D.,Peach,J.M.,Laurin,K.,Friesen,J.,Zanna,M.,etal.(2009).Inequality,discrimination,andthepowerofthestatusquo:Directevidenceforamotivationtoseethewaythingsareasthewaytheyshouldbe.JournalofPersonalityandSocialPsychology,421434.Kobrynowicz,D.,&Biernat,M.(1998).Consideringcorrectness,contrast,andcategori-zationinstereotypingphenomena.InR.S.WyerJr.(Ed.),Advancesinsocialcogni-Vol.11.(pp.109126)Mahwah,NJ:Erlbaum.Livingston,R.W.,&Pearce,N.A.(2009).Theteddybeareffect:DoeshavingababyfacetblackchiefexecutiveofPsychologicalScience,1229Livingston,R.W.,&Richardson,E.V.(submittedforreview).Thehubrispenalty:BiasedresponsestodisplaysofBlackNFLplayers.MacKinnon,D.P.,Fritz,M.S.,Williams,J.,&Lockwood,C.M.(2007).Distributionoftheproductcondencelimitsfortheindirecteffect:ProgramPRODCLIN.BehaviorRe-searchMethods,384MacKinnon,D.P.,Lockwood,C.M.,&Williams,J.(2004).Condencelimitsforthein-directeffect:Distributionoftheproductsandresamplingmethods.BehavioralResearch,99Manis,M.,Nelson,T.E.,&Shedler,J.(1988).Stereotypesandsocialjudgment:Extremity,assimilation,andcontrast.JournalofPersonalityandSocialPsychology,2836.Moss-Racusin,C.A.,Phelan,J.E.,&Rudman,L.A.(2010).Whenmenbreakthegenderrules:statusincongruityandbacklashtowardmodestmen.PsychologyofMenand,140Nieva,V.E.,&Gutek,B.A.(1981).Womenandwork:Apsychologicalperspective.York:Praeger.Parks-Stamm,E.J.,Heilman,M.E.,&Hearns,K.A.(2008).Motivatedtopenalize:Women'sstrategicrejectionofsuccessfulwomen.PersonalityandSocialPsychology,237Phelan,J.E.,&Rudman,L.A.(2010).Actors'reactionstoethnicdeviance:Theroleofbacklashinracialstereotypepreservation.JournalofPersonalityandSocialPsychol-,265Pratto,F.,&Pitpitan,E.V.(2008).Ethnocentrismandsexism:Howstereotypeslegiti-mizesixtypesofpower.SocialandPersonalityPsychologyCompass,2159Prentice,D.A.,&Carranza,E.(2002).Whatwomenandmenshouldbe,shouldn'tbe,areallowedtobe,anddon'thavetobe:Thecontentsofprescriptivegenderstereo-PsychologyofWomenQuarterly,269Prentice,D.A.,&Miller,D.T.(2007).Psychologicalessentialismofhumancategories.CurrentDirectionsinPsychologicalScience,202Richardson,E.,Phillips,K.W.,Rudman,L.A.,&Glick,P.(submittedforpublication).Doublejeopardyorgreaterlatitude:DoBlackwomenescapebacklashfordomi-nancedisplays?Ridgeway,C.L.(2001).Gender,status,andleadership.JournalofSocialIssuesRidgeway,C.,&Bourg,C.(2004).Genderasstatus:Anexpectationstatestheoryap-proach.InA.H.Eagly,A.E.Beall,&R.J.Sternberg(Eds.),Thepsychologyofgender(pp.217241).(2nded.).NewYork:GuilfordPress.Rosenberg,S.,Nelson,C.,&Vivekananthan,P.S.(1968).Amultidimensionalapproachtothestructureofpersonalityimpressions.JournalofPersonalityandSocialPsychology,283294.Rudman,L.A.(1998).Self-promotionasariskfactorforwomen:Thecostsandbeneofcounterstereotypicalimpressionmanagement.JournalofPersonalityandSocial,629Rudman,L.A.,&Fairchild,K.(2004).Reactionstocounterstereotypicbehavior:Theroleofbacklashinculturalstereotypemaintenance.JournalofPersonalityandSocialPsychology,157Rudman,L.A.,Feinberg,J.M.,&Fairchild,K.(2002).Minoritymembers'implicitatti-tudes:Ingroupbiasasafunctionofgroupstatus.SocialCognition,294Rudman,L.A.,&Glick,P.(1999).Feminizedmanagementandbacklashtowardagenticwomen:Thehiddencoststowomenofakinder,gentlerimageofmiddleman-JournalofPersonalityandSocialPsychology,1004L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165 Rudman,L.A.,&Glick,P.(2001).Prescriptivegenderstereotypesandbacklashtowardagenticwomen.JournalofSocialIssues,732Rudman,L.A.,&Glick,P.(2008).Thesocialpsychologyofgender:Howpowerandintima-cyshapegenderrelations.NewYork:Guilford.Rudman,L.A.,&Kilianski,S.E.(2000).Implicitandexplicitattitudestowardfemaleau-PersonalityandSocialPsychologyBulletin,1315Rudman,L.A.,&Mescher,K.(inpress).Penalizingmenwhorequestafamilyleave:Isexibilitystigmaafemininitystigma?JournalofSocialIssues.Rudman,L.A.,&Phelan,J.E.(2008).Backlasheffectsfordisconrminggenderstereo-typesinorganizations.InA.P.Brief,&B.M.Staw(Eds.),ResearchinorganizationalVol.4.(pp.6179)NewYork:Elsevier.Shrout,P.T.,&Bolger,N.(2002).Mediationinexperimentalandnonexperimentalstudies:Newproceduresandrecommendations.PsychologicalMethodsSnodgrass,S.E.,&Rosenthal,R.(1984).Femalesincharge:Effectsofsexofsubordinateandromanticattachmentstatusuponself-ratingsofdominance.JournalofPerson-,355Tesser,A.,&Smith,J.(1980).Someeffectsoftaskrelevanceandhelping:Youdon'tal-wayshelptheoneyoulike.JournalofExperimentalSocialPsychology,582Touhey,J.C.(1974).Effectsofadditionalwomenprofessionalsonratingsofoccupa-tionalprestigeanddesirability.JournalofPersonalityandSocialPsychologyTwenge,J.M.(1997).Attitudestowardwomen,19701995:Ameta-analysis.SexRoles,35Valian,V.(1999).Whysoslow?Theadvancementofwomen.Cambridge,MA:TheMITWilliams,J.E.,&Best,D.L.(1990).Measuringsexstereotypes:Amultinationstudy.buryPark,CA:Sage.L.A.Rudmanetal./JournalofExperimentalSocialPsychology48(2012)165