Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
HermanAguinisAvram TuckerDistinguishedScholar&
ProfessorofManagementSchoolofBusiness
GeorgeWashingtonUniversitywww.hermanaguinis.com
It’sBeenanExtraordinaryFirstYear@GW
THANK YOU!!
ResearchPhilosophyandInterests• Professionalandlifeagenda• Problem-focused,interdisciplinary,andmulti-method
research:• Talentmanagement• Howtoturntalentmanagementintoasourceof
inimitableandsustainablecompetitiveadvantageand,simultaneously,improveemployeewell-being?
• Researchmethodsandanalysis• Howtoimproveresearchmethodsandanalysisto
makeimportanttheoryadvancementswithclearimplicationsforpracticeandpolicies?
ExamplesofRecentProjectsHowcanweenhanceemployeeengagement?• Aguinis,H.,&Glavas,A.inpress.Oncorporatesocialresponsibility,
sensemaking,andthesearchformeaningfulnessthroughwork.JournalofManagement.doi:10.1177/0149206317691575
Whatisthecausalchainofleadership?• Bergh,D.D.,Aguinis,etal.,2016.Usingmeta-analyticstructuralequation
modelingtoadvancestrategicmanagementresearch:Guidelinesandanempiricalillustrationviathestrategicleadership-performancerelationship.StrategicManagementJournal,37:477-497.
Howtoproduceandmanagestarperformers?• Joo,H.,Aguinis,H.,&Bradley,K.J.2017.Notallnon-normaldistributionsare
createdequal:Improvedtheoreticalandmeasurementprecision.JournalofAppliedPsychology,102:1022-1053.
Today’sPresentation• Starperformers
• Credibilitycrisisinscience
• Insightsfromtalentmanagementresearchandtheorieshelpusunderstand:• Howtoproduceandmanagestarperformers• Howtoaddressthecredibilitycrisisconceptualizedas
aperformanceproblem(i.e.,“researchperformance”problem)
Argentinaearly1970s…
StarPerformersinSports
• 1974:Soccercards—SoccerWorldCupinMunich• Grzegorz Lato scored7 goals,JohanNeeskensandAndrzej
Szarmach scored5 goals,3otherplayers4 goals,6players3goals,9players2 goals,29players1 goal,andabout170playersscoredzerogoals
• 5ofthe6topscorersinthethreebestteams(outof16)
NotJustinSports…• MarkZuckerberg:
“Someonewhoisexceptionalintheirroleisnotjustalittlebetterthansomeonewhoisprettygood…Theyare100timesbetter”
NotJustinSports…• MarcAndreessen:“Fivegreat
programmerscancompletelyoutperform1,000mediocreprogrammers”
NotJustinSports…• JackWelch:“Theteamwiththe
bestplayerswins”
DifferentSamples&Measures• Multiplestudiesincludingmorethan600,000people:
• Researchersinmorethan50academicdisciplines• Actors,actresses,directors,choreographers,andlighting
specialistsinthemovieandTVindustries• Fictionandnonfictionwriters• Musicians• ElectedofficialsintheUnitedStatesandothercountry-level
legislativebodies(e.g.,Australia,Canada,Ireland,Estonia)• Professionalandcollegiateathletesinfootball,baseball,
basketball,cricket,swimming,trackandfield,skiing,tennis,andothersports(individualandteamsports)
• Othertypesofworkersincludingbanktellers,callcenteremployees,grocerycheckers,electricalfixtureassemblers,wirers
HowMuchBetteraretheBest?• About65.8%ofresearchersfallbelowthemean
numberoftop-tierjournalarticlespublished
• BasedontheEmmy-nominatedentertainers,83.3%fallbelowthemeanintermsofnumberofnominations
• ForUSRepresentatives,67.9%fallbelowthemeanintermsofnumberoftimeselected
• ForNBAplayers,71.1%arebelowthemeanintermsofpointsscored
Let’sFormalizetheseResults…
𝜇Normal
Performance
Num
ber
of W
orke
rs
Low High
High
Low
𝜇Heavy-tailed
𝑝 𝑥 ∝ 𝑥%& ,wherealpha(α)(>1)istherateofdecay
The“Normal”ViewofPerformance
NationalPublicRadio’sMorningEdition(900stations,20millionlisteners)
JustThreeExamples(of100s)
HowGoogle’sVPforPeopleOperationsSummarizedourResults
ResearchProgram:RelevantSources• Availableatwww.hermanaguinis.com:• O’Boyle,E.,&Aguinis,H.2012.Thebestandtherest:Revisitingthe
normofnormalityofindividualperformance.PersonnelPsychology,65:79-119.
• Aguinis,H.,&O’Boyle,E.2014.Starperformersintwenty-first-centuryorganizations.PersonnelPsychology,67:313-350.
• Crawford,G.C.,Aguinis,H.,etal.2015.Powerlawdistributionsinentrepreneurship:Implicationsfortheoryandresearch.JournalofBusinessVenturing,30:696-713.
• Aguinis,H.,O’Boyle,E.,Gonzalez-Mulé,E.,&Joo,H.2016.Cumulativeadvantage:Conductorsandinsulatorsofheavy-tailedproductivitydistributionsandproductivitystars.PersonnelPsychology,69:3-66.
• Joo,H.,Aguinis,H.,&Bradley,K.J.2017.Notallnon-normaldistributionsarecreatedequal:Improvedtheoreticalandmeasurementprecision.JournalofAppliedPsychology, 102:1022-1053.
SomeImplicationsforTheory• Increasedstarpresencein21st-centuryorganizations:job
complexity,reducedsituationalconstraints,moreautonomy,flexiblehierarchies,moreaccesstoknowledgeandinformation
• Theadditionanddepartureofstarshasextraordinaryconsequencesonoverallorganizationalproductivity
• Therelationbetweenperformanceandvaluefollowsanon-linearfunctionwithapproximately30%ofvaluevestedinthetop10%ofworkersand50%ofvaluevestedinthetop25%
• Thenon-linearrelationshipbetweenindividualperformanceandvalueismoderatedbyaposition’sdegreeofproximitytoanorganization’sstrategiccorecompetence:steepercurveasproximityincreases
• Performance-basedpaydispersion(i.e.,payforperformance)leadstoincreasedstarperformerretentionandlesspaydispersionleadstoincreasedstarperformanceturnover
GivenTheseResults…Implicationsforpractice:
• Yoursuggestionsforproducingstarperformers?• Yoursuggestionsfor“cloning”starperformers?• Yoursuggestionsformanagingstarperformers?• Yoursuggestionsforretainingstarperformers?• Yoursuggestionsformanagingaverageperformers?
SomeImplicationsforPractice• Minimizesituationalconstraints(i.e.,ceilingconstraints)facedby
workerstoallowfortheemergenceofmorestarperformers• Allowstarperformerstorevolveinandoutofteamsbecausethis
widensthestar’snetworkandtakesfulladvantageofknowledgetransfertorisingstars
• Firmsexperiencingfinancialdifficultiesshouldpayspecialattentiontostarperformersasbudgetcuts,downsizing,andothercostcuttingmeasuressignalsthattheorganizationisindeclineandleadstopreemptivestardeparture
• I-dealsforstarsshouldbeclearlyarticulatedandappliedfairly—everyonecanbecomeastar
• Managersinvestingadisproportionateamountoftheirresourcesintostarsgenerategreateroveralloutputandcreatepositivegainspirals
SomeImplicationsforPractice• Nothingwrongwithbeingan
averageperformer!• AbbyWambach(184goalsin
internationalmatches):“I’veneverscoredagoalwithoutgettingapassfromsomeoneelse.”
• Managementpracticessuchasnon-performancebasedincentives,limitedpaydispersion,andlongevity-basedpromotiondecisionsthatemphasizehomogeneitydemotivatestarsandleadtotheirpsychologicalwithdrawal,emotionaldetachment,andeventually(whentherightopportunityappears)theirdeparture
StarPerformers• Otherthoughts?• Otherreactions?
Let’sSwitchGears…
CredibilityCrisisinScience
• Between30%to60%ofthepublishedempiricalstudiesarenotreplicableinseveralscientificfields…• Psychology• Economics• Management• Notjustasocialscienceproblem…
ResultsofaNature Survey
Source:Baker,M.2016.1,500scientistsliftthelidonreproducibility.Nature,533:452-454.
Psychology• 100experimentalandnon-experimentalstudiesinthree
leadingpsychologyjournals• 270authorsconductedthe100replications• Replicationswithhighfidelitytotheoriginaldesigns• Results:
• Meanr declinedfrom.40to.20• Numberofstatisticallysignificantresultsdeclinedfrom
97%to36%• Effectsizer declinedin53%oftheeffectsizes• Replicationfailurerate:60%
Source:OpenScienceCollaboration.2015.Estimatingthereproducibilityofpsychologicalscience.Science,349(6251):aac4716.
Economics• 18experimentsinAmericanEconomicReviewandQuarterly
JournalofEconomics,2011-2014• 18authorsandresearchdesignwasverifiedbyoriginal
authors• Results:
• Overall,40%notreplicable• Meaneffectsize=.47originalstudies,.28replication• Non-experimentalresearch:morethan75%not
replicable
Source:Camerer,C.F.,etal.2016.Evaluatingreplicabilityoflaboratoryexperimentsineconomics.Science,7/9/17351(6280):1433-1436.
Management• Reproducibility:Retestthehypothesesusingthesamedataas
intheoriginalstudy• 88OLSregressionorSEMmodelsasreportedinStrategic
ManagementJournal• 70%reportedincompletedatatoallowforanytypeof
reproducibilityattempt• Ofthosethatcouldbereproduced,about1in3hypothesistests
ledtostatisticallynon-significantratherthansignificantconclusions
Source:Bergh,D.D.,Sharp,B.M.,Aguinis,H.,&Li,M.2017.Isthereacredibilitycrisisinstrategicmanagementresearch?Evidenceonthereproducibilityofstudyfindings.StrategicOrganization,15:423-436.
The Perfect Storm…
• Increasedpressuretopublishin“A”journalsandincreasedrejectionratesandnumbersubmissionsat“A”journals
• Increasedmethodologicalsophistication• Decreasedresourcesfordoctoral&facultytraining& retooling• Increaseddemandstobecome“daVinciAcademics”
OnTorturingtheDataUntiltheyConfess
WhatDoFrenchCuisineandManyJournalArticlesHaveinCommon?• Thefinal“perfect”product…
TorturingtheDataUntiltheyConfess• Whathappenedinthe“researchkitchen”?• Selectionofvariablestoincludeinamodel• Useofcontrolvariables• Handlingofoutliers• Reportingofp-values• Hypothesizingafterresultsareknown(HARKing)
TalentManagementInsights:PerformanceProblem• Performanceisdeterminedmostlybytwomajorfactors(e.g.,
VanIddekinge,Aguinis,Mackey,&DeOrtentiis,inpress,JournalofManagement):• Knowledge,skills,andabilities(KSAs)
• Oneneedstoknow howtodothejobwell• Motivation
• Oneneedstowant todothejobwell• ResearchersneedtohavethenecessaryKSAstocreate
reproducibleandreplicableknowledge—knowledgethatwillbecredible,trustworthy,anduseful
• Researchersneedtowanttocreatereproducibleandreplicableknowledge
“ResearchPerformance”Problem• KSAs:
• Reducedtrainingfordoctoralstudents• Methodologycoursesoften“outsourced”• Reducedresourcesforre-toolingofmoreseniorresearchers• Someuniversityadministratorserroneouslyperceive
doctoralprogramsasa“costcenter”ratherthan“research-capabilities&research-culture-buildingcenter”
• Motivation:• Pressuretopublishin“A”journalsbyallmeansnecessary• 1994vs.2013inJournalofAppliedPsychology:
• Tellinga“goodstory”seemscriticalbasedonnumberofpages:Introduction(2.49to3.90),Discussion(1.71to2.69),Method(1.58to1.81),Results(2.41to2.34)
ShouldWeDoAnythingAboutthisinOurFields?AtGWSB?• Accounting?• Decisionsciences?• Finance?• Informationsystems&technologymanagement?• Internationalbusiness?• Management?• Marketing?• Strategicmanagementandpublicpolicy?
OurModestProposals• Availableatwww.hermanaguinis.com:• Aguinis,H.,Ramani,R.S.,&Alabduljader,N.inpress.Whatyouseeis
whatyouget?Enhancingmethodologicaltransparencyinmanagementresearch.AcademyofManagementAnnals.
• Aguinis,H.,Cascio,W.F.,&Ramani,R.S.2017.Science’sreproducibilityandreplicabilitycrisis:Internationalbusinessisnotimmune.JournalofInternationalBusinessStudies,48:653-663.
• Gonzalez-Mulé,E.,&Aguinis,H.inpress.Advancingtheorybyassessingboundaryconditionswithmeta-regression:Acriticalreviewandbest-practicerecommendations.JournalofManagement.
• McKenny,A.F.,Aguinis,H.,Short,J.C.,&Anglin,A.H.inpress.Whatdoesn’tgetmeasureddoesexist:Improvingtheaccuracyofcomputer-aidedtextanalysis.JournalofManagement.
OurModestProposals• Aguinis,H.,Edwards,J.R.,&Bradley,K.J.2017.Improvingour
understandingofmoderationandmediationinstrategicmanagementresearch.OrganizationalResearchMethods,20:665-685
• Bernerth,J.&Aguinis,H.2016.Acriticalreviewandbest-practicerecommendationsforcontrolvariableusage.PersonnelPsychology,69:229-283.
• Bosco,F.A.,Aguinis,H.,Field,J.G.,Pierce,C.A.,&Dalton,D.R.2016.HARKing’s threattoorganizationalresearch:Evidencefromprimaryandmeta-analyticsources.PersonnelPsychology,69:709-750.
• Aguinis,H.,Gottfredson,R.K.,&Joo,H.2013.Best-practicerecommendationsfordefining,identifying,andhandlingoutliers.OrganizationalResearchMethods,16:270-301.
TalentManagementResearchandTheoriesofferUsefulInsights…• Ifwewanttoproduce,manage,retain,and“clone”star
performers
• Ifwewanttounderstandandaddressthecredibilitycrisisandthechallenging“researchperformanceproblem”inmanagement,economics,psychology,andmanyotherfields
• Articlesavailableatwww.hermanaguinis.com
THANKYOU!!!