資源描述：
《graphical causal models》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在工程資料-天天文庫。
1、Chapter13GraphicalCausalModelsFelixElwertAbstractThischapterdiscussestheuseofdirectedacyclicgraphs(DAGs)forcausalinferenceintheobservationalsocialsciences.ItfocusesonDAGs?mainuses,discussescentralprinciples,andgivesappliedexamples.DAGsarevisualrepresentationsofqualitat
2、ivecausalassumptions:Theyencoderesearchers?beliefsabouthowtheworldworks.Straightforwardrulesmapthesecausalassumptionsontotheassociationsandindependenciesinobservabledata.ThetwoprimaryusesofDAGsare(1)determiningtheidentiTabilityofcausaleffectsfromobserveddataand(2)deriv
3、ingthetestableimplicationsofacausalmodel.ConceptscoveredinthischapterincludeidentiTcation,d-separation,confounding,endogenousselection,andovercontrol.Illustrativeapplicationsthendemonstratethatconditioningonvariablesatanystageinacausalprocesscaninduceaswellasremovebias
4、,thatconfoundingisafundamentallycausalratherthananassociationalconcept,thatconventionalapproachestocausalmediationanalysisareoftenbiased,andthatcausalinferenceinsocialnetworksinherentlyfacesendogenousselectionbias.Thechapterdiscussesseveralgraphicalcriteriafortheidenti
5、Tcationofcausaleffectsofsingle,time-pointtreatments(includingthefamousbackdoorcriterion),aswellidentiTcationcriteriaformultiple,time-varyingtreatments.IntroductionVisualrepresentationsofcausalmodelshavealonghistoryinthesocialsciences,Trstgainingprominencewithpathdiagra
6、msforlinearstructuralequationmodelsinthe1960s(Blalock1964;Duncan1975).Sincethesebeginnings,methodologistsinvariousdisciplineshavemaderemarkableprogressindevelopingformaltheoriesforgraphicalcausalmodelsthatnotonlygeneralizethelinearpathdiagramsofyoreintoafullynonparamet
7、ricframeworkbutalsointegrategraphicalmodelswiththereigningpotentialoutcomesframeworkofcausalinference.Bestofall,methodologistshavedevelopedasystemthatisbothrigorousandeasytouse.Inrecentyears,graphicalcausalmodelshavebecomelargelysynonymouswithdirectedacyclicgraphs(DAGs
8、).Ontheirown,DAGsarejustmathematicalobjectsbuiltfromdotsandarrows.Withafewassumptions,however,DAGscanberigorouslyrela
