基于经典计量模型的统计数据质量评估方法

[1]BreitungJ.andC.Gouriroux(1997).Ranktestsforunitroots,JournalofEconometrics81,2-27.[2]Breitung,J.(2001).Ranktestsfornonlinearcointegration.JournalofBusinessandEconomicStatistics,19,pp.331-40.[3]Chang,Y.,Park,J.Y.andP.C.B.Phillips(2001):Nonlineareconometricmodelswithcointegratedanddeterministicallytrendingregressors,EconometricsJournal,4,1-36.[4]Chinn,M.D.,1991.Somelinearandnonlinearthoughtsonexchangerates.JournalofInternationalMoneyandFinance10,214-230.[5]Dufrnot,G.andV.Mignon(2002).Recentdevelopmentsinnonlinearcointegrationwithapplicationstomacroeconomicsandfinance,KluwerAcademicPublishers.[6]Escribano,A.andS.Mira(2002)Nonlinearerrorcorrectionmodels,JournalofTimeSeriesAnalysis,Vol.23,No.5,509-522.[7]Franses,H.G.andM.McAleer(1995).Testingforunitrootsandnon2lineartransformations,Journaloftimeseriesanalysis,19,147-164.[8]Gonzalo,JandJ.Pitarakis(2006):ThresholdEffectsinCointegratingRelationships,OxfordBulletinofEconomicsandStatistics,68,Supplement813-834.[9]Granger,C.W.J.,1991.Somerecentgeneralizationsofcointegrationandtheanalysisoflong2runrelationships.In:Engle,R.F.,Granger,C.W.J.(Eds.),Long2RunEconomicRelationships.OxfordUniversityPress,pp.277-287.,,,,,,(:)3:,,,,,(),,,:;;;:C811:A:1002-4565(2009)03-0091-06AnEvaluationMethodofStatisticalDataQualityBasedontheClassicalEconometricModelLiuHong&HuangYanAbstract:Basedoneconomictheoryandtheentireeconomicsystem,thispaperusesrelatedfactorsoftheresearchsubjectstobuildeconometricmodel.Anddependingontheestablishedmodel,thispaperusesthetestofabnormalvaluesandtheprinciplesofstatisticaldiagnosetoevaluatethequalityofdataquantitatively.Bychoosingtheappropriatemodel,thepapersimulatesthechangesofinspectedobjects,identifiesexceptionaldata(outliers),judgesitssignificanceandinvestigatesthecauses.Finally,thepaperempiricallyanalysesthequalityofourcountrysstatisticaldata.Keywords:thequalityofstatisticaldata;classicaleconometricmodel;exceptionaldata;empiricalanalysis3,,,,26320093StatisticalResearchVol.26,No13Mar.2009,,,,,GDP,,;,,,,,,,,,,,,,:,();;,,(),,;,,,,,,,,,,;,C2D2(Cobb2Dauglas),C2DY=AKL,tr,Y=A0ertKL,Y,K,L,r,A0,Y=A0ertKL,,:lnY=lnA0+rt+lnK+lnL+,+=1,:ln(YPL)=lnA0+rt+ln(KPL)+,,,,,,,,,,,,,,():,,,,,,112070,,,,:9220093,,:,,;,,,,,,,,,;,;,,,,,,,,BechmanCook(1983),:,,,(contaminant),,:,,,,,,,^,,,,,,,,,,,,,,,,,,,,21,,:yi=0+1xi1++ixik+i,(i=1,2,n)(1)ii,Y=X+,Y=(y1,,yn),=(1,,n),=(0,1,,k),Xnp,i(1,xi1,,xik),H=X(XX)-1X,hiiHi,,,hii,,hii(1)^i,,,,^i:ri,ti:ri=^is1-hii,ti=^i[s(i)1-hii],s,s(i)i,263:93tit,t(n-k-1),|ti|,i;rirititiri:t2i=(n-k-1)r2in-k-r2i(2)cook,Cook(1977)cook,:(1),Rp,^,^,,i^^(i),^(i)i,cook:Di=(^(i)-^)XX(^(i)-^)ks2s2(1),cook,(3)w2kDIFFITS,DF,FITWekschKuh(xi,yi)xi,w2k:wki=hii1-hii1P2tiri,SPSS1115,ritiri,SPSS1115Leveragevalues,,hii,hih(h=23kPn),hih,ik,n,,:ln(YPL)=lnA0+rt+ln(KPL),1978,1978-2004,,,,,(),,,,1978GDPt(),,,Lt,SNA,1993,(),,,,,,,,9420093,Goldsmith,:Kt=It+(1-t)Kt-1,Ktt,Kt-1t-1,Itt,tt:1119521978,1978,,(2004),197810%211993,1991,()Pt1978(1978-2000),(1978)Pt,:=PtPPt-1PtPPt-1,31SNA,,1993,,,41,916%,,916%,:ln(GDPtPLt)=lnA0+rTt+ln(KtPLt)+t,Yt=ln(GDPtPLt),Ct=ln(KtPLt),T1=1,T=1,SPSS1115:Yt=0+1Ct+2Tt+t(2):^Yt=31983+01369Ct+01053Tt(15135)(9179)(141399)p(01000)(01000)(01000)(3)01997,F35991839,FP01000,95%,,,SPSS,11(3)tiricookwkihi1(1978)-214683212497018765-0103730126442(1979)0101290101310100000100010117963(1980)0105100105200100020100030110984(1981)-014762014837010092-0100230107145(1982)0114900115200100070100060105046(1983)-018233018286010186-0100270103917(1984)1134761132620104370100390103038(1985)2170722141830116110100720102489(1986)11566711523001050001003901020610(1987)11175411166501028501003001021211(1988)01958301959901023601003001034412(1989)01007101007201000001000001053613(1990)-019243019270010353-01004501073214(1991)-118345117534011642-01010501090715(1992)-111624111544010653-01006601089616(1993)-016452016528010164-01003001068617(1994)-011892011930010012-01000801052918(1995)-012092012133010010-01000601029819(1996)-011803011839010007-01000401019220(1997)-010623010636010001-01000201022121(1998)-010907010926010002-01000301035422(1999)-012770012823010027-01001201058023(2000)-012280012324010024-01001201084624(2001)01065501066901000201000401108425(2002)01286501291901005101002101121026(2003)01242201246901003701001801122527(2004)015534015612010197010041011245tit(n-k-1),(1)n=27,k=2,5%,tt0105(24)=11711,1,1(1978)8(1985)14(1991)ti-214683217072-118345,t0105(23),ti,263:95,ri118625211583212312,ri,cook,cook1(1978)8(1985)14(1991);w2k,wki1(1978)14(1991)8(1985),cookw2k,,184,184hi,h=23kPnh,h=011481,1(1978)2(1979)h,,1(1978)8(1985)14(1991),7(1984)9(1986)tiricookwki,,,,11,,,21,,,,,,,,,,31,,,,,,,,,,,,,[1].[J].,2003(4).[2],,.[J].,2004(8).[3],.[J].,2006(3).[4],.[J].,2007(8).[5],,.[J].,1997(2).[6],.[J].,2000(10).[7].[M].,2003.[8],.[J].,1999(7).[9].,[M].1991:128.[10],,,,.[J].,2004(9).[11].[J].,1999(4).[12].[M].,2005.[13].GDP[J].,2002(5).[14],,1:1952-2000[J]1,2004(10).,,:,(:)9620093