實驗三多元線性回歸模型的估計和檢驗

1、廉束財經(jīng)大學舉商學院HUASHANGCOLLEGEGUANGDONGUN1VERS1TVOFFINANCESECONOMICS實驗報告課程名稱：計量經(jīng)濟學實驗項目：實驗三多元線性回歸模型的估計和檢驗實驗類型：綜合性口設計性口驗證性專業(yè)班別：姓名：學號：實驗課室：指導教師：石立實驗日期：2014年5月12日廣東商學院華商學院教務處制、實驗項目訓練方案小組合作：是口否小組成員：無實驗目的：掌握多元線性回歸模型估計和檢驗的方法。實驗場地及儀器、設備和材料實驗室：普通配置的計算機，Eviews軟件及常用辦公軟件實驗訓練內(nèi)容（包括實驗原理和操作步驟）：【實驗步驟】（一）國內(nèi)生產(chǎn)總值的增長模型：分析廣東省

2、國內(nèi)生產(chǎn)總值的增長，根據(jù)廣東數(shù)據(jù)（數(shù)據(jù)見“表：廣東省宏觀經(jīng)濟數(shù)據(jù)-第三章.xls”文件，各變量的表示按照試驗指導課本上的來表示）選擇不變價GDP（GDPB）、不變價資本存量（ZC）和從業(yè)人員（RY）,把GDPB作為因變量，ZC和RY作為兩個解釋變量進行二元線性回歸分析。要求：按照試驗指導課本R00P02,分別作：1 .作散點圖（GDPB同ZC,GDPB同RY）（結(jié)果控制在本頁）RPDG2 .進行因果關系檢驗（GDPB同ZC,GDPB同RY）（結(jié)果控制在本頁）PairwiseGrangerCausalityTestsDate:05/12/14Time:12:10Sample:19782005La

3、gs:2NullHypothesis:ObsF-StatisticProb.ZCdoesnotGrangerCauseGDPB263.849390.0376GDPBdoesnotGrangerCauseZC19.07482.E-05PairwiseGrangerCausalityTestsDate:05/12/14Time:12:10Sample:19782005Lags:3NullHypothesis:ObsF-StatisticProb.RYdoesnotGrangerCauseGDPB252.887440.0641GDPBdoesnotGrangerCauseRY3.463090.038

4、2從因果關系檢驗看，ZC明顯影響GDPB,RY不明顯，這是可以理解的，計劃經(jīng)濟時期存在著隱性失業(yè)，使得勞動力的變化對產(chǎn)出的影響不明顯。3 .作GDPB同ZC和RY的多元線性回歸，寫出模型估計的結(jié)果，并分析模型檢驗是均否通過？（三個檢驗）（結(jié)果控制在本頁）DependentVariable:GDPBMethod：LeastSquaresDate：05/12/14Time:12:13Sample:19782005Includedobeervations:23CoefficientStd.Errort-StafiEticProbZC03771700.00835545.1426500000RY0353

5、689004275732720280.0000C-800.59971137822-703624700000Rrsquared0999152Meandependentvar1754112djustedR-squared0.999085SD.dependentvar1683.912S.E.ofregression50.94570.Akaikeinfocritenon10.80035Sumsquaredresid64886,61Schwarzcriterion10.94309Loglikelihood-148.2050Hannan-Quinnenter.10.S4399F-etailstic1473

6、632Durbin-Walsonstat0443992ProbfF-stalistic0.000000得到的估計方程GDPB=0.377170*ZC+0.353689*RY-800.59974 .將建立的二元回歸模型（GDPB同ZC和RY）同一元回歸模型（GDPB同ZC、GDPB同RY）相比較，分析優(yōu)點。（結(jié)果控制在本頁）DependentVariable:GDPBMethod:LeastSquaresDate：05/12/14Time:12:16Sample197B2005includedobservations:28CoefTidentStdErrort-StatlsticProb.ZC0

7、.4428980004S9690460000.0000c133972125.570545.2393140.0000R-squagd0.996833Meandependentvar1764112AdjustedR-squared0.996711S.D.dependentvar1683912S.Eofregression9657302Akaikeinfocriterion12,04722Sumsquaredresid242485.QSctiwarzaiteilon1214239Loglikelihood-1666611Hannan-Quinncriter.12.07632F-statisticS1

8、83,011Durbin-Watsonstat0167556Prob（F-statistic）0000000DependentVariable:GDPBMethod:LeastSquaresDate:05/12J14Time:12:26Sample.197S2005Includedobservations.海CoefTidentSidErrort-StatistiGFob.RY2.1693170.11773518.5952300000C-5S19.1274004253-13.7931900000R-squared093皿67Meandependentvsr1754112AdjustedR-sq

9、uared0.927377,D.dependentvar1633912S.E.o1regression4537907AkaiKeinfocrilerion15.14190Sumsquaredrasid5354077Schwarzcriterion1523706Loglikelihood-2099866Hannan-Quinnenter.1517099F-statistc3457S44Durtin-Watsonstat007S64iProh(F-statistic)0.0000005 .結(jié)合相關的經(jīng)濟理論，分析估計的二元回歸模型的經(jīng)濟意義。（結(jié)果控制在本頁）估計方程的乎U定系數(shù)人R2接近1;參數(shù)

10、顯著性t檢驗均大于2;方程顯著性F檢驗顯著。調(diào)整的判定系數(shù)為0.99085,比上面的一元回歸有明顯改善。（二）宏觀經(jīng)濟模型：根據(jù)廣東數(shù)據(jù)，研究廣東省居民消費行為、固定資產(chǎn)投資行為、貨物和服務凈出口行為和存貨行為，分別建立居民消費模型、固定資產(chǎn)投資模型、貨物和服務凈出口模型和存貨增加模型。要求：按照試驗指導課本R05Ri2,分別作出以下模型，并對需要改進的模型進行改進。寫出最終估計的模型結(jié)果，并結(jié)合相關的經(jīng)濟理論，分析模型的經(jīng)濟意義。（數(shù)據(jù)見“表：廣東省宏觀經(jīng)濟數(shù)據(jù)-第三章.X1S”文件，各變量的表示按照試驗指導課本上的來表小O）1 .居民消費模型（結(jié)果控制在本頁）2,0004,0006,000

11、8,00010,000LB根據(jù)經(jīng)濟理論居民消費XFJ取決于勞動報酬LB,看散點圖和因果關系檢驗。10,000_8,000_6,000JFX4,0002,0000_0PairwiseGrangerCausalityTestsDate:05/12/14Time:12:34Sample:19782005Lags:2NullHypothesis:ObsF-StatisticProb.LBdoesnotGrangerCauseXFJ267.190100.0042XFJdoesnotGrangerCauseLB5.455160.0124從散點圖看它們之間具有線性關系，從因果關系檢驗看它們之間似乎具有雙向因

12、果關系。宏觀經(jīng)濟中確實如此。進行一元線性回歸如下：DependentVariable:XFJMethod:LeastSquaresDate:06/12/14Time:12:36Sample:19732005Indudedobservations:28CaefficientStdErrort-StatisticProb.L日0986702001691659.330100.0000c-75.9966259.99073-126680602165R-squared0.992416LTeandependentvar2362277AdjustedR-squared0992125SD.dependentva

13、r2565.722S.e.orregression227.6&09AkaiKeinfocriterion13.76260Sumsquaredresid1347921.Schwancriterion1385775Loglikelihood-190.6765HannanQuinnenter13.79169FStatistic3402401Durbin-Watsonstat0.701578ProbF-statistic)0000000得到回歸方程XFJ=0.986702*LB-75.996622 .固定資產(chǎn)投資模型（結(jié)果控制在本頁）固定資產(chǎn)投資TZC顯然取決于固定資產(chǎn)折舊ZJ、營業(yè)盈余YY和財政支出

14、CZ,進行三元線性回歸如下：DependentVariable:TZGMethod:LeastSquaresDate:05/12/14Time:12:45Sample：19732005Includedobservations:28CoeflidentStdEitqt-StatisticProb.ZJ11118640.2431524.5727160.0001YY04316920.0525568.2123520.0000CZ014321004053080.35333607269C31,2762527,8251711240270.2721R-squared0.997573Meandependentv

15、ar1628997AdjustedR-squared0.997270S.D.dependentvar2003852S.E.ofregression104.7010AkaiIceinfocriterion1227166Sumsquaredresid2630951Schwarzcriterion1246197Loglikelihood-167.8032Hannan-Quinncriter1232984F-statistic3288.646Durbin-Watsonstat1298515Prob(F-statistic)0.000000分別去掉一個解釋變量進行三個二元線性回歸如下：Dependent

16、Variable:TZGMethod:LeastSquaresDate:05/12/14Time:11247Sample:19782005Indudedobservations:2SCoefndentStd.Errort-StatisticProb.ZJYYC1.1918780.43842233,656130.08699313700918。4812g910936526,520921.25904100000ooooo0.2161R-squ百mdAdjustedR-squaredS.EofregressionSumsquaredresidLoglikelihoodF-statisticProtHF

17、-statistic)0.9975610.997360102S5212644637-167.87585111.8520000000MeandependentvarS.D.dependentwarAkaikeinfocriterionSchwarzcriterionHannan-Quinnalter.Durbin-Watsonstat1629.997200395212,2054212.3431512.249051370345DependentVariable:tzgMethod:LeastSquaresDate:05712/14Time:12:48Sample:19782005includedo

18、bservations:28CoefficientStdErrorbStatisticProb.ZJCZC1.09857S1.349301-45.613940.4650212.3624280.722479136760150,11293-09102230.02620.073603714R*squaredAdjustedR-squaredS.E.ofregressionSumsquaredresidLoglikelihoodF-statisticProb(F-statistic)09907540990014200.242110024224310.000000Meandep

19、endentvarSDdependentvarAkaikeinfocriterionSchwarzcriterionHannan-Quinncriter.Durbin-Watsonstat1628997200385213,53739136806213581520.436795DependentVariable.TZGMethod:LeastSquaresDate：05/12i14Time：12:49Sample-19782005Includedobservations28CoeHicientSilError1-SlatjsiicProt.YYCZ0.4500530070453S1C470900

20、00018692780197S46944313500000G20,918g33717015056衣8805736R*squared0.993459Meanaependentvar1626997AdjustedR-sciuareti0.995096SOdependentvar2003852S.Ecfregression140.3301AkaikeinfoGfiierion1之嵬睡Sumsquaredresid452313.8Schwarzcriterign12,96957LoglikelihoodT7s.5756Hannan-Quinncrikr,12,67047F-statistic2740.

21、226DurbinAVatsonstat0751924Prostic；0.000000從上面三個回歸結(jié)果可以看出，只要固定資產(chǎn)折舊ZJ和財政支出CZ其中一個不在方程中，回歸就能得到很好的擬合?，F(xiàn)在暫且去最后一個回歸方程來使用，方程為TZG=0.430093*YY+1.869278*CZ+20.918933 .貨物和服務凈流出模型（結(jié)果控制在本頁）先考慮影響貨物和服務凈流出CK的因素為支出法的國內(nèi)生產(chǎn)總值GDP,看散點圖和因果關系檢驗。PairwiseGrangerCausalityTestsDate:05/12/14Time：13:00Sample:197S2005Lags:2NullHypo

22、thesis:ObsF-StatisticProbCKdoesnotGrangerCauseGDP2610,25630.0008GDPdoesnotGrangerCauseCK7440170.0036從散點圖和因果關系檢驗看它們具有關系，進行一元線性回歸如下:DependsntVariableGDPMethod:L&wstSquaresDate:05,112/14Time:13:02Sample:19792005Includedobservations:28CoefficientStd.Errort-StatisticProbCK9.3122230.5205S117.88813O.OQ9QC1

23、460.714400,226536497伯0.0012R-5quared0.924852M&andepend&rit,;ar5437,386AdjustedR-squareti0.921962S.Ddependentvar&3M032SE町蚓口口1761.049Akaikeinfocriterion17,35396Sumsquaredggid80633654Scliwarzcriterion17,94911LogliKAlitiaod*2479554HannanQuinncritei.1730395F-statistic319.9853Duroin-Z/atsonstat0.868719Pro

24、b(F-statstc)0.000000在所有收集到的統(tǒng)計數(shù)據(jù)中，年利率LL是一個可以考慮引入的因素，引入LL進行二元線性回歸如下：Dependentvariable:CKMethod：LeastSquaresDate:05/12/14Time:13:04Sample:19782005Includedobservations:28CoefficientStdErrort-StatisticProb.GDP0.088239000552515.9705700000LL-42.6598911,83054-3.6058800.0014C202.217395,2503821230080.0438R-sq

25、uared0.950564Meandependentvar4270379AdjustedR-squared0.946609S.D.dependentvar651.0303S.Eofregression150.4304Akaikeinfocriterion12,96584Sumsquaredresid565732.7Schwarzcriterion13,10857Loglikelihood-170.5217Hannan-Quinncrrter.13,00947F-statistic2403512Durbin-Watsonstat1,504205Prob(F-statiStic)0.000000C

26、K=0.88239*GDP-42.65989*LL+202.21734,存貨增加模型（結(jié)果控制在本頁）存貨增加TZC顯然取決于城鄉(xiāng)儲蓄CX和商品零售價格指數(shù)PSL,進行二元線性回歸如下：DependentVariable:TZCMethod:LeastSquaresD3te:C5/12/i4Time:13:12Sample-19782005Includedobservations:2ECoefficientStJ.Errort-StatisticPro&.CX0.0306330m473g646369S00000PSL17808060190659395511200000G-2D9054645E4

27、519-45600130.0001R-squared0.952473Meandependemtvar4243629AdjustedR-squared0.946fi71S.D.dependentvar3922360S.E.ofregressionSS.S6445AkaikeInfocriterion11,91305Sumsquaredresidl197422.3Schwarzcriterion1205579Loglikelihood-1637826Hannan-Quinnenter11,95669F-Btati&ic250.5102DurbinWatsonstat2164713Prob(F-statistic)0OOOODC方程為TZC=0.030633*CX+1.780806*PSL-209.054