實驗三多元線性回歸模型的估計和檢驗

1、 實 驗 報 告課程名稱： 計量經(jīng)濟(jì)學(xué) 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 設(shè)計性 驗證性R專業(yè)班別： 姓 名： 學(xué) 號： 實驗課室： 厚德樓B503 指導(dǎo)教師： 石立 實驗日期： 2016年4月29日 廣東商學(xué)院華商學(xué)院教務(wù)處 制 一、實驗項目訓(xùn)練方案小組合作：是 否R小組成員：無實驗?zāi)康模赫莆斩嘣€性回歸模型估計和檢驗的方法。實驗場地及儀器、設(shè)備和材料實驗室：普通配置的計算機(jī)，Eviews軟件及常用辦公軟件。實驗訓(xùn)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)

2、濟(jì)數(shù)據(jù)-第三章.xls”文件，各變量的表示按照試驗指導(dǎo)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(jìn)行二元線性回歸分析。要求：按照試驗指導(dǎo)課本，分別作：1作散點(diǎn)圖（GDPB同ZC，GDPB同RY）（結(jié)果控制在本頁）2進(jìn)行因果關(guān)系檢驗（GDPB同ZC，GDPB同RY）（結(jié)果控制在本頁）Pairwise Granger Causality TestsDate: 04/29/16 Time: 14:35Sample: 1978 2005Lags: 2 Null Hypothesis:ObsF-Sta

3、tisticProb.  ZC does not Granger Cause GDPB 26 3.849390.0376 GDPB does not Granger Cause ZC 19.07482.E-05Pairwise Granger Causality TestsDate: 04/29/16 Time: 14:36Sample: 1978 2005Lags: 2 Null Hypothesis:ObsF-StatisticProb.  RY does not Granger Cause GDPB

4、 26 0.096030.9088 GDPB does not Granger Cause RY 4.728560.02023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）Dependent Variable: GDPBMethod: Least SquaresDate: 04/29/16 Time: 14:40Sample: 1978 2005Included observations: 28CoefficientStd. Errort-StatisticProb. 

5、0;ZC0.3771700.00835545.142650.0000RY0.3536890.0427578.2720280.0000C-800.5997113.7822-7.0362470.0000R-squared0.999152    Mean dependent var1754.112Adjusted R-squared0.999085    S.D. dependent var1683.912S.E. of regression50.94570    Akaike i

6、nfo criterion10.80035Sum squared resid64886.61    Schwarz criterion10.94309Log likelihood-148.2050    Hannan-Quinn criter.10.84399F-statistic14736.32    Durbin-Watson stat0.443892Prob(F-statistic)0.000000得到估計方程GDPB=0.37716969694502*ZC+0.353

7、688537498*RY-800.599732335估計方程的判定系數(shù)R2接近1；參數(shù)顯著性t檢驗值均大于2；方程顯著性F檢驗顯著。調(diào)整的判定系數(shù)為0.999085，比下面的一元回歸有明顯改善。4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(diǎn)。（結(jié)果控制在本頁）一元回歸模型：Dependent Variable: GDBMethod: Least SquaresDate: 04/29/16 Time: 14:52Sample: 1978 2005Included observations: 28CoefficientStd. Err

8、ort-StatisticProb.  ZC0.4428980.00489690.460000.0000C133.972125.570545.2393140.0000R-squared0.996833    Mean dependent var1754.112Adjusted R-squared0.996711    S.D. dependent var1683.912S.E. of regression96.57302    Akaike info cr

9、iterion12.04722Sum squared resid242485.0    Schwarz criterion12.14238Log likelihood-166.6611    Hannan-Quinn criter.12.07632F-statistic8183.011    Durbin-Watson stat0.167556Prob(F-statistic)0.000000Dependent Variable: GDPBMethod: Least Squa

10、resDate: 04/29/16 Time: 14:54Sample: 1978 2005Included observations: 28CoefficientStd. Errort-StatisticProb.  RY2.1893170.11773518.595280.0000C-5519.137400.4253-13.783190.0000R-squared0.930067    Mean dependent var1754.112Adjusted R-squared0.927377   

11、0;S.D. dependent var1683.912S.E. of regression453.7907    Akaike info criterion15.14190Sum squared resid5354077.    Schwarz criterion15.23706Log likelihood-209.9866    Hannan-Quinn criter.15.17099F-statistic345.7844    D

12、urbin-Watson stat0.078643Prob(F-statistic)0.000000問題3的二元回歸模型與一元回歸模型比較，可以得出估計方程的判定系數(shù)R2、參數(shù)顯著性t檢驗、方程顯著性F檢驗和調(diào)整的判定系數(shù)有些比一元回歸有改進(jìn)，表明這些確實應(yīng)該進(jìn)行二元回歸。 5結(jié)合相關(guān)的經(jīng)濟(jì)理論，分析估計的二元回歸模型的經(jīng)濟(jì)意義。（結(jié)果控制在本頁）因為GDPB=1.669146*ZC-0.057889*RY-476.1072系數(shù)說明，不變價資本存量ZC每增加1.669146個單位，同時從業(yè)人員RY0.057889個單位，國內(nèi)生產(chǎn)總值GDPS增加1個單位，因此符合經(jīng)濟(jì)理論。（二）宏觀經(jīng)濟(jì)模型：根

13、據(jù)廣東數(shù)據(jù)，研究廣東省居民消費(fèi)行為、固定資產(chǎn)投資行為、貨物和服務(wù)凈出口行為和存貨行為，分別建立居民消費(fèi)模型、固定資產(chǎn)投資模型、貨物和服務(wù)凈出口模型和存貨增加模型。要求：按照試驗指導(dǎo)課本，分別作出以下模型，并對需要改進(jìn)的模型進(jìn)行改進(jìn)。寫出最終估計的模型結(jié)果，并結(jié)合相關(guān)的經(jīng)濟(jì)理論，分析模型的經(jīng)濟(jì)意義。（數(shù)據(jù)見“表：廣東省宏觀經(jīng)濟(jì)數(shù)據(jù)-第三章.xls”文件，各變量的表示按照試驗指導(dǎo)課本上的來表示。）1 居民消費(fèi)模型（結(jié)果控制在本頁）Pairwise Granger Causality TestsDate: 04/29/16 Time: 15:15Sample: 1978 2005Lags: 2

14、60;Null Hypothesis:ObsF-StatisticProb.  LB does not Granger Cause XFJ 26 7.190100.0042 XFJ does not Grager Cause LB 5.455160.0124Dependent Variable: XFJMethod: Least SquaresDate: 04/29/16 Time: 15:17Sample: 1978 2005Included observations: 28VariableCoefficientStd. Error

15、t-StatisticProb.  LB0.7408080.03289322.521990.0000YY0.3620750.0464527.7946920.0000C46.9151336.602821.2817350.2117R-squared0.997789    Mean dependent var2362.277Adjusted R-squared0.997612    S.D. dependent var2565.722S.E. of regression125.3710 &#

16、160;  Akaike info criterion12.60139Sum squared resid392946.9    Schwarz criterion12.74412Log likelihood-173.4194    Hannan-Quinn criter.12.64502F-statistic5641.541    Durbin-Watson stat1.122075Prob(F-statistic)0.0000002固定資產(chǎn)投資模型（結(jié)果

17、控制在本頁）Dependent Variable: TZGMethod: Least SquaresDate: 04/29/16 Time: 15:21Sample: 1978 2005Included observations: 28VariableCoefficientStd. Errort-StatisticProb.  ZJ1.1118640.2431524.527160.0001YY0.4316920.0525668.2123520.0000CZ0.1432100.4053080.3533380.7269C31.2762527.825171.1240270.272

18、1R-squared0.997573    Mean dependent var1628.997Adjusted R-squared0.997270    S.D. dependent var2003.852S.E. of regression104.7010    Akaike info criterion12.27166Sum squared resid263095.1    Schwarz criterion12.46197Log

19、 likelihood-167.8032    Hannan-Quinn criter.12.32984F-statistic3288.646    Durbin-Watson stat1.298515Prob(F-statistic)0.0000003貨物和服務(wù)凈流出模型（結(jié)果控制在本頁）Dependent Variable: CKMethod: Least SquaresDate: 04/29/16 Time: 15:25Sample: 1978 2005Included observations: 28Var

20、iableCoefficientStd. Errort-StatisticProb.  GDP0.0882390.00552515.970670.0000LL-42.6598911.83064-3.6058800.0014C202.217395.250382.1230080.0438R-squared0.950564    Mean dependent var427.0379Adjusted R-squared0.946609    S.D. dependent var651.0303S.E.

21、of regression150.4304    Akaike info criterion12.96584Sum squared resid565732.7    Schwarz criterion13.10857Log likelihood-178.5217    Hannan-Quinn criter.13.00947F-statistic240.3512    Durbin-Watson stat1.504205Prob(F-s

22、tatistic)0.0000004存貨增加模型（結(jié)果控制在本頁）Dependent Variable: TZCMethod: Least SquaresDate: 04/29/16 Time: 15:27Sample: 1978 2005Included observations: 28VariableCoefficientStd. Errort-StatisticProb.  CX0.0306330.0047396.4638880.0000PSL1.7808060.1988598.9551120.0000C-209.054645.84519-4.5600130.0001

23、R-squared0.952473    Mean dependent var424.3629Adjusted R-squared0.948671    S.D. dependent var392.2360S.E. of regression88.86446    Akaike info criterion11.91306Sum squared resid197422.3    Schwarz criterion12.05579Log likelihood-163.7828    Hannan-Quinn criter.11.95669F-statistic250.5102    Durbin-Watson stat2.164713Prob(F-statistic)0.000000二、實驗總結(jié)與評價實驗