計量經(jīng)濟(jì)作業(yè)

1、第三章課后習(xí)題3.3 家庭書刊消費收家庭收入及戶主受教育年數(shù)的影響。Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:19Sample: 1 18Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C-50.0163849.46026-1.0112440.3279X10.0864500.0293632.9441860.0101X252.370315.20216710.067020.0000R-squared0.951235

2、Mean dependent var755.1222Adjusted R-squared0.944732 S.D. dependent var258.7206S.E. of regression60.82273 Akaike info criterion11.20482Sum squared resid55491.07 Schwarz criterion11.35321Log likelihood-97.84334 F-statistic146.2974Durbin-Watson stat2.605783 Prob(F-statistic)0.000000（1）設(shè)家庭書刊消費的計量經(jīng)濟(jì)模型：其

3、中Y為家庭書刊年消費支出 X為家庭月平均收入、T為戶主受教育年數(shù)（2）根據(jù)表中的數(shù)據(jù)，模型估計的結(jié)果為（3） 由估計檢驗結(jié)果, 戶主受教育年數(shù)參數(shù)對應(yīng)的t 統(tǒng)計量為10.06702, 明顯大于t的臨界值，同時戶主受教育年數(shù)參數(shù)所對應(yīng)的P值為0.0000，明顯小于，均可判斷戶主受教育年數(shù)對家庭書刊消費支出確實有顯著影響。（4）本模型說明家庭月平均收入和戶主受教育年數(shù)對家庭書刊消費支出有顯著影響，家庭月平均收入增加1元，家庭書刊年消費支出將增加0.086元，戶主受教育年數(shù)增加1年，家庭書刊年消費支出將增加52.37元。3.7 1970-1982年某國實際通貨膨脹率、失業(yè)率和預(yù)期通貨膨脹率（1）設(shè)回

4、歸模型為：對模型參數(shù)進(jìn)行估計，根據(jù)回歸結(jié)果得：模型估計結(jié)果表明：在預(yù)期通貨膨脹率不變的情況下，失業(yè)率每增長1個百分點，實際通貨膨脹率平均降低1.393115個百分點；在失業(yè)率不變的情況下，預(yù)期通貨膨脹率每增長1個百分點，實際通貨膨脹率平均增長1.480674個百分點。（2）檢驗t檢驗:給定顯著性水平，查表得，與對應(yīng)的t統(tǒng)計量分別為4.390321、-4.493196、8.217506，其絕對值均大于，這說明應(yīng)當(dāng)拒絕。表明在其他解釋變量不變的情況下，“失業(yè)率”和“預(yù)期通貨膨脹率”分別對被解釋變量“實際通貨膨脹率”有顯著影響。 F檢驗：針對，給定顯著性水平，由于，所以拒絕原價設(shè)，說明回歸方程顯著，

5、即“失業(yè)率”和“預(yù)期通貨膨脹率”聯(lián)合起來確實對被解釋變量“實際通貨膨脹率”有顯著影響。擬和優(yōu)度：，修正的可決系數(shù)： ，這說明模型對樣本擬和的很好。（2）修正的可決系數(shù)第四章4.6理論上認(rèn)為影響能源消費需求總量的因素。1985年-2002年（1） 建立對數(shù)線性多遠(yuǎn)回歸模型進(jìn)行最小二乘估計后得到 由此可見，該模型R2=0.989327，=0.989327可決系數(shù)很高，F(xiàn)檢驗值132.4180，明顯顯著。但是當(dāng)0.05時，=2.228,不僅、系數(shù)的t檢驗不顯著，而且、系數(shù)的符號與預(yù)期的相反，這表明很可能存在多重共線性。計算各解釋變量的相關(guān)關(guān)系，得到：由相關(guān)系數(shù)矩陣可以看出，各解釋變量相互之間的相關(guān)系

6、數(shù)較高，證實尋在多重共線性。修正多重共線性運用OLS方法逐一求y對各個解釋變量的回歸。結(jié)合經(jīng)濟(jì)意義和統(tǒng)計經(jīng)驗選出擬合效果最好的一元線性回歸方程。結(jié)果如下：從結(jié)果來看，對Y的線性關(guān)系最強(qiáng)，擬合程度最好。（）逐步回歸Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:09Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C86357.352841.43330.392180.0000X418.54

7、1547.8302782.3679290.0318X1-0.4756720.454371-1.0468820.3117R-squared0.922471 Mean dependent var116222.1Adjusted R-squared0.912134 S.D. dependent var23538.34S.E. of regression6977.273 Akaike info criterion20.68972Sum squared resid7.30E+08 Schwarz criterion20.83811Log likelihood-183.2074 F-statistic89

8、.23847Durbin-Watson stat0.335003 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:10Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C86259.352803.49130.768550.0000X419.573337.9580502.4595630.0265X2-0.5269020.454185-1.1

9、601030.2641R-squared0.923657 Mean dependent var116222.1Adjusted R-squared0.913477 S.D. dependent var23538.34S.E. of regression6923.737 Akaike info criterion20.67431Sum squared resid7.19E+08 Schwarz criterion20.82271Log likelihood-183.0688 F-statistic90.74026Durbin-Watson stat0.347108 Prob(F-statisti

10、c)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:11Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C85732.172895.25529.611270.0000X49.6590461.5200806.3543030.0000X30.1229720.2190400.5614120.5828R-squared0.918519 Mean dependent var116

11、222.1Adjusted R-squared0.907655 S.D. dependent var23538.34S.E. of regression7152.915 Akaike info criterion20.73944Sum squared resid7.67E+08 Schwarz criterion20.88783Log likelihood-183.6550 F-statistic84.54591Durbin-Watson stat0.314153 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least Squar

12、esDate: 12/12/14 Time: 16:11Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C84833.922917.47429.077860.0000X413.914993.0359934.5833410.0004X5-2.7506222.287992-1.2021990.2479R-squared0.924118 Mean dependent var116222.1Adjusted R-squared0.914001 S.D. dependent v

13、ar23538.34S.E. of regression6902.771 Akaike info criterion20.66825Sum squared resid7.15E+08 Schwarz criterion20.81664Log likelihood-183.0142 F-statistic91.33793Durbin-Watson stat0.313571 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:12Sample: 1985 2002Inc

14、luded observations: 18VariableCoefficientStd. Errort-StatisticProb. C83786.874994.56116.775620.0000X47.9024324.8479461.6300580.1239X6124.7067240.19210.5191960.6112R-squared0.918275 Mean dependent var116222.1Adjusted R-squared0.907379 S.D. dependent var23538.34S.E. of regression7163.592 Akaike info c

15、riterion20.74242Sum squared resid7.70E+08 Schwarz criterion20.89082Log likelihood-183.6818 F-statistic84.27173Durbin-Watson stat0.242887 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:12Sample: 1985 2002Included observations: 18VariableCoefficientStd. Erro

16、rt-StatisticProb. C164797.082887.901.9881910.0654X410.972360.99785210.995980.0000X7-1183.2041242.400-0.9523540.3560R-squared0.921550 Mean dependent var116222.1Adjusted R-squared0.911090 S.D. dependent var23538.34S.E. of regression7018.597 Akaike info criterion20.70153Sum squared resid7.39E+08 Schwar

17、z criterion20.84992Log likelihood-183.3137 F-statistic88.10265Durbin-Watson stat0.347945 Prob(F-statistic)0.000000從結(jié)果來看、對Y的擬合最好，繼續(xù)添加變量得如下結(jié)果：Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:16Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C836

18、94.695218.97316.036620.0000X49.52293816.727500.5692990.5782X5-4.6950627.648324-0.6138680.5491X10.4016601.5026570.2673000.7931R-squared0.924503 Mean dependent var116222.1Adjusted R-squared0.908326 S.D. dependent var23538.34S.E. of regression7126.884 Akaike info criterion20.77427Sum squared resid7.11E

19、+08 Schwarz criterion20.97213Log likelihood-182.9684 F-statistic57.14635Durbin-Watson stat0.297856 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:16Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C85164.534589.72718.

20、555470.0000X415.4022915.866560.9707390.3482X5-2.1280556.927209-0.3072020.7632X2-0.1311051.370944-0.0956310.9252R-squared0.924168 Mean dependent var116222.1Adjusted R-squared0.907918 S.D. dependent var23538.34S.E. of regression7142.714 Akaike info criterion20.77870Sum squared resid7.14E+08 Schwarz cr

21、iterion20.97656Log likelihood-183.0083 F-statistic56.87266Durbin-Watson stat0.321437 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:17Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C84744.583006.24328.189530.0000X41

22、3.200773.5204103.7497830.0022X5-2.6352302.366856-1.1133880.2843X30.0959110.2186660.4386160.6676R-squared0.925147 Mean dependent var116222.1Adjusted R-squared0.909107 S.D. dependent var23538.34S.E. of regression7096.455 Akaike info criterion20.76571Sum squared resid7.05E+08 Schwarz criterion20.96357L

23、og likelihood-182.8914 F-statistic57.67760Durbin-Watson stat0.362047 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:17Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C55475.695935.6289.3462210.0000X40.9911003.0994580

24、.3197660.7539X5-15.107332.758634-5.4763790.0001X61445.868279.05585.1812850.0001R-squared0.973991 Mean dependent var116222.1Adjusted R-squared0.968418 S.D. dependent var23538.34S.E. of regression4183.077 Akaike info criterion19.70861Sum squared resid2.45E+08 Schwarz criterion19.90647Log likelihood-17

25、3.3775 F-statistic174.7603Durbin-Watson stat0.945925 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:18Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C226134.783767.342.6995570.0173X417.045643.4127464.9947010.0002X5-

26、4.3665302.361625-1.8489510.0857X7-2128.6081261.221-1.6877360.1136R-squared0.936947 Mean dependent var116222.1Adjusted R-squared0.923436 S.D. dependent var23538.34S.E. of regression6513.120 Akaike info criterion20.59415Sum squared resid5.94E+08 Schwarz criterion20.79202Log likelihood-181.3474 F-stati

27、stic69.34517Durbin-Watson stat0.656021 Prob(F-statistic)0.000000從結(jié)果來看，、和的擬合有度最好，但的系數(shù)為負(fù)，不符合經(jīng)濟(jì)意義，綜合比較、對Y的解釋是最好的。再次添加到方程中得到結(jié)果如下：Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:26Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C84005.485451.19315

28、.410470.0000X10.2635221.5946510.1652540.8713X30.0869710.2330480.3731890.7150X410.3858017.420810.5961720.5613X5-3.9216978.162314-0.4804640.6389R-squared0.925304 Mean dependent var116222.1Adjusted R-squared0.902320 S.D. dependent var23538.34S.E. of regression7356.616 Akaike info criterion20.87472Sum s

29、quared resid7.04E+08 Schwarz criterion21.12205Log likelihood-182.8725 F-statistic40.25953Durbin-Watson stat0.347820 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:26Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-StatisticProb. C85

30、450.374766.41017.927620.0000X2-0.2837601.450401-0.1956420.8479X30.1064030.2328480.4569640.6552X416.3417016.463690.9925900.3390X5-1.2751437.371855-0.1729740.8653R-squared0.925367 Mean dependent var116222.1Adjusted R-squared0.902402 S.D. dependent var23538.34S.E. of regression7353.521 Akaike info crit

31、erion20.87388Sum squared resid7.03E+08 Schwarz criterion21.12120Log likelihood-182.8649 F-statistic40.29616Durbin-Watson stat0.382583 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:27Sample: 1985 2002Included observations: 18VariableCoefficientStd. Errort-

32、StatisticProb. C55376.166034.5219.1765630.0000X61446.306283.63585.0991650.0002X30.0973000.1310100.7426860.4709X40.2626213.2994900.0795940.9378X5-14.994012.808052-5.3396470.0001R-squared0.975050 Mean dependent var116222.1Adjusted R-squared0.967373 S.D. dependent var23538.34S.E. of regression4251.722

33、Akaike info criterion19.77817Sum squared resid2.35E+08 Schwarz criterion20.02549Log likelihood-173.0035 F-statistic127.0099Durbin-Watson stat1.195203 Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 16:27Sample: 1985 2002Included observations: 18VariableCoeffic

34、ientStd. Errort-StatisticProb. C225770.086256.722.6174190.0213X7-2124.4251298.678-1.6358370.1258X30.0935040.2066520.4524710.6584X416.343193.8416754.2541830.0009X5-4.2508572.445103-1.7385190.1057R-squared0.937925 Mean dependent var116222.1Adjusted R-squared0.918824 S.D. dependent var23538.34S.E. of r

35、egression6706.383 Akaike info criterion20.68964Sum squared resid5.85E+08 Schwarz criterion20.93697Log likelihood-181.2068 F-statistic49.10570Durbin-Watson stat0.726028 Prob(F-statistic)0.000000從擬合結(jié)果和經(jīng)濟(jì)意義上來看，加入似乎是合適的，但對Y的影響并不顯著，故將刪去。得到最終結(jié)果。Date: 12/12/14 Time: 16:17Sample: 1985 2002Included observati

36、ons: 18VariableCoefficientStd. Errort-StatisticProb. C84744.583006.24328.189530.0000X413.200773.5204103.7497830.0022X5-2.6352302.366856-1.1133880.2843X30.0959110.2186660.4386160.6676R-squared0.925147 Mean dependent var116222.1Adjusted R-squared0.909107 S.D. dependent var23538.34S.E. of regression709

37、6.455 Akaike info criterion20.76571Sum squared resid7.05E+08 Schwarz criterion20.96357Log likelihood-182.8914 F-statistic57.67760Durbin-Watson stat0.362047 Prob(F-statistic)0.000000第五章5.9 1978年-2000年四川省農(nóng)村人均純收入和人均生活費支出。1、估計參數(shù)Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 20:40Sample(a

38、djusted): 1978 2000Included observations: 23 after adjusting endpointsVariableCoefficientStd. Errort-StatisticProb. C41.4444819.567912.1179820.0463X0.8163490.02104838.785530.0000R-squared0.986232 Mean dependent var632.6430Adjusted R-squared0.985577 S.D. dependent var489.9925S.E. of regression58.8465

39、0 Akaike info criterion11.07068Sum squared resid72721.12 Schwarz criterion11.16942Log likelihood-125.3129 F-statistic1504.317Durbin-Watson stat0.389107 Prob(F-statistic)0.000000（1） 建立樣本回歸函數(shù)：=41.44448+0.816349X（2.117982） （38.78553）R2=0.986232 F=1504.317 DW=0.389107 2、檢驗?zāi)Ｐ偷漠惙讲?X與E2的散點圖 因此模型存在異方差 Y與E2的

40、散點圖 如圖看出存在異方差（2） Whitet檢驗White Heteroskedasticity Test:F-statistic8.702383 Probability0.001910Obs*R-squared10.70210 Probability0.004743Test Equation:Dependent Variable: RESID2Method: Least SquaresDate: 12/12/14 Time: 20:49Sample: 1978 2000Included observations: 23VariableCoefficientStd. Errort-Stati

41、sticProb. C-1326.7362317.812-0.5724090.5734X6.9016176.8054251.0141350.3226X2-0.0005900.003347-0.1761550.8619R-squared0.465309 Mean dependent var3161.788Adjusted R-squared0.411840 S.D. dependent var5009.390S.E. of regression3841.782 Akaike info criterion19.46637Sum squared resid2.95E+08 Schwarz crite

42、rion19.61448Log likelihood-220.8632 F-statistic8.702383Durbin-Watson stat1.599367 Prob(F-statistic)0.001910由上表可知，nR2=10.70210 給定，在自由度為2下，查卡方分布表，得臨界值為，顯然，nR2=10.70210>5.9915,則拒絕原假設(shè)，說明模型存在異方差。3、修正異方差。 取權(quán)數(shù)為W2Dependent Variable: YMethod: Least SquaresDate: 12/12/14 Time: 21:08Sample: 1978 2000Include

43、d observations: 23Weighting series: W2VariableCoefficientStd. Errort-StatisticProb. C6.6590270.25376126.241330.0000X0.8686910.000985881.79380.0000Weighted StatisticsR-squared1.000000 Mean dependent var224.0761Adjusted R-squared1.000000 S.D. dependent var988.1865S.E. of regression0.206384 Akaike info

44、 criterion-0.235219Sum squared resid0.894478 Schwarz criterion-0.136481Log likelihood4.705022 F-statistic777560.3Durbin-Watson stat1.281139 Prob(F-statistic)0.000000Unweighted StatisticsR-squared0.980282 Mean dependent var633.0004Adjusted R-squared0.979343 S.D. dependent var490.5345S.E. of regression70.50182 Sum squared resid104380.6Durbin-Watson stat0.279924Y = 6.65902728 + 0.8686910728*X White 檢驗White Heteroskedasticity Test:F-statistic1.021337 Prob