《计量经济学课后题答案.doc》由会员分享，可在线阅读，更多相关《计量经济学课后题答案.doc（10页珍藏版）》请在三一办公上搜索。
1、3.5 一 模型设置 线形图 可以看出Y X1 X2都是逐年增长的，但增长率有所变动，而X2 在多数年份呈现出水平波动。说明变量间不一定是线性关系，可探索建立该地区城镇居民人均全年耐用消费品支出关于人均年可支配收入和耐用消费品价格指数的回归模型： Y = C + aX1 +bX2二 参数估计利用Eviews 估计参数模型可得回归结果Dependent Variable: YMethod: Least SquaresDate: 05/13/13 Time: 10:20Sample: 1991 2001Included observations: 11VariableCoefficientStd.
2、 Errort-StatisticProb.C158.5398121.80711.3015640.2293X10.0494040.00468410.547860.0000X2-0.9116840.989546-0.9213160.3838R-squared0.947989Mean dependent var190.4827Adjusted R-squared0.934986S.D. dependent var79.29127S.E. of regression20.21757Akaike info criterion9.077982Sum squared resid3270.001Schwar
3、z criterion9.186499Log likelihood-46.92890Hannan-Quinn criter.9.009577F-statistic72.90647Durbin-Watson stat1.035840Prob(F-statistic)0.000007 Y = 158.539835512 + 0.0494037966638*X1 - 0.911684215976*X2 121.8071 0.004684 0.989546 t= 1.301564 10.54786 -0.921316 =0.947989 =0.93499 F=72.90647 n =11四模型检验是由
4、估计和检验结果可看出，该地区人均年可支配收入的参数的t检验值为10.54786，其绝对值大于临界值；而且对应的P值为0.0000，也明显小于。说明人均年可支配收入对该地区城镇居民人均全年耐用消费品支出确实有显著影响。但是，该地区耐用消费品价格指数的参数的t检验值为-0.921316，其绝对值小于临界值；而且对应的P值为0.3838，也明显大于。这说明该地区耐用消费品价格指数对城镇居民人均全年耐用消费品支出并没有显著影响.这样的结果并不合理，耐用消费品价格指数对人均耐用消费品支出应有影响，在耐用消费品价格指数上升时，人均耐用消费品支出会有所下降。由于y与增长趋势并不明显 ，为减小X2的变化对其取
5、对Dependent Variable: LNYMethod: Least SquaresDate: 05/13/13 Time: 10:34Sample: 1991 2001Included observations: 11VariableCoefficientStd. Errort-StatisticProb.C4.9940763.9385411.2680010.2405X10.0002483.01E-058.2599050.0000LNX2-0.1194440.819221-0.1458020.8877R-squared0.923317Mean dependent var5.170096
6、Adjusted R-squared0.904146S.D. dependent var0.418850S.E. of regression0.129677Akaike info criterion-1.020536Sum squared resid0.134529Schwarz criterion-0.912019Log likelihood8.612947Hannan-Quinn criter.-1.088940F-statistic48.16271Durbin-Watson stat0.914876Prob(F-statistic)0.000035LNY = 4.99407624656
7、+ 0.000248245410851*X1 - 0.119444065471*LNX2Lnx2系数仍没有通过t检验。为减小X1的变化对其取对：Dependent Variable: YMethod: Least SquaresDate: 05/13/13 Time: 10:54Sample: 1991 2001Included observations: 11VariableCoefficientStd. Errort-StatisticProb.C-720.8880161.8193-4.4548950.0021LNX1142.337620.480746.9498260.0001X2-1.6
8、129541.518866-1.0619460.3193R-squared0.889828Mean dependent var190.4827Adjusted R-squared0.862284S.D. dependent var79.29127S.E. of regression29.42503Akaike info criterion9.828569Sum squared resid6926.658Schwarz criterion9.937086Log likelihood-51.05713Hannan-Quinn criter.9.760164F-statistic32.30672Du
9、rbin-Watson stat0.660676Prob(F-statistic)0.000147Y = -720.888011503 + 142.337598779*LNX1 - 1.61295354357*X2X2系数仍没有通过t检验。3.6 一 模型设置 线形图二 参数估计(1)利用Eviews 估计参数对数模型可得回归结果Dependent Variable: LNYMethod: Least SquaresDate: 05/13/13 Time: 11:09Sample: 1960 1982Included observations: 23VariableCoefficientStd
10、. Errort-StatisticProb.C1.5495040.09011317.195080.0000LNX10.9969230.01911052.166340.0000LNX2-0.3313640.024310-13.630860.0000R-squared0.994130Mean dependent var4.412077Adjusted R-squared0.993543S.D. dependent var0.224107S.E. of regression0.018008Akaike info criterion-5.074916Sum squared resid0.006486
11、Schwarz criterion-4.926808Log likelihood61.36153Hannan-Quinn criter.-5.037667F-statistic1693.652Durbin-Watson stat0.807846Prob(F-statistic)0.000000LNY = 1.54950389461 + 0.996922957751*LNX1 - 0.331364251849*LNX20.090113 0.019110 0.024310t= 17.19508 52.16634 -13.63086=0.994 F=1693.652 n=23回归系数意义：0.994
12、指在其他变量不变的情况下，实际GDP指数每上升1%，能源需求指数平均上升0.9969%；- 0.33136指在其他变量不变的情况下，能源价格指数每上升1%，能源需求指数平均下降0.33136%。P值检验：回归系数P值均远小于0.05，回归系数显著，即实际GDP指数、能源价格指数对能源需求指数有显著影响。(2) 利用Eviews 估计参数线性回归模型可得回归结果Dependent Variable: YMethod: Least SquaresDate: 05/13/13 Time: 11:17Sample: 1960 1982Included observations: 23VariableC
13、oefficientStd. Errort-StatisticProb.C28.255061.42148819.877090.0000X10.9808490.01945450.419000.0000X2-0.2584260.015282-16.910310.0000R-squared0.993890Mean dependent var84.34348Adjusted R-squared0.993279S.D. dependent var17.50999S.E. of regression1.435479Akaike info criterion3.681982Sum squared resid
14、41.21199Schwarz criterion3.830090Log likelihood-39.34279Hannan-Quinn criter.3.719230F-statistic1626.707Durbin-Watson stat0.977840Prob(F-statistic)0.000000Y = 28.2550550619 + 0.98084861146*X1 - 0.258425759715*X2 1.421488 0.019454 0.015282 t= 19.87709 50.41900 -16.91031=0.993890 F =1626.707 n=23回归系数意义
15、：说明收入GDP指数增加1个单位时，平均说来能源需求指数将增长0.980849个单位; 价格指数增加1个单位时，平均说来能源需求指数将降低0.258426个单位 由P值可知, 收入和价格对能源需求的影响是显著的.P值检验：回归系数P值均远小于0.05，回归系数显著，即实际GDP指数、能源价格指数对能源需求指数有显著影响。3.71) 根据所建立得线性回归模型：预期的x1符号为正，x2符号为正，x3符号为正，x4、x5、x6符号为负。2)Dependent Variable: YMethod: Least SquaresDate: 05/15/13 Time: 11:18Sample: 1974
16、1987Included observations: 14VariableCoefficientStd. Errort-StatisticProb.C-3.49656330.00659-0.1165260.9101X20.1253300.0591392.1192450.0669X30.0736670.0378771.9448970.0877X42.6775891.2572932.1296460.0658X53.4534482.4508501.4090820.1965X6-4.4911172.214862-2.0277190.0771R-squared0.970442Mean dependent
17、 var142.7129Adjusted R-squared0.951968S.D. dependent var26.09805S.E. of regression5.719686Akaike info criterion6.623232Sum squared resid261.7185Schwarz criterion6.897114Log likelihood-40.36262Hannan-Quinn criter.6.597879F-statistic52.53086Durbin-Watson stat1.972755Prob(F-statistic)0.000007Y = -3.496
18、5627 + 0.12532958*X2 + 0.07366672*X3 + 2.67758946*X4 + 3.4534482*X5 - 4.49111676*X6 30.00659 0.059139 0.037877 1.257293 2.450850 2.214862 t= -0.116526 2.119245 1.944897 2.129646 1.409082 -2.027719=0.970442 F=52.53086 n=14只有X2和X6的符号与预期一致，X3、X4、X5的符号均与预期相反3）拟合优度：=0.970442，说明模型对样本的拟合比较好.F检验：在时， 临界值F=3.
19、69，而方程F=52.53086，说明各解释变量结合起来对粮食年销售量有显著影响。t 检验：，所以经过对比得任何一个回归系数都不能通过检验，均需接受各系数为0的假设。对模型及各个解释变量的显著性发现：虽然可决系数和修正的可决系数都较高，F=13.5823,检验表明也显著,但是所有的解释变量的t检验却都不显著, 这种矛盾现象说明此模型存在严重的问题（存在严重多重共线性）。4.3 在Ln=模型下X1=GDP X2=CPI（1） 对数回归模型Dependent Variable: LNYMethod: Least SquaresDate: 05/13/13 Time: 11:35Sample: 19
20、85 2007Included observations: 23VariableCoefficientStd. Errort-StatisticProb.C-3.5571440.304243-11.691780.0000LNGDP1.5841250.08158719.416490.0000LNCPI-0.8108780.186482-4.3482870.0003R-squared0.993281Mean dependent var9.155303Adjusted R-squared0.992609S.D. dependent var1.276500S.E. of regression0.109
21、745Akaike info criterion-1.460211Sum squared resid0.240878Schwarz criterion-1.312103Log likelihood19.79242Hannan-Quinn criter.-1.422962F-statistic1478.217Durbin-Watson stat0.934574Prob(F-statistic)0.000000LNY = -3.55714351097 + 1.5841246105*LNGDP - 0.810878444967*LNCPI 0.304243 0.081587 0.186482t= -
22、11.69178 19.41649 -4.348287=0.99328 F=1478.217 n=23(2)居民消费价格指数的回归系数的符号不能进行合理的经济意义解释，且CPI与进口之间的简单相关系数呈现正向变动。可能数据中有多重共线性。 计算相关系数:相关系数矩阵：1.0000000.9852320.7174510.9852321.0000000.8147660.7174510.8147661.000000(3)Dependent Variable: LNYMethod: Least SquaresDate: 05/15/13 Time: 11:43Sample: 1985 2007Incl
23、uded observations: 23VariableCoefficientStd. Errort-StatisticProb.C-4.3138180.339695-12.699090.0000LNGDP1.2436830.03123439.818360.0000R-squared0.986928Mean dependent var9.155303Adjusted R-squared0.986306S.D. dependent var1.276500S.E. of regression0.149380Akaike info criterion-0.881710Sum squared res
24、id0.468600Schwarz criterion-0.782971Log likelihood12.13966Hannan-Quinn criter.-0.856877F-statistic1585.502Durbin-Watson stat0.490475Prob(F-statistic)0.000000LNY = -4.31381765484 + 1.24368337208*LNGDPDependent Variable: LNYMethod: Least SquaresDate: 05/15/13 Time: 11:49Sample: 1985 2007Included obser
25、vations: 23VariableCoefficientStd. Errort-StatisticProb.C-5.4424201.253662-4.3412180.0003LNCPI2.6637900.22804611.680910.0000R-squared0.866619Mean dependent var9.155303Adjusted R-squared0.860268S.D. dependent var1.276500S.E. of regression0.477166Akaike info criterion1.441037Sum squared resid4.781435S
26、chwarz criterion1.539775Log likelihood-14.57192Hannan-Quinn criter.1.465869F-statistic136.4437Durbin-Watson stat0.152312Prob(F-statistic)0.000000LNY = -5.442420219 + 2.66378992828*LNCPIDependent Variable: LNGDPMethod: Least SquaresDate: 05/15/13 Time: 11:54Sample: 1985 2007Included observations: 23V
27、ariableCoefficientStd. Errort-StatisticProb.C-1.1901060.771199-1.5431890.1377LNCPI2.1934310.14028415.635610.0000R-squared0.920896Mean dependent var10.83002Adjusted R-squared0.917129S.D. dependent var1.019656S.E. of regression0.293532Akaike info criterion0.469282Sum squared resid1.809383Schwarz crite
28、rion0.568020Log likelihood-3.396739Hannan-Quinn criter.0.494114F-statistic244.4722Durbin-Watson stat0.141471Prob(F-statistic)0.000000LNGDP = -1.19010631836 + 2.1934312176*LNCPI 可以看到在对模型进行变形后，各解释变量都能通过t检验，但是变形后反而变小，模型没有前面拟合得好。(4) 利用逐步回归法，把没有通过显著性检验的变量剔除，看和F是否有所改善，如果有改善，则需将变量剔除；若没有改善，则保留该变量。Acknowledg
29、ements My deepest gratitude goes first and foremost to Professor aaa , my supervisor, for her constant encouragement and guidance. She has walked me through all the stages of the writing of this thesis. Without her consistent and illuminating instruction, this thesis could not havereached its presen
30、t form. Second, I would like to express my heartfelt gratitude to Professor aaa, who led me into the world of translation. I am also greatly indebted to the professors and teachers at the Department of English: Professor dddd, Professor ssss, who have instructed and helped me a lot in the past two y
31、ears. Last my thanks would go to my beloved family for their loving considerations and great confidence in me all through these years. I also owe my sincere gratitude to my friends and my fellow classmates who gave me their help and time in listening to me and helping me work out my problems during
32、the difficult course of the thesis. My deepest gratitude goes first and foremost to Professor aaa , my supervisor, for her constant encouragement and guidance. She has walked me through all the stages of the writing of this thesis. Without her consistent and illuminating instruction, this thesis cou
33、ld not havereached its present form. Second, I would like to express my heartfelt gratitude to Professor aaa, who led me into the world of translation. I am also greatly indebted to the professors and teachers at the Department of English: Professor dddd, Professor ssss, who have instructed and help
34、ed me a lot in the past two years. Last my thanks would go to my beloved family for their loving considerations and great confidence in me all through these years. I also owe my sincere gratitude to my friends and my fellow classmates who gave me their help and time in listening to me and helping me work out my problems during the difficult course of the thesis.
