1、第二章2.2（1）对于浙江省预算收入与全省生产总值的模型，用 Eviews 分析结果如下：Dependent Variable: YMethod: Least SquaresDate: 12/03/14 Time: 17:00Sample (adjusted): 1 33Included observations: 33 after adjustmentsVariable Coefficient Std. Error t-Statistic Prob. X 0.176124 0.004072 43.25639 0.0000C -154.3063 39.08196 -3.948274 0.000
2、4R-squared 0.983702 Mean dependent var 902.5148Adjusted R-squared 0.983177 S.D. dependent var 1351.009S.E. of regression 175.2325 Akaike info criterion 13.22880Sum squared resid 951899.7 Schwarz criterion 13.31949Log likelihood -216.2751 Hannan-Quinn criter. 13.25931F-statistic 1871.115 Durbin-Watso
3、n stat 0.100021Prob(F-statistic) 0.000000由上可知，模型的参数：斜率系数 0.176124，截距为154.3063关于浙江省财政预算收入与全省生产总值的模型，检验模型的显著性：1）可决系数为 0.983702，说明所建模型整体上对样本数据拟合较好。2）对于回归系数的 t 检验： t（2 ）=43.25639t 0.025(31)=2.0395，对斜率系数的显著性检验表明，全省生产总值对财政预算总收入有显著影响。用规范形式写出检验结果如下：Y=0.176124X154.3063(0.004072) (39.08196)t= (43.25639) （-3.9
4、48274）R2=0.983702 F=1871.115 n=33经济意义是：全省生产总值每增加 1 亿元，财政预算总收入增加 0.176124 亿元。（2）当 x=32000 时，进行点预测，由上可知 Y=0.176124X154.3063，代入可得：Y= Y=0.176124*32000154.3063=5481.6617进行区间预测：先由 Eviews 分析：X YMean 6000.441 902.5148Median 2689.280 209.3900Maximum 27722.31 4895.410Minimum 123.7200 25.87000Std. Dev. 7608.02
5、1 1351.009Skewness 1.432519 1.663108Kurtosis 4.010515 4.590432Jarque-Bera 12.69068 18.69063Probability 0.001755 0.000087Sum 198014.5 29782.99Sum Sq. Dev. 1.85E+09 58407195Observations 33 33由上表可知，x2=（X iX） 2=2x(n1)= 7608.0212 x (331)=1852223.473(XfX)2=(32000 6000.441)2=675977068.2当 Xf=32000 时，将相关数据代入
6、计算得到：5481.66172.0395x175.2325x1/33+1852223.473/675977068.2Yf5481.6617+2.0395x175.2325x1/33+1852223.473/675977068.2即 Yf 的置信区间为（5481.6617 64.9649, 5481.6617+64.9649）(3) 对于浙江省预算收入对数与全省生产总值对数的模型，由 Eviews 分析结果如下：Dependent Variable: LNYMethod: Least SquaresDate: 12/03/14 Time: 18:00Sample (adjusted): 1 33
7、Included observations: 33 after adjustmentsVariable Coefficient Std. Error t-Statistic Prob. LNX 0.980275 0.034296 28.58268 0.0000C -1.918289 0.268213 -7.152121 0.0000R-squared 0.963442 Mean dependent var 5.573120Adjusted R-squared 0.962263 S.D. dependent var 1.684189S.E. of regression 0.327172 Akai
8、ke info criterion 0.662028Sum squared resid 3.318281 Schwarz criterion 0.752726Log likelihood -8.923468 Hannan-Quinn criter. 0.692545F-statistic 816.9699 Durbin-Watson stat 0.096208Prob(F-statistic) 0.000000模型方程为：lnY=0.980275lnX-1.918289由上可知，模型的参数：斜率系数为 0.980275，截距为-1.918289关于浙江省财政预算收入与全省生产总值的模型，检验其
9、显著性：1）可决系数为 0.963442，说明所建模型整体上对样本数据拟合较好。2）对于回归系数的 t 检验： t（ 2）= 28.58268t0.025(31)=2.0395，对斜率系数的显著性检验表明，全省生产总值对财政预算总收入有显著影响。经济意义：全省生产总值每增长 1%，财政预算总收入增长 0.980275%2.4（1 ）对建筑面积与建造单位成本模型，用 Eviews 分析结果如下：Dependent Variable: YMethod: Least SquaresDate: 12/01/14 Time: 12:40Sample: 1 12Included observations:
10、 12Variable Coefficient Std. Error t-Statistic Prob. X -64.18400 4.809828 -13.34434 0.0000C 1845.475 19.26446 95.79688 0.0000R-squared 0.946829 Mean dependent var 1619.333Adjusted R-squared 0.941512 S.D. dependent var 131.2252S.E. of regression 31.73600 Akaike info criterion 9.903792Sum squared resi
11、d 10071.74 Schwarz criterion 9.984610Log likelihood -57.42275 Hannan-Quinn criter. 9.873871F-statistic 178.0715 Durbin-Watson stat 1.172407Prob(F-statistic) 0.000000由上可得：建筑面积与建造成本的回归方程为：Y=1845.475-64.18400X（2）经济意义：建筑面积每增加 1 万平方米，建筑单位成本每平方米减少 64.18400 元。（3）首先进行点预测，由 Y=1845.475-64.18400X 得，当 x=4.5，y=1
12、556.647再进行区间估计：用 Eviews 分析：Y XMean 1619.333 3.523333Median 1630.000 3.715000Maximum 1860.000 6.230000Minimum 1419.000 0.600000Std. Dev. 131.2252 1.989419Skewness 0.003403 -0.060130Kurtosis 2.346511 1.664917Jarque-Bera 0.213547 0.898454Probability 0.898729 0.638121Sum 19432.00 42.28000Sum Sq. Dev. 18
13、9420.7 43.53567Observations 12 12由上表可知，x2=（X iX） 2=2x(n1)= 1.9894192 x (121)=43.5357(XfX)2=(4.5 3.523333)2=0.95387843当 Xf=4.5 时，将相关数据代入计算得到：1556.6472.228x31.73600x1/12+43.5357/0.95387843Yf1556.647+2.228x31.73600x1/12+43.5357/0.95387843即 Yf 的置信区间为（1556.647 478.1231, 1556.647+478.1231）第三章3.2）对出口货物总额计量
14、经济模型，用 Eviews 分析结果如下：Dependent Variable: YMethod: Least SquaresDate: 12/01/14 Time: 20:25Sample: 1994 2011Included observations: 18Variable Coefficient Std. Error t-Statistic Prob. X2 0.135474 0.012799 10.58454 0.0000X3 18.85348 9.776181 1.928512 0.0729C -18231.58 8638.216 -2.110573 0.0520R-squared
15、0.985838 Mean dependent var 6619.191Adjusted R-squared 0.983950 S.D. dependent var 5767.152S.E. of regression 730.6306 Akaike info criterion 16.17670Sum squared resid 8007316. Schwarz criterion 16.32510Log likelihood -142.5903 Hannan-Quinn criter. 16.19717F-statistic 522.0976 Durbin-Watson stat 1.17
16、3432Prob(F-statistic) 0.000000由上可知，模型为：Y = 0.135474X2 + 18.85348X3 - 18231.58对模型进行检验：1）可决系数是 0.985838，修正的可决系数为 0.983950，说明模型对样本拟合较好2） F 检验，F=522.0976F （ 2,15）=4.77，回归方程显著3） t 检验，t 统计量分别为 X2 的系数对应 t 值为 10.58454，大于 t（15）=2.131，系数是显著的，X3 的系数对应 t 值为 1.928512，小于 t（15）=2.131，说明此系数是不显著的。（2 ）对于对数模型，用 Eviews
17、 分析结果如下：Dependent Variable: LNYMethod: Least SquaresDate: 12/01/14 Time: 20:25Sample: 1994 2011Included observations: 18Variable Coefficient Std. Error t-Statistic Prob. LNX2 1.564221 0.088988 17.57789 0.0000LNX3 1.760695 0.682115 2.581229 0.0209C -20.52048 5.432487 -3.777363 0.0018R-squared 0.98629
18、5 Mean dependent var 8.400112Adjusted R-squared 0.984467 S.D. dependent var 0.941530S.E. of regression 0.117343 Akaike info criterion -1.296424Sum squared resid 0.206540 Schwarz criterion -1.148029Log likelihood 14.66782 Hannan-Quinn criter. -1.275962F-statistic 539.7364 Durbin-Watson stat 0.686656P
19、rob(F-statistic) 0.000000由上可知，模型为：LNY=-20.52048+1.564221 LNX2+1.760695 LNX3对模型进行检验：1）可决系数是 0.986295，修正的可决系数为 0.984467，说明模型对样本拟合较好。2） F 检验，F=539.7364 F（2,15）=4.77，回归方程显著。3） t 检验，t 统计量分别为-3.777363，17.57789 ，2.581229，均大于 t（15 ）=2.131，所以这些系数都是显著的。（3 ）（1）式中的经济意义：工业增加 1 亿元，出口货物总额增加 0.135474 亿元，人民币汇率增加 1，出
20、口货物总额增加 18.85348 亿元。（2）式中的经济意义：工业增加额每增加1% ，出口货物总额增加1.564221%，人民币汇率每增加1%，出口货物总额增加1.760695%3.3（1 ）对家庭书刊消费对家庭月平均收入和户主受教育年数计量模型，由 Eviews 分析结果如下：Dependent Variable: YMethod: Least SquaresDate: 12/01/14 Time: 20:30Sample: 1 18Included observations: 18Variable Coefficient Std. Error t-Statistic Prob. X 0.0
21、86450 0.029363 2.944186 0.0101T 52.37031 5.202167 10.06702 0.0000C -50.01638 49.46026 -1.011244 0.3279R-squared 0.951235 Mean dependent var 755.1222Adjusted R-squared 0.944732 S.D. dependent var 258.7206S.E. of regression 60.82273 Akaike info criterion 11.20482Sum squared resid 55491.07 Schwarz crit
22、erion 11.35321Log likelihood -97.84334 Hannan-Quinn criter. 11.22528F-statistic 146.2974 Durbin-Watson stat 2.605783Prob(F-statistic) 0.000000模型为：Y = 0.086450X + 52.37031T-50.01638对模型进行检验：1）可决系数是 0.951235，修正的可决系数为 0.944732，说明模型对样本拟合较好。2） F 检验，F=539.7364 F（2,15）=4.77，回归方程显著。3） t 检验，t 统计量分别为 2.944186，
23、10.06702，均大于 t（15）=2.131，所以这些系数都是显著的。经济意义：家庭月平均收入增加 1 元，家庭书刊年消费支出增加 0.086450 元，户主受教育年数增加 1 年，家庭书刊年消费支出增加 52.37031 元。（2）用 Eviews 分析：Dependent Variable: YMethod: Least SquaresDate: 12/01/14 Time: 22:30Sample: 1 18Included observations: 18Variable Coefficient Std. Error t-Statistic Prob. T 63.01676 4.5
24、48581 13.85416 0.0000C -11.58171 58.02290 -0.199606 0.8443R-squared 0.923054 Mean dependent var 755.1222Adjusted R-squared 0.918245 S.D. dependent var 258.7206S.E. of regression 73.97565 Akaike info criterion 11.54979Sum squared resid 87558.36 Schwarz criterion 11.64872Log likelihood -101.9481 Hanna
25、n-Quinn criter. 11.56343F-statistic 191.9377 Durbin-Watson stat 2.134043Prob(F-statistic) 0.000000Dependent Variable: XMethod: Least SquaresDate: 12/01/14 Time: 22:34Sample: 1 18Included observations: 18Variable Coefficient Std. Error t-Statistic Prob. T 123.1516 31.84150 3.867644 0.0014C 444.5888 4
26、06.1786 1.094565 0.2899R-squared 0.483182 Mean dependent var 1942.933Adjusted R-squared 0.450881 S.D. dependent var 698.8325S.E. of regression 517.8529 Akaike info criterion 15.44170Sum squared resid 4290746. Schwarz criterion 15.54063Log likelihood -136.9753 Hannan-Quinn criter. 15.45534F-statistic
27、 14.95867 Durbin-Watson stat 1.052251Prob(F-statistic) 0.001364以上分别是 y 与 T，X 与 T 的一元回归模型分别是：Y = 63.01676T - 11.58171X = 123.1516T + 444.5888（3 ）对残差进行模型分析，用Eviews分析结果如下：Dependent Variable: E1Method: Least SquaresDate: 12/03/14 Time: 20:39Sample: 1 18Included observations: 18Variable Coefficient Std.
28、Error t-Statistic Prob. E2 0.086450 0.028431 3.040742 0.0078C 3.96E-14 13.88083 2.85E-15 1.0000R-squared 0.366239 Mean dependent var 2.30E-14Adjusted R-squared 0.326629 S.D. dependent var 71.76693S.E. of regression 58.89136 Akaike info criterion 11.09370Sum squared resid 55491.07 Schwarz criterion 1
29、1.19264Log likelihood -97.84334 Hannan-Quinn criter. 11.10735F-statistic 9.246111 Durbin-Watson stat 2.605783Prob(F-statistic) 0.007788模型为：E1 = 0.086450E2 + 3.96e-14参数：斜率系数 为 0.086450，截距为 3.96e-14（3 ）由上可知，2 与 2 的系数是一样的。回归系数与被解释变量的残差系数是一样的，它们的变化规律是一致的。第五章5.3(1)由 Eviews 软件分析得：Dependent Variable: YMeth
30、od: Least SquaresDate: 12/10/14 Time: 16:00Sample: 1 31Included observations: 31Variable Coefficient Std. Error t-Statistic Prob. X 1.244281 0.079032 15.74411 0.0000C 242.4488 291.1940 0.832602 0.4119R-squared 0.895260 Mean dependent var 4443.526Adjusted R-squared 0.891649 S.D. dependent var 1972.07
31、2S.E. of regression 649.1426 Akaike info criterion 15.85152Sum squared resid 12220196 Schwarz criterion 15.94404Log likelihood -243.6986 Hannan-Quinn criter. 15.88168F-statistic 247.8769 Durbin-Watson stat 1.078581Prob(F-statistic) 0.000000由上表可知，2007 年我国农村居民家庭人均消费支出（x）对人均纯收入（y ）的模型为：Y=1.244281X+242.
32、4488（2）由图形法检验01,0,02,0,03,0,04,0,05,0,06,0,02,04,06,08,010,XE2由上图可知，模型可能存在异方差。Goldfeld-Quanadt 检验1）定义区间为 1-12 时，由软件分析得：Dependent Variable: Y1Method: Least SquaresDate: 12/10/14 Time: 11:34Sample: 1 12Included observations: 12Variable Coefficient Std. Error t-Statistic Prob. X1 1.485296 0.500386 2.96
33、8297 0.0141C -550.5492 1220.063 -0.451247 0.6614R-squared 0.468390 Mean dependent var 3052.950Adjusted R-squared 0.415229 S.D. dependent var 550.5148S.E. of regression 420.9803 Akaike info criterion 15.07406Sum squared resid 1772245. Schwarz criterion 15.15488Log likelihood -88.44437 Hannan-Quinn cr
34、iter. 15.04414F-statistic 8.810789 Durbin-Watson stat 2.354167Prob(F-statistic) 0.014087得e 1i2=1772245.2）定义区间为 20-31 时，由软件分析得：Dependent Variable: Y1Method: Least SquaresDate: 12/10/14 Time: 16:36Sample: 20 31Included observations: 12Variable Coefficient Std. Error t-Statistic Prob. X1 1.086940 0.148
35、863 7.301623 0.0000C 1173.307 733.2520 1.600141 0.1407R-squared 0.842056 Mean dependent var 6188.329Adjusted R-squared 0.826262 S.D. dependent var 2133.692S.E. of regression 889.3633 Akaike info criterion 16.56990Sum squared resid 7909670. Schwarz criterion 16.65072Log likelihood -97.41940 Hannan-Qu
36、inn criter. 16.53998F-statistic 53.31370 Durbin-Watson stat 2.339767Prob(F-statistic) 0.000026得e 2i2=7909670.3）根据 Goldfeld-Quanadt 检验，F 统计量为：F=e 2i2 /e 1i2 =7909670./ 1772245=4.4631在 =0.05 水平下，分子分母的自由度均为 10，查分布表得临界值 F0.05（10,10）=2.98，因为 F=4.4631 F0.05（10,10）=2.98 ，所以拒绝原假设，此检验表明模型存在异方差。（3）1）采用 WLS 法估
37、计过程中，用权数 w1=1/X,建立回归得：Dependent Variable: YMethod: Least SquaresDate: 12/09/14 Time: 11:13Sample: 1 31Included observations: 31Weighting series: W1Variable Coefficient Std. Error t-Statistic Prob. X 1.425859 0.119104 11.97157 0.0000C -334.8131 344.3523 -0.972298 0.3389Weighted StatisticsR-squared 0.
38、831707 Mean dependent var 3946.082Adjusted R-squared 0.825904 S.D. dependent var 536.1907S.E. of regression 536.6796 Akaike info criterion 15.47102Sum squared resid 8352726. Schwarz criterion 15.56354Log likelihood -237.8008 Hannan-Quinn criter. 15.50118F-statistic 143.3184 Durbin-Watson stat 1.369081Prob(F-statistic) 0.000000
