计量经济学第二版 第四章答案.docVIP

4.1 (1) 存在。 因为 当之间的相关系数为零时，离差形式的 有 同理有： (2) 因为 ，，， 则 (3) 存在。 因为 当时， 同理，有 4.3(1)建立中国商品进口额为Y与国内生产总值x1、居民消费价格指数x2得回归模型 估计模型参数，结果为 Dependent Variable: LNY Method: Least Squares Date: 05/16/12 Time: 19:15 Sample: 1985 2007 Included observations: 23 Variable Coefficient Std. Error t-Statistic Prob. C -3.060149 0.337427 -9.069059 0.0000 LNX1 1.656674 0.092206 17.96703 0.0000 LNX2 -1.057053 0.214647 -4.924618 0.0001 R-squared 0.992218 Mean dependent var 9.155303 Adjusted R-squared 0.991440 S.D. dependent var 1.276500 S.E. of regression 0.118100 Akaike info criterion -1.313463 Sum squared resid 0.278952 Schwarz criterion -1.165355 Log likelihood 18.10482 F-statistic 1275.093 Durbin-Watson stat 0.745639 Prob(F-statistic) 0.000000 参数估计结果如下： （2）)数据中有多重共线性，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，且其简单相关系数呈现正向变动。 （3） Dependent Variable: LNY Method: Least Squares Date: 05/16/12 Time: 19:17 Sample: 1985 2007 Included observations: 23 Variable Coefficient Std. Error t-Statistic Prob. C -4.090667 0.384252 -10.64579 0.0000 LNX1 1.218573 0.035196 34.62222 0.0000 R-squared 0.982783 Mean dependent var 9.155303 Adjusted R-squared 0.981963 S.D. dependent var 1.276500 S.E. of regression 0.171438 Akaike info criterion -0.606254 Sum squared resid 0.617208 Schwarz criterion -0.507515 Log likelihood 8.971921 F-statistic 1198.698 Durbin-Watson stat 0.364369 Prob(F-statistic) 0.000000 Dependent Variable: LNY Method: Least Squares Date: 05/16/12 Time: 19:18 Sample: 1985 2007 Included observations: 23 Variable Coefficient Std. Error t-Statistic Prob. C -5.442420 1.253662 -4.341218 0.0003 LNX2 2.663790 0.228046 11.68091 0.0000 R-squared 0.866619 Mean dependent var 9.155303 Adjusted R-squared 0.860268 S.D. dependent var 1.276500 S.E. of regression 0.477166 Akaike info criterion 1.441037 Sum squared resid 4.781435 Schwarz criterion 1.539775 Log like