Dependent Variable: LOG(Y)    
Method: Least Squares    
Date: 03/31/09   Time: 16:46    
Sample: 1978 2008    
Included observations: 31    
    
Variable Coefficient Std. Error t-Statistic Prob. 
    
LOG(Y1) 0.208501 0.059399 3.510146 0.0016
LOG(Y2) 0.128395 0.089743 1.430699 0.1640
LOG(Y3) 0.618978 0.085272 7.258833 0.0000
C 1.424104 0.127462 11.17275 0.0000
    
R-squared 0.999157     Mean dependent var  5.152778
Adjusted R-squared 0.999064     S.D. dependent var  1.504901
S.E. of regression 0.046044     Akaike info criterion  -3.198525
Sum squared resid 0.057241     Schwarz criterion  -3.013495
Log likelihood 53.57714     F-statistic  10673.44
Durbin-Watson stat 0.382212     Prob(F-statistic)  0.000000
    
 帮忙分析一下阿，DW是不是太小了阿

logy1 log2,log3系数为正，说明他们与因变量之间的关系为同向关系，logy1,logy3的系数在1%的水平显著，logy2的系数不显著，R-squared0.9以上显示方程的拟合程度很好，Prob(F-statistic)  0.000000显示整个方程的系数显著，Durbin-Watson stat 0.382212显示可能出现自相关问题！  

Dependent Variable: LOG(Y)        

Method: Least Squares         

Date: 04/01/09   Time: 19:25            

Sample (adjusted): 1981 2008             

Included observations: 28 after adjustments 

Convergence achieved after 74 iterations     

                           

                           

Variable   Coefficient      Std. Error       t-Statistic Prob. 

                           

                           

LOG(Y1) 0.381591 0.033475 11.39938 0.0000

LOG(Y2) 0.263109 0.033433 7.869720 0.0000

LOG(Y3) 0.278914 0.032578 8.561466 0.0000

C     137.1963 64991.51 0.002111 0.9983

AR(1)     1.432084 0.229260 6.246542 0.0000

AR(2)     -0.175256       0.413342 -0.423997       0.6759

AR(3)     -0.256878       0.269402 -0.953512       0.3512

                           

                           

R-squared       0.999933     Mean dependent var       5.396471

Adjusted R-squared       0.999914     S.D. dependent var 1.370737

S.E. of regression   0.012712     Akaike info criterion      -5.680181

Sum squared resid  0.003394     Schwarz criterion   -5.347130

Log likelihood 86.52253     F-statistic       52317.39

Durbin-Watson stat 1.844143     Prob(F-statistic)     0.000000

                           

                           

Inverted AR Roots        1.00              .77             -.33

 

[此贴子已经被作者于2009-4-1 19:40:21编辑过]

看样还是得再加紧学习呀

还是不怎么懂！

本人认为，你加入了AR选项，但Ar（2)及Ar（3)选项不显著，若变量为内生变量需要考虑寻找合适的外生变量使用两阶段最小二乘法。

残差不平稳说明什么？

没有啊

残差性检验包括平稳性检验？

学了一学期的全忘了

哇，大神好厉害