在eviews中,如果检验时间序列的异方差不太明显,但有修正自相关后(如下表),t 检验不太好,怎么办?可以不用OLS模型,用arch模型吗?
Dependent Variable: LNY
Method: Least Squares
Date: 08/06/10 Time: 15:21
Sample (adjusted): 1983 2008
Included observations: 26 after adjustments
Convergence achieved after 9 iterations
Variable Coefficient Std. Error t-Statistic Prob.
C -25.78310 9.894413 -2.605824 0.0169
LNX1 0.708442 0.462148 1.532934 0.1410
LNX2 2.106640 1.294927 1.626840 0.1194
LNI -1.230464 1.219296 -1.009160 0.3250
AR(1) 0.785863 0.227825 3.449417 0.0025
AR(2) -0.138208 0.234613 -0.589088 0.5624