下面的是回归的结果,用的是截面加权的混合最小二乘法,不显著的系数包括CSHLISTED、REMUNERACOMM和SENIORFEMALE,三者都是虚拟变量。我的问题是:是不是加权的时候虚拟变量也除了相应的权数,所以才导致回归结果中系数不显著的?如果是的话,该怎样解决呢?求高手帮忙解答!先表示感谢!
Dependent Variable: TER?
Method: Pooled EGLS (Cross-section weights)
Date: 04/21/12 Time: 18:28
Sample: 2006 2010
Included observations: 5
Cross-sections included: 59
Total pool (unbalanced) observations: 269
Linear estimation after one-step weighting matrix
Variable Coefficient Std. Error t-Statistic Prob.
C 0.087853 0.007836 11.21083 0.0000
SHFOREIGN? 0.004370 0.001622 2.694564 0.0075
CSHLISTED? -0.000473 0.000712 -0.664252 0.5071
BS? -0.000888 0.000269 -3.299849 0.0011
ID? -0.016968 0.007419 -2.287141 0.0230
REMUNERACOMM? -0.000362 0.001429 -0.253565 0.8000
SENIORFEMALE? 6.36E-05 0.001235 0.051543 0.9589
SIZE? -0.002301 0.000303 -7.580302 0.0000
LAGPER? -6.72E-06 3.48E-06 -1.928686 0.0549
TIMECONSTR? 0.000302 0.000148 2.044609 0.0419
Weighted Statistics
R-squared 0.314221 Mean dependent var 0.033682
Adjusted R-squared 0.290391 S.D. dependent var 0.023793
S.E. of regression 0.009503 Sum squared resid 0.023390
F-statistic 13.18586 Durbin-Watson stat 1.941041
Prob(F-statistic) 0.000000
Unweighted Statistics
R-squared 0.117873 Mean dependent var 0.022318
Sum squared resid 0.024946 Durbin-Watson stat 1.785582