我知道LM_error和LM_lag检验是对ols估计的残差进行检验,但是ols估计的时候是否要包含截距项xcontant,包不包括对后面的LM检验有影响,如下:
LMsarsem_panel(results8,W,y,[xconstant x]);
LM test no spatial lag, probability = 458.4931, 0.0000
robust LM test no spatial lag, probability = 1.8848, 0.1698
LM test no spatial error, probability = 456.6086, 0.0000
robust LM test no spatial error, probability = 0.0003, 0.9857 这是包含截距项的检验
LMsarsem_panel(results,W,y,x);
LM test no spatial lag, probability = 403.5287, 0.0000
robust LM test no spatial lag, probability = 79.7228, 0.0000
LM test no spatial error, probability = 331.0173, 0.0000
robust LM test no spatial error, probability = 7.2114, 0.0072 这是不包含截距项的检验
两者不同,选哪一种啊,求好心人帮忙