本人使用的软件是R 和Gretl, 多次在两个软件对16年2个国家的面板数据尝试random effects regression不成功. 我的目的是想运行 random effects regression 以后进行Hausman test来看数据适合Random effects 还是Fixed effects.
在Gretl中,软件显示: insufficient degrees of freedom for regression.
R软件中显示: Error in if (sigma2$id < 0) stop(paste("the estimated variance of the", : missing value where TRUE/FALSE needed
在Gretl软件中,在Fixed effects regression 结果后,附了一个检验:
Test for differing group intercepts -
Null hypothesis: The groups have a common intercept
Test statistic: F(1, 26) = 0.34254
with p-value = P(F(1, 26) > 0.34254) = 0.563412
还没有高手过来答疑解惑,我只好自己用不同的panel data 来尝试 Random effects model. 结果是,至少要有3 个 corss-sections 的 Panel data 才能运行 Random effects model.
但是上面提到的Fixed effects 的结果:
Test for differing group intercepts -
Null hypothesis: The groups have a common intercept
Test statistic: F(1, 26) = 0.34254
with p-value = P(F(1, 26) > 0.34254) = 0.563412
还是需要高手给我解释下,这个说明什么,数据适不适合用Fixed effects model?
非常感谢!