(2) Your data have clusters and thus IID violations. This requires the use of cluster-corrected standard errors or possibly even HLM. Since most regression packages allow for cluster-corrected SEs only on one level, please show the ICC(1) values for each possible area of clustering (in particular, industry, region).
(4) As a robustness test, please include in the paper results using a Tobit regression specification. Again, this will require cluster-corrected SEs.
同意xddlovejiao1314对第二个修改意见的说明。
对第一个修改意见,没建议一定要用HLM软件做,只说“or possibly even HLM”,表明HLM可用可不用。
第一个修改意见,主要在please show the ICC(1) values for each possible area of clustering (in particular, industry, region). 不懂ICC(1)的意思,不知是不是分别显示vce(cluster industry)、vce(cluster region)等估计结果。
help xttobit看看呗。里面有的哇,如下: vce(vcetype) specifies the type of standard error reported, which includes types that are derived from asymptotic theory and that use bootstrap or jackknife methods; see [XT] vce_options.祝好运~