连老师您好,
最近导师要求做一篇面板DID的文章,需要用到kernel matching,但自己对这块实在不太了解,特向连老师请教!
我首先用logit模型找到propesity scores,然后用psmatch2找common support,结果如下:
-------------------------------------------------------------------------------
> ---------
Variable Sample | Treated Controls Difference S.E.
> T-stat
----------------------------+--------------------------------------------------
> ---------
lnecp Unmatched | 1.45286887 .426829267 1.0260396 .902619808
> 1.14
ATT | 1.24405947 1.005086 .238973465 .491525812
> 0.49
----------------------------+--------------------------------------------------
> ---------
Note: S.E. does not take into account that the propensity score is estimated.
psmatch2: | psmatch2: Common
Treatment | support
assignment | Off suppo On suppor | Total
-----------+----------------------+----------
Untreated | 0 509 | 509
Treated | 2 34 | 36
-----------+----------------------+----------
Total | 2 543 | 545
但我不太明白这个结果,而且如何继续在common support上跑回归,进而用kernel density weighting来做generalized DID matching estimator 呢?
由于导师一直在催,是否焦急,麻烦连老师了,非常感谢!


雷达卡




京公网安备 11010802022788号







