之前用DID方法写了篇论文，返回的修改意见是The econometric modellingmethodology is not applied (or at least not reported) to the required extent,e.g. panel model specification tests are not carried out, nor tests forpossible violations of assumptions regarding the error terms. Such violations,if they exist, may affect standard errors and thus may have sizeable impacts onthe results, especially on the effect of interest that, despite the very largesample size, is only significant at the 5% level.
前半部看不是很懂，后半部跟 standard errors 有关，不知道你有没对你的标准差做 cluster（这是现在投稿基本要求）？请参考
Bertrand, Marianne, Esther Duflo, and Sendhil Mullainathan, 2004. “How Much Should We Trust Differences-in-Differences Estimates?” Quarterly Journal of Economics 119 (1), 249–275.
扫码加好友，拉您进群
全部版块
我的主页
收藏
