悬赏 10 个论坛币 未解决
面板设定为xtset name date,回归命令为:xtivreg y x1 (x2=iv),fe。现在分组为gen by_year = (year<2012),我想比较这两组回归结果的系数是否有差异。分组回归命令为:xtivreg y x1 (x2=iv) if year<2012,fe
est store a
xtivreg y x1 (x2=iv) if year>2011,fe
est store b
suest a b
然后报错a was estimated with a nonstandard vce (conventional)
用bdiff, group(by_year) m(xtivreg y x1 (x2=iv),fe) surest
然后还是报错GrouP1 was estimated with a nonstandard vce (conventional)
请问各位朋友该用哪种方式检验回归结果系数的差异性?