面板回归中xtreg的R^2与reg不一致该选用哪一个？

Miss_姚姚

2181

收藏 2022-04-06

悬赏 10 个论坛币未解决

求助一个计量问题：

在进行面板数据回归并对时间、个体进行双向控制时，两个等价的命令：
xtreg y x1 x2 ... i.year, fe r
与  reg y x1 x2 ... i.ID i.year, vce(cluster ID)
得到的R^2不同且差异较大，应该选用哪一个报告呢？下面两部分为用两个不同命令的示例，可以看到系数相同，但reg的R2与xtreg的within R2有很大差异 reg y x i.ID i.year, vce(cluster ID)

Linear regression                            Number of obs    =    1,660
                                             F(3, 414)       =       .
                                             Prob > F       =       .
                                             R-squared       =    0.9426
                                             Root MSE       =    .04819

                                 (Std. Err. adjusted for 415 clusters in ID)
------------------------------------------------------------------------------
         |             Robust
         y |    Coef. Std. Err.    t P>|t|    [95% Conf. Interval]
-------------+----------------------------------------------------------------
         x |  -.0286474 .005314 -5.39 0.000 -.0390931 -.0182017

xtreg y x i.year, fe r

Fixed-effects (within) regression             Number of obs    =    1,660
Group variable: ID                            Number of groups  =       415

R-sq:                                        Obs per group:
   within  = 0.0532                                        min =       4
   between = 0.0631                                        avg =       4.0
   overall = 0.0218                                        max =       4
——————————————————————————————————————————————————-
                                             F(4,414)       =    21.43
corr(u_i, Xb)  = -0.2376                      Prob > F       =    0.0000

                                 (Std. Err. adjusted for 415 clusters in ID)
------------------------------------------------------------------------------
         |             Robust
         y |    Coef. Std. Err.    t P>|t|    [95% Conf. Interval]
-------------+----------------------------------------------------------------
         x |  -.0286474 .0046016 -6.23 0.000 -.0376927 -.0196021