面板数据如何做sobel检验，尤其是控制变量怎么办？-经管之家官网！

各位大虾，本人最近要做sobel中介效应分析看了相关文献后，设计了三条路径进行检验，分别是Path1、path2、path3.但是3条路径中的控制变量不一样，若用下面语句
sgmediation detemine var, mv(mediate var) iv(independent var) cv(control vars)
则三条路径中的控制将会是同一组，求大神解答，如何三条路径分别能够加入控制变量。
另，大神可否再详述一下sobel检验后结果的分析过程。
不胜感激！！！
Sobel-Goodman Mediation Tests
Coef Std Err Z P>Z
Sobel 2.396e-06 .00004634 .0517 .95876956
Goodman-1 (Aroian)2.396e-06 .00009393.02551 .97965086
Goodman-2 2.396e-06 . . .
Coef Std Err Z P>Z
a coefficient = -.000693 .001227-.564845 .57218
b coefficient = -.003458 .066609-.051916 .958596
Indirect effect =2.4e-06 .000046 .051698 .95877
Direct effect = -.02 .007431-2.69138 .007116
Total effect = -.019997 .00743-2.69127 .007118
Proportion of total effect that is mediated:-.0001198
Ratio of indirect to direct effect: -.00011979
Ratio of total to direct effect: .99988021
若使用bootstrap之后 结果分析
bootstrap r(ind_eff) r(dir_eff),reps(1000):sgmediation Dependent var, iv(indenpendent var) mv(mediate var)
_bs_1:r(ind_eff)
_bs_2:r(dir_eff)
Observed Bootstrap Normal-based
Coef. Std. Err. z P>z [95% Conf. Interval]
_bs_1 .0000237 .0000777 0.30 0.760 -.0001287 .000176
_bs_2 -.0207014 .0061517 -3.37 0.001 -.0327585 -.0086443
. estat bootstrap, percentile bc
Bootstrap results Number of obs = 11620
Replications = 1000
command:sgmediation Fncskew, iv(lnREL200) mv(REM3)
_bs_1:r(ind_eff)
_bs_2:r(dir_eff)
Observed Bootstrap
Coef. Bias Std. Err.[95% Conf. Interval]
_bs_1 .00002369 4.66e-09 .00007773 -.0001246 .0002086 (P)
-.0000658 .0002995(BC)
_bs_2 -.02070141 .0002408 .00615168 -.0325487-.0081198 (P)
-.0335503-.0094498(BC)
(P) percentile confidence interval
(BC) bias-corrected confidence interval