Hi,

I am a new STATA user and found this forum has been incredibly helpful for me. I can't seem to find answers to my specific situation so I decide to post this problem and hopefully someone can help me.

I am trying to run logit regressions with mediating variables. I have tried to use GSEM, but the logit regression won't converge. After I went through some of the posts in the forum that recommended relogit and Firthlogit, I decided to run separate regressions, and multiply the coefficient together to get the indirect effect.

Luckily, I have been able to find results using relogit. However, it is hard for me to export the results, as I am running multiple mediating variables with multiple predicting variables. I have created two varlists, one for predicting variables and one for mediating variables. My problem is when I am trying to export my results to excel, it can only append vertically. Is there any way I can append the results horizontally when I am running through my predicting variables list?


My code:

local indivar inshold top1_hold top5_hold mot_hold lt_time_holding lt_turnover_holding ind_holding ind_holding2 tra_holding //predicting variable list
local medvar risk_com duality //mediating variable list

foreach m of local medvar {

foreach x of local indivar{

relogit `m' `x' ta firmage tobin roa sperf sgrowth lev2 srv fci rdass capxass assint f500 hrfd unique ind_tobin total_inshold
eststo a: nlcom (_b[`x']), post


relogit target `m' `x' ta firmage tobin roa sperf sgrowth lev2 srv fci rdass capxass assint f500 hrfd unique ind_tobin total_inshold
eststo b: nlcom(_b[`m']) , post

relogit target `m' `x' ta firmage tobin roa sperf sgrowth lev2 srv fci rdass capxass assint f500 hrfd unique ind_tobin total_inshold
eststo c: nlcom (_b[`x']), post
esttab a b c using test_gsem.csv, starlevels(* 0.10 ** 0.05 *** 0.01) append

}
esttab a b c using test_gsem.csv, starlevels(* 0.10 ** 0.05 *** 0.01) append

}

Thank you so much.