Hello,
I'm trying to use the esttab command to create a table with 3 columns: the 1st reporting (baseline) mean values of 4 variables for treated individuals in my study, the 2nd reporting the mean values of the same 4 variables for control individuals, and the 3rd reporting p-values from several regressions in which the variables listed in the first two columns are regressed on the treatment variable with controls (!). I'm having difficulty in combining the p-values from several regressions and have esttab recognise that the LHS variables are the same as the ones listed in the summary tables. Is there a way to do this?
I'll paste below the method that I attempted, but this doesn't seem to work. (Stata says that 'estimation result r(coefs, is not found'.)
Thank you very much!
I'm trying to use the esttab command to create a table with 3 columns: the 1st reporting (baseline) mean values of 4 variables for treated individuals in my study, the 2nd reporting the mean values of the same 4 variables for control individuals, and the 3rd reporting p-values from several regressions in which the variables listed in the first two columns are regressed on the treatment variable with controls (!). I'm having difficulty in combining the p-values from several regressions and have esttab recognise that the LHS variables are the same as the ones listed in the summary tables. Is there a way to do this?
I'll paste below the method that I attempted, but this doesn't seem to work. (Stata says that 'estimation result r(coefs, is not found'.)
Thank you very much!
Code:
eststo clear
local time " igaNew school tuse1_iga tuse1_educ "
foreach v of local time {
eststo est_`v': qui: areg `v'B treat2 , a(s0a_district) cl(s0a_community)
local list="`list' `v'B"
}
eststo treatment: estpost summarize `list' if treat2==1
eststo control: estpost summarize `list' if treat2==0
esttab est_*, keep(treat2)
esttab treatment control r(coefs, transpose) using "$path/0.tex", replace ///
prehead("\def\sym#1{\ifmmode^{#1}\else\(^{#1}\)\fi} \setlength{\tabcolsep}{4pt} \begin{tabularx}{\textwidth}{p{0.33\linewidth} *{3}{>{\centering\arraybackslash}X} } \multicolumn{4}{l}{\textbf{Balance}} \\[2ex] \hline\hline") ///
posthead("\hline\hline \multicolumn{4}{l}{\textbf{Panel A}} \\[0.4ex] \hline") ///
mtitle("Treated" "Control" "Difference (p-value)") nonote collabels( ,none) ///
cell( ///
mean(pattern(1 1 0) fmt(2)) & p(pattern(0 0 1) fmt(2)) ///
) ///
fragment ///
varlabels(igaNewB "=1 if started new IGA" schoolB "=1 if attended school" tuse1_igaB "Time spent with income generating activities" tuse1_educB "Time spent with education") ///
postfoot("\hline\hline \end{tabularx}") noobs
