Outreg2 with mixed-effect model with multiple columns

[Wei Chen]

mixed pupil_stimulus mean_val rating_diff trial ///
|| subject: $randvars 
outreg2 using  reg_mean_diff_rating_pupil , tex replace pvalue 

mixed pupil_choice   mean_val rating_diff trial ///
|| subject: $randvars 
outreg2 using  reg_mean_diff_rating_pupil , tex append pvalue

Hi,

I found that when using outreg2, I can't seems to eliminate those variance (sigma) estimation results, and make a clean table.
The above code is my original code, and the result of it is a table with 12 equations, and only equation (1) and (6) are the model estimation results that I care about.
The other 10 equations are all estimation of several different sigma.

I tried using the "onecol" option. This sort of works (I still have to manually delete stuff, but it is easy) when the dependent variable is the same across these two equations.
In the above case, the table will still be very messy, because outreg2 put the second equation below the first one.
In other words, the independent variables are not aligned.

Lastly, I tried "eqkeep(pupil_stimulus )" to try to keep just the first equation, but somehow the estimation of those sigma will still sit below the constant term.
Moreover, when I append the second equation, I got so many of those estimation of those sigma, and they don't even make sense now, because the order are all mixed.

I am wondering if there is a way that I can easily get what I want, like a linear regression table's output?

Thanks.

[Oded Mcdossi]

The variance (sigma) are the random effect part of your model so you should think about including them in your results. Anyway outreg2 (SSC) is not the best tool to deal with exporting results from mixed effect models. Try esttab (SSC) as it much more flexible tool.