mlogit output tables

[Elizabeth Unger]

Hello Stata users,

I am producing output tables after I run mlogit command

My goal is to have the estimated coefficients and standard errors in the same column, and to have parentheses around the standard errors.

I am running this command:

estout dk1_dk2_price_full using mlogitout, replace c(b(fmt(%9.3f)) se(fmt(%9.3f))) ///
s(ll chi2 N) label legend varlabel(_cons Constant) ///
unstack posthead("") prefoot("") postfoot("")

The output table is produced so that there is an extra row below the estimated coefficient where the corresponding standard error is printed. I am not using LaTex. Can someone provide some suggestions for a code?

Thanks in advance,
Elizabeth

[Oded Mcdossi]

If you want to add parentheses around the standard error use the par() option:

cells(b(fmt(3) star) se(fmt(3) par))

By the way, with esttab, the syntax is much more simple:

esttab dk1_dk2_price_full using mlogitout.csv,  replace b(3) se(3)  label scalars(ll chi2 N) nonumbers unst

[Eric Lin]

I have been using esttab to output mlogit results. I am wondering if there is a way to have the coefficients reported as relative risk ratios. If you are looking at the mlogit output for just one model in stata, you can use the option rrr for relative risk ratios (pretty much exponentiating the coefficients). But when you use esttab, you are reporting the regular multinomial logistic coefficients. Is there a way to report the relative risk ratios? If we were doing logit models, you can run the model "logistic" instead of "logit", and esttab out the rrr's.