Please ignore-- I realize I can calculate the login transformed CIs using the formula provided in the manual.
Sorry.


I'm using -proportion- to get estimates and the logit-transformed confidence intervals, and I want to be able to save these results in a file, combine them with others and use them to make a table or graph.

I thought to gather these results using Roger Newson's -parmest- or Ben Jann's -estout- but each of these uses the standard error to calculate the 95% CI based on a t-distribution given the degrees of freedom. Here is a simple example.

. eststo clear

. sysuse auto.dta, clear
(1978 Automobile Data)

. proportion rep78

Proportion estimation Number of obs = 69

--------------------------------------------------------------
| Logit
| Proportion Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
rep78 |
1 | .0289855 .0201966 .0070794 .1110924
2 | .115942 .0385422 .058317 .2173648
3 | .4347826 .0596787 .3214848 .5553295
4 | .2608696 .0528625 .1695907 .3788629
5 | .1594203 .0440694 .0895793 .267702
--------------------------------------------------------------

. parmest, saving(myexample,replace)
file myexample.dta saved

. eststo
(est1 stored)

. use myexample.dta, clear

. list parm min95 max95, clean noobs

parm min95 max95
1.rep78 -.01131623 .06928724
2.rep78 .03903231 .19285175
3.rep78 .31569563 .55386958
4.rep78 .15538409 .36635504
5.rep78 .07148126 .24735932

. estout, cells("b ci_l ci_u" )

---------------------------------------------------
est1
b min95 max95
---------------------------------------------------
1.rep78 .0289855 -.0113162 .0692872
2.rep78 .115942 .0390323 .1928518
3.rep78 .4347826 .3156956 .5538696
4.rep78 .2608696 .1553841 .366355
5.rep78 .1594203 .0714813 .2473593
---------------------------------------------------

Is there another *simple* way to get the logit transformed 95% CI values using these commands or another command? I don't doubt that -putexcel- might be made to do this, but for my purposes it is much easier to just use -Copy table- and paste into Excel.
Thank you,
Dave Harless