Derivation of confidence intervals for the difference in proportions

[Megan Moreton]

Hello

If I use prtest to calculate the difference in proportions between groups, is it appropriate to say the confidence intervals were derived using a asymptotic normal approximation?

prtest outcome, by(group)

Is there a way in Stata to derive confidence intervals using an unconditional exact method? I am using Stata version 16.1.

Thank you for any help!

* Example generated by -dataex-. For more info, type help dataex
clear
input float(group outcome)
0 0
0 0
0 0
0 1
0 1
0 1
0 1
0 1
0 1
0 1
1 0
1 0
1 0
1 0
1 0
1 0
1 1
1 1
1 1
1 1
end

[Clyde Schechter]

With any official Stata command, you can go to the chapter for that command in the PDF manuals that come installed with Stata. On the first page of that chapter, there will be a link to Methods and Formulas that shows, in detail, how the command calculates its results. In this case, if you go there, you will see that -prtest- uses an asymptotic normal approximation.