Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Standard deviations from OLS fixed effects regressions

From   Chris <>
To   statalist <>
Subject   st: Standard deviations from OLS fixed effects regressions
Date   Thu, 6 Jun 2013 11:42:12 +0200

Dear Statalist,

I want to calculate summary statistics for two groups after
controlling for heterogeneity via fixed effects. Groups are denoted by
dummy D=1 and D=0. To do this, I run

xtreg depvar D, fe

Now the mean (after controlling for fixed effects) is just _cons (for
D=0) and _cons+_b[D] (for D=1). Clearly, standard errors are reported,
but I have no idea how to get the SD for each of the two group means.
(Just multiplying the SE of the constant with the number of
observations from the D=0 group doesn't give the right answer; I
checked without FE, i.e., reg). Using predict, one can obtain the mean
by averaging predictions over all observations of the group, but I can
hardly use

predict se, stdp

and average standard errors of individual predictions. But how else am
I going to get those SDs? Any help is very much appreciated.

(I posted this question also here with a more detailled description,
but I believe they are mostly using R:
*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index