Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Bootstrap predicted probabilities


From   "Jun Xu" <mystata@hotmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Bootstrap predicted probabilities
Date   Mon, 22 Sep 2003 11:48:31 -0500

Sorry for my long-delayed thanks to Axel Heitmueller for his help on this. I would also like to know how Stata's bootstrap command deals with this problem, or any other suggestions than Axel's.

My question again:

I am writing some bootstrap routine for getting the confidence interval of predicted probabilities in binary, ordered, mlogit, and count regression models. However, sometimes the random samples I draw do not have certain categories of the dependent variables, (for example, in binary regression, ordered, mlogit models), which means I cannot calculate the predicted probability of that value category (or, can we assume it is zero?). I'd love to know how the Stata bootstrap command deals with it or how I should deal with this problem? P.S. I have looked at the bs codes, but kind of hard to decode. Really appreciate.

Jun Xu
Department of Sociology
Indiana University at Bloomington

**********************************************************************************
Jun Xu,

I had a similar problem awhile ago with probit and heckprob
bootstraps and the question is whether one can tread the sub-
samples as independent and random in order to draw separately
from each sub-group (in your case the four groups) rather than from
the entire sample which would resolve your problem.
Because there seems not much literature out there, David Moore
and I are currently running some simulations for the simple probit
case and preliminary results suggest that over very many samples
there is not much difference between the two methods. If these
results turn out to be robust this would mean for your case that you
could draw from each of the four groups separately as long as they
represent independent and random samples.

axel

Axel Heitmueller,
Centre for Economic Reform and Transformation, CERT
School of Management
Heriot-Watt University
Edinburgh
EH14 4AS
UK
phone +44(0)131 451 3969
fax +44 (0)131 451 3296
a.heitmueller@hw.ac.uk
www.som.hw.ac.uk/somah3/
*********************************************************************************

_________________________________________________________________
Share your photos without swamping your Inbox. Get Hotmail Extra Storage today! http://join.msn.com/?PAGE=features/es

*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/




© Copyright 1996–2021 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index