Thanks, Maarten, for sharing your ideas with me. the solutions you proposed
sound quite reasonable. and, i agree with you that OLS for categorical
variables is always heteroscedastistic.
Wishing the very best to you,
hyojoung
Hyojoung Kim
Assistant Professor of Sociology
University of Washington
202 Savery Hall, Box 353340
Seattle, WA 98195-3340
(Phone) 206-543-9644
(Fax) 206-543-2516
(E-mail) [email protected]
----- Original Message -----
From: "maartenbuis" <[email protected]>
To: <[email protected]>
Sent: Thursday, December 02, 2004 5:29 AM
Subject: Re: st: heteroscedasticity test for logistic regression
> Dear hyojoung,
>
> I just remembered that in a linear probability model the errors are
> always hetroscedastic. So doing a hettest after an OLS regression
> should always give you evidence that there is heteroscadasticity,
> regardless of whether or not that is also the case in logistic
> regression. See for instance (Long 1997, p. 38-39).
>
> J. Scott Long, 1997, Regression Models for Categorical and Limited
> Dependent Variables. Sage. (This is not the Long and Freese book wich
> is often referd to on this list, since the Long and Freese book does
> not discuss the linear probability model)
>
> Maarten
>
>
>
> *
> * 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/
>
*
* 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/