Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

st: -mim-, -firthlogit-, and penalized profile likelihood?


From   Pavel Roshanov <roshanp@mcmaster.ca>
To   statalist@hsphsun2.harvard.edu
Subject   st: -mim-, -firthlogit-, and penalized profile likelihood?
Date   Mon, 1 Aug 2011 18:55:25 -0400

Dear Statalist,

I have recently learned how to use Stata, and although I'm a
first-time poster, I have learned much from this list. I hope that you
can help me with the following problem.

I'm trying to perform logistic regression with sparse data and a
relatively small sample size. I discovered separation in the
complete-case analysis and have tackled it using -exlogistic- and
-firthlogit-.

In my multiply-imputed data, however, I find myself unable to use
-exlogistic-. -firthlogit- worked but, if I understand this correctly,
the default parameter confidence intervals are based on Wald and may
be incorrect. The penalized profile likelihood method is recommended
but I don't know how to implement it and especially how to do this in
-mim-.

Thank you for your time and any suggestions are very much appreciated!



Pavel Roshanov
Graduate Student
Department of Clinical Epidemiology and Biostatistics
McMaster University
Hamilton, ON, Canada
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index