| Date: | Wed, 9 Feb 2005 08:41:50 -0500 |
|---|
| Reply-To: | cbautista@hivresearch.org |
|---|
| Sender: | "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU> |
|---|
| From: | Christian Bautista <cbautista@hivresearch.org> |
|---|
| Subject: | Logistic regression for few outcomes |
|---|
| In-Reply-To: | <023b01c5080d$e541a9d0$4300a8c0@NOTEBOOK> |
|---|
| Content-type: | text/plain; charset=us-ascii |
|---|
Hi list,
This is not a specific question about SPSS but maybe someone in this list
can give some suggestions how to analyze this information.
I have a prevalence or cross-sectional study where the number of events is
very low ( 1%, 11/1389). In my analysis I apply logistic regression to
estimate potential risk factors associated to the disease. However, in this
case due to the proportion of the events to be very low maybe logistic
regression analysis is not very appropiate in this study. A second idea is
to use Poisson regression.
Thanks in advance for all your comments as well as suggestions.
/Christian
| 
