Quoting "Schmidt, Michael" <mschmidt@neuro.columbia.edu>:
> Is there a function to get the odds ratios for each coefficient when
> using glm with family=binomial?
>
> Thanks.
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> J. Michael Schmidt, Ph.D.
It is usually not a good idea to associate an odds ratio with a coefficient as a
predictor is often represented by multiple coefficients (unless you make the
unreasonable assumption of linearity in the log odds). You can easily get odds
ratios in the general case using the Design library summary (actually
summary.Design) function and the even more general contrast.Design function.
>
> Project Manager
> Columbia University Subarachnoid Hemorrhage Outcomes Project
> 177 Fort Washington, Milstein Hospital 8 Garden South, Room 306
> W: 212-305-0943
> F: 212-305-2792
> mschmidt@neuro.columbia.edu
>
> -----Original Message-----
> From: Harrell, Frank E [mailto:f.harrell@Vanderbilt.Edu]
> Sent: Wednesday, July 21, 2004 9:19 PM
> To: r-help@stat.math.ethz.ch
> Cc: S-news@lists.biostat.wustl.edu
> Subject: [S] Regression Modeling Strategies Short Course
>
> I will be giving a one-day short course related to my book Regression
> Modeling
> Strategies in Toronto as part of the Joint Statistical Meetings on
> August 8.
> For more information visit the American Statistical Association web site
>
> amstat.org and biostat.mc.vanderbilt.edu/rms. The course applies to
> both R and
> S-Plus and assumes a good general background in multiple regression.
>
> -----------------------------------------------------------------
> Harrell, Frank E
> Vanderbilt University
>
> --------------------------------------------------------------------
> This message was distributed by s-news@lists.biostat.wustl.edu. To
> unsubscribe send e-mail to s-news-request@lists.biostat.wustl.edu with
> the BODY of the message: unsubscribe s-news
> --------------------------------------------------------------------
> This message was distributed by s-news@lists.biostat.wustl.edu. To
> unsubscribe send e-mail to s-news-request@lists.biostat.wustl.edu with
> the BODY of the message: unsubscribe s-news
>
-----------------------------------------------------------------
Harrell, Frank E
Vanderbilt University
|