Date:         Wed, 25 Jan 2006 15:42:29 -0300
Reply-To:     Hector Maletta <hmaletta@fibertel.com.ar>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Hector Maletta <hmaletta@fibertel.com.ar>
Subject:      Re: logistic regression interpretations
Comments: To: Ola S Rostant <rostanto@msu.edu>
In-Reply-To:  <E1F1mOg-0002Dm-RL@sys23.mail.msu.edu>
Content-Type: text/plain; charset="us-ascii"

I assume Male=1 and Female=0 in your dependent variable (make the necessary
adjustments in this answer if the reverse was the case). The coefficient
Exp(B) equals the log odds of the event of preferring a male doctor over a
female one, per additional point of clinical competence. If the clinical
competence scale goes up by one point, the odds of preferring a male doctor
go up by 40.7%.
Hector

-----Mensaje original-----
De: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] En nombre de Ola S
Rostant
Enviado el: Wednesday, January 25, 2006 12:17 PM
Para: SPSSX-L@LISTSERV.UGA.EDU
Asunto: logistic regression interpretations

Dear list,

I'm having trouble interpreting the odds of a continuous predictor in a
logistic regression.

This is my model:
Gender preference for OB/GYN doctor(DV)= race (majority vs. minority) + educ
(hs vs. college) + age (<35 or >36)+ clinical competence (this is a
continuous variable which represents perception of physician competence).

In my output, the Exp(B) for clinical competence is 1.407. How do I
interpret this? Can anyone suggest an online resource?

All suggestions are welcomed,

Ola