Date:         Tue, 22 Jun 2004 20:38:14 +0200
Reply-To:     Marta García-Granero
              <biostatistics@terra.es>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Marta García-Granero
              <biostatistics@terra.es>
Organization: Asesoría Bioestadística
Subject:      Re: CI for odds ratios
In-Reply-To:  <D412E9F5D2DCD411AAF200B0D079D66401761611@12-5-57-137.bnsi.net>
Content-Type: text/plain; charset=ISO-8859-1

Hi Tony,

TB> Does anyone know the algorithm SPSS uses to obtain the confidence intervals
TB> for odds ratios in logistic regression?  Specifically, how does it generate
TB> the standard errors used in obtaining the CI?

SPSS calculates asymptotic CI for OR &RR:

OR:
SE(logOR)=SQRT(1/a+1/b+1/c+1/d)
95%CI=OR*EXP(-1.96*SE(logOR));OR*EXP(1.96*SE(logOR))

RR:
SE(logRR)=SQRT(1/a-1/n1+1/c-1/n0)
95%CI=RR*EXP(-1.96*SE(logRR));RR*EXP(1.96*SE(logRR))

Being:
      Case Control
Exp+   a       b    n1
Exp-   c       d    n0

For exact CI, I recommend you WinPepi.

HTH

Marta Garcia-Granero

TB> Any help is much appreciated.

TB> Tony

TB> _______________________

TB> The Epsilon Group, LLC
TB> 1410 Sachem Place
TB> Charlottesville, VA 22901 USA
TB> 434.975.0097 x302
TB> 434.975.0477 (fax)
TB> tbaglioni@epsilongroup.com
TB> www.epsilongroup.com



--
Saludos,
 Marta                            mailto:biostatistics@terra.es