Date:         Wed, 2 Mar 2011 15:09:08 -0500
Reply-To:     Jonas Bilenas <jonas.bilenas@CHASE.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Jonas Bilenas <jonas.bilenas@CHASE.COM>
Subject:      Re: Survey logistic warnings
Comments: To: Kelly Thevenet-Morrison
          <Kelly_Thevenet-morrison@URMC.ROCHESTER.EDU>

Have you run a collinearity diagnostic on your information matrix?
Preferably using PROC REG with the COLLIN option in the MODEL statement.

Jonas Bilenas

On Wed, 2 Mar 2011 14:22:52 -0500, Thevenet-Morrison, Kelly <Kelly_Thevenet-
morrison@URMC.ROCHESTER.EDU> wrote:

>Hello,
>I am running a proc survey logistic on an NHANES sample. I run similar
models on the sample but I change the classification of my primary exposure
variable.
>
> I increased the max iteration from 25-50 and had the majority of the
models converge with no errors or warnings.
>
>However, a few of the models gave the following warnings:
>
>WARNING: Ridging has failed to improve the loglikelihood. You may want to
use a different ridging technique (RIDGING= option),
>         or switch to using linesearch to reduce the step size
(RIDGING=NONE), or specify a new set of initial estimates
>         (INEST= option).
>WARNING: The SURVEYLOGISTIC procedure continues in spite of the above
warning. Results shown are based on the last maximum
>         likelihood iteration. Validity of the model fit is questionable.
>
>I changed the ridging to absolute in some of the models and that seemed to
take care of the error but not in all of them.
>
>I also get the warning
>
>WARNING: The information matrix is singular and thus the convergence is
questionable.
>
>I change the convergence to the different options
>
>Convergence Criteria
>Four convergence criteria are allowed: ABSFCONV=, FCONV=, GCONV=, and
XCONV=.
>
>But I continue to get the information matrix warning.
>
>I also checked the SAS-L archives but I only saw the option about changing
the ridging.
>
>Any help with these warnings would be greatly appreciated.
>
>Thanks,
>Kelly
>
>
>Kelly Thevenet-Morrison MS
>Lead Programmer Analyst
>Department of Community and Preventive Medicine
>University of Rochester School of Medicine and Dentistry
>601 Elmwood Ave., box 644
>Rochester, NY 14642
>Phone: 585-275-1817
>e-mail: kelly_thevenet-morrison@urmc.rochester.edu