[prev in list] [next in list] [prev in thread] [next in thread] 

List:       sas-l
Subject:    Re: PLEASE HELP
From:       David L Cassell <davidlcassell () MSN ! COM>
Date:       2006-02-22 20:58:49
[Download RAW message or body]

fberenice@YAHOO.COM wrote back:
>Thanks a lot!!

We'll see if you still feel that way when you read what I have to say.

>Ok what I am trying to check is if the smoking status of a person can
>decline on death.
>
>What I have is two groups: Smokers and non-smokers, for each group I
>have other measurements, such as cholesterol level, sex, age, weight,
>etc.
>
>My response variable is a dichotomous variable: Death (1=yes 0=no)
>
>I want to get a model to predict the decline of the subjects into
>death, so far I did the logistic regression of my total data and I got
>a particular model for it, after that I got a paired data and I got a
>totally different model for it, but I saw that there is a big
>difference of age between the paired and the total data, so I want to
>see if there is any method I can use to separate my data by age groups
>and match without getting ride of the extreme people, that way I am
>hopping I will have a more similar paired data to the original data and
>therefore more similar results.

Yes.  The large discrepancy in AGE may be causal, or related to the study
issues, or at least related to the way the data were collected.  Which means
that you should NOT be trying to wedge your results into this sort of
paired-
comparison model, since you don't really have one.  You do not have a
case-control study, and you cannot squeeze your data into one just by
wishing really hard for the ages to match up.

And, if your data come from a sample survey instead of some manner of
observationla study, then you probably analyzed the data incorrectly to
begin with, since sample survey data ought to be analyzed properly using
the SAS SURVEYxxxx procs.  In your case, PROC SURVEYLOGISTIC .  But
I'm just guessing on that part.

HTH,
David
--
David L. Cassell
mathematical statistician
Design Pathways
3115 NW Norwood Pl.
Corvallis OR 97330

_________________________________________________________________
Is your PC infected? Get a FREE online computer virus scan from McAfeeŽ
Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

[prev in list] [next in list] [prev in thread] [next in thread]