Logistic Regression on SAS, all variables I put in to test are showing up as insignif

#1
Hello,

I need help really badly.

I am using SAS to do my statistical modelling. I have several different categorical variables with the outcome variable being OBGYAsk --> which is whether or not the family doctor advises their patient about their future health.

So the independent variables I have are like Age, Marital status etc etc and they are all categorical.

I want to run a logistic regression to see whether socio demographic factors influence whether or not their doc advises them on their future health.

I have tested every single independent variable together and separately with the outcome and every single time it shows as insignificant.

I have about 203 observations. i also tried using Exact Logistic Regression, which showed the same outcome.

I tried with a different outcome, such as whether or not they got a pregnancy complication based on socio demogrpahics and the same results --> INSIGNIFICANT.

I really need help. I'm not sure whether my coding is wrong, or whether my same size is too small, or what is going on.

please help.
thanks!
 
#2
Re: Logistic Regression on SAS, all variables I put in to test are showing up as insi

I would start by saying that "advising about future health" is a very vague outcome. Maybe you could provide us some more information regarding your variables. As I see it, age is not categorical in nature, unless you have made it categorical. If you can give is a brief look at the variables and the way the data are entered, this might be a useful place to start.
 

hlsmith

Omega Contributor
#3
Re: Logistic Regression on SAS, all variables I put in to test are showing up as insi

Are patients cluster within provider? Meaning is a provider in the dataset more than once? If so that needs to be controlled for in the modeling!

A la multilevel modelling.