You can standardize (or divide by 2 SDs - Gelman recommends that in a
paper somewhere, and in his regression book), or somehow equate the
scales. Or you can try to make a sensible comparison - e.g. in one
analysis I did, I said that smoking is equivalent to being 8 years
older, in terms of the odds of having some condition (I think it was
Dupuytrens).
However, if you standardize, you throw away the original information
about the scales, which can be a problem, because your comparison is
sensitive to the SD of the data in your sample.
J
2010/1/27 Dan <[log in to unmask]>:
> In binary logistic regression, how is this possible?
>
> Some of my predictors are categorical dummies, some are continuous, and I
> have one of a 12 point scale and one on a 54 point scale. Obviously, the
> odds ratios are not comparible due to the varying scales, so how do I assess
> the relative association of each predictor with the dependent variable?
>
> Many thanks for any help
>
> Dan
>
--
Jeremy Miles
Psychology Research Methods Wiki: www.researchmethodsinpsychology.com
|