2
$\begingroup$

What difference does it make in estimation of model equation if a variable is specified in offset option in proc logistic?

I know, if I specify a variable in offset option; the variable will be included in the model equation with coefficient as 1. If the variable was already coming as a predictor then will it make any difference to the model equation if that variable is specified in offset option.

$\endgroup$
  • 1
    $\begingroup$ Yes it will alter the coefficient. The overall model should be the same (if I understand what you mean). But why not try it and see? You will not break SAS. $\endgroup$ – mdewey May 17 '18 at 11:30
  • 1
    $\begingroup$ This is on topic. Offsets are a statistical issue, not a programming one. $\endgroup$ – Peter Flom May 18 '18 at 10:48
0
$\begingroup$

Partially answered in comments:

Yes it will alter the coefficient (1 will be subtracted). The overall model should be the same (if I understand what you mean). But why not try it and see? You will not break SAS. – mdewey

offsets is mostly used with poisson (or negative binomial) regression, to model rates. It is not often used with logistic regression, but you can find examples and discussion here: Offset in Logistic regression: what are the typical use cases?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.