topic Re: proc genmod estimate statement in Statistical Procedures

proc genmod estimate statement

docfermi — Thu, 02 Aug 2018 12:50:51 GMT

proc genmod data = eyestudy;
 class carrot id;
 model lenses = carrot/ dist = poisson link = log;
 repeated subject = id/ type = unstr;
 estimate 'Beta' carrot 1 -1/ exp;
run;

above was copied from below:

https://stats.idre.ucla.edu/sas/faq/how-can-i-estimate-relative-risk-in-sas-using-proc-genmod-for-co...

Above statement allows exp(beta) and exp(confidence intervals).

But I don't understand, the part of estimate statement, 1 -1. Is it because carrot is binary variable? If the variable has more than 2 categories or continuous, how estimate statement will be different. I would very much appreciate your help!

proc genmod estimate statement

docfermi — Thu, 02 Aug 2018 12:48:39 GMT

proc genmod data = eyestudy;
 class carrot id;
 model lenses = carrot/ dist = poisson link = log;
 repeated subject = id/ type = unstr;
 estimate 'Beta' carrot 1 -1/ exp;
run;

above was copied from below:

https://stats.idre.ucla.edu/sas/faq/how-can-i-estimate-relative-risk-in-sas-using-proc-genmod-for-common-outcomes-in-cohort-studies/

Above statement allows exp(beta) and exp(confidence intervals).

Sun

Re: proc genmod estimate statement

SAS_Rob — Thu, 02 Aug 2018 13:28:44 GMT

Yes, it is because it has two levels. The number of coefficients will correspond with the number of levels and the ordering of the levels.

There are a number of good examples and explanations of estimate/contrast statements found here:

http://support.sas.com/kb/24/447.html