|From||Michael Ingre <Michael.Ingre@ipm.ki.se>|
|Subject||Re: st: Re: clogit or logistic for matched pairs|
|Date||Fri, 7 Nov 2003 01:48:13 +0100|
Thank You all
gene byte baseline=1Roger Newson pointed me in the right direction. I think RR is what I want together with the risk for unexposed twins (or baseline). I noticed that RR also can be computed by -mcc- with data in the wide format (CI:s differs slightly though).
glm dv iv baseline, link(log) family(binomial) cluster(twinpairid) eform noconst