Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Tomas Lind" <tomas.lind@ki.se> |
To | <statalist@hsphsun2.harvard.edu> |
Subject | st: clogit with multiple cases |
Date | Fri, 5 Mar 2010 11:38:46 +0100 |
Dear Statalisters, Model 1 has 52 cases in stratum 1 with exposure 11 and 3 controls with exposure 8, 9 and 10. Stratum 2 has 63 cases etc. Model 2 is similar but here I use a weight (to "repeat" each stratum "weight" number of times). Conditional logistic regression with -clogit- handles stratum with a different numbers of cases and controls. However, model 1 below does not work, bur model 2 works fine. Why? (I am using Stata 10.) ------------------- Model 1 ------------------- id death expo 1 52 11 1 0 8 1 0 9 1 0 10 2 63 14 2 0 8 2 0 4 2 0 11 etc etc (lots of stratum) clogit death expo , group(id) -------------------------- Model 2 -------------------------- id death expo weight 1 1 11 52 1 0 8 52 1 0 9 52 1 0 10 52 2 1 14 63 2 0 8 63 2 0 4 63 2 0 11 63 etc etc (lots of stratum) clogit death expo [fw=weight] , group(id) /Tomas * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/