Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Steve Samuels <sjsamuels@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: weighted logistic regression in stata |
| Date | Wed, 20 Jun 2012 09:12:34 -0400 |
You are describing what used to be called the "errors in variables" problem; now it is known as the "measurement error" problem. The kind of weighting you describe does not solve this problem, as far as I know. (I'd be happy to learn otherwise.) It is ordinarily used to accommodate outcomes, not covariates, that are measured with unequal precision. In this application, the weights are known as "analytic" weights (in Stata, "aweights"; see the -help- for "weights") For Stata solutions to the problem of measurement error in covariates. See http://stata.com/merror/ . Steve sjsamuels@gmail.com On Jun 19, 2012, at 7:32 PM, Katherine Nishimura wrote: I have an epidemiologic dataset and want to do a weighted logistic regression. I have a continuous exposure variable that I want to see if it is associated with a binary outcome. The accuracy of the exposure variable is highly variable and I want to do a logistic regression with assigns greater weight to exposure values that are highly accurate, and lower weight to values that have lower accuracy. Can stata do this type of analysis? What command should I use? Thanks, Katie * * 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/ * * 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/