Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Lachenbruch, Peter" <Peter.Lachenbruch@oregonstate.edu> |
To | "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |
Subject | st: RE: summary statistics with mi multiple imputation |
Date | Tue, 20 Jul 2010 09:40:10 -0700 |
Alan, I assume you mean 20 imputations and not 20 data sets being imputed individually. I also assume that some of the background/demographic variables have missing values and you want to get good estimates of these. So... Within the mi estimate command, you can use means for these and if you want percentiles, you can use qreg. I have a Stata Tip coming out in vol 10 #3 on this. Tony Peter A. Lachenbruch Department of Public Health Oregon State University Corvallis, OR 97330 Phone: 541-737-3832 FAX: 541-737-4001 -----Original Message----- From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Alan Acock Sent: Monday, July 19, 2010 2:35 PM To: statalist@hsphsun2.harvard.edu Subject: st: summary statistics with mi multiple imputation When imputing 20 datasets and dong a logistic regression, I still need some descriptive statistics on background/demographic variables to describe the sample. a. Should I report the demographic means/sd's for each variable using the original dataset and N for each variable? b. Should I report the grand mean treating the 20 datasets as one big dataset? c. What is the best practice? Is there a way to get confidence intervals that around the means that take the multiple imputation into account? Perhaps I'm missing something that is quite obvious. --Alan Acock * * 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/