Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Airey, David C" <david.airey@vanderbilt.edu> |
To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
Subject | re: st: repeated measures |
Date | Wed, 27 Feb 2013 14:16:51 +0000 |
. Sounds like you tried some models and anova complained about the bse term. I find the description a little hard to follow. What is a multiplex test? What are the outcome/dependent variables and measures? What are the predictors? What was the study design? It sounds complicated. Were the measures "resulta_1" and "resulta_2" replicates at the same time (technical replicates you would be willing to average because they represent trying to get a better measure) or were they collected at different times (you were interested in change over time in those measures)? If you used -anova- or -xtmixed- you would need to reshape your data to long format. -manova- and -sem- would allow analysis of repeated measures in wide format. -Dave > I have a dataset that includes results from a multiplex test by 2 > collection methods, and a criterion standard result for each > collection method. In wide form, by subject, it looks like this: > subject resulta_1 resultb_1 resultc_1 criterion_1 resulta_2 resultb_2 > resultc_2 criterion_2 > > In order to assess the effect of each result as predictor of the > criterion standard, I know that I have to adjust for non-independence > of the observations, but how do I do that for the outcome? I would > use repeated measures ANOVA, but the dependent variable is also > repeated. And what would be my bse term?? > > A little lost. Thank you > > -- > Harrison Alter, MD, MS, FACEP * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/