Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "O'Ceilleachair, Alan" <A.OCeilleachair@ucc.ie> |
To | "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |
Subject | st: Three-way interaction with continouous dependent variable and categorical explanatory variables STATA 10 |
Date | Wed, 24 Apr 2013 08:55:14 +0000 |
Hi all, I want to investigate a three-way interaction using STATA 10 (all we have a= vailable here, unfortunately) My model is comprised of the log of drug prescribing cost as the dependent = variable and three explanatory categorical variables; age category (10 cate= gories), gender (two categories) and region of residence (eight categories)= . Particularly, I wish to investigate the significance of the individual inte= raction levels, so the variables on their own, the three two-way interactio= ns and the three way interaction. Previously I used desmat and destest to do this but I am informed that this= is incorrect, namely that the destest function does not test for the right= specification of the model and that the F-test is subsequently wrong. How do I best do this in STATA 10? Any help you could give would be greatly appreciated Alan Alan Ó Céilleachair Centre for Policy Studies University College Cork Phone: 021 4902636 * * 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/