Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Tracey West <t.west@griffith.edu.au> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: RE ordered logit |
Date | Wed, 9 Jan 2013 19:11:45 +1000 |
Hi, I have a large panel of survey data and ran the following ordered logit with no problems, svy jackknife : ologit financialrisk age1 age3 age4 age5 age6 agesq degree vocational yr11 marfe sinfe sinma childunder15 selfemployed ownhome Europe Asia Othercob excellenthealth vgoodhealth fairhealth poorhealth Otherrel Atheist rhincome netwealth1 netwealth3 netwealth4 cash_rate ASX200 Unemplrate consconf CPI Bond but I need to correct it to adjust for random effects and also control for individuals in the same household. I have downloaded the gllamm package and have been looking for examples of how to use it. This is my random effects ordered logit command: gllamm financialrisk age1 age3 age4 age5 age6 agesq degree vocational yr11 marfe sinfe sinma childunder15 selfemployed ownhome Europe Asia Othercob excellenthealth vgoodhealth fairhealth poorhealth Otherrel Atheist rhincome netwealth1 netwealth3 netwealth4 cash_rate ASX200 Unemplrate consconf CPI Bond, i(hhid) family(binomial) link(ologit) adapt where 'hhid' is the household id. Stata produces the estimates but I'm not sure if I used the right syntax to control for households, and I'm not sure what the 'family(binomial)' and 'adapt' commands do. Also, do I need to worry about specifying that it's survey data with the gllamm command? Does it mean that using a jackknife for the survey sample is not viable? Thankyou very much in advance. Cheers, Tracey. * * 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/