Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Maarten buis <maartenbuis@yahoo.co.uk> |
To | statalist@hsphsun2.harvard.edu |
Subject | bug fix seqlogit now available [was: st: seqlogit: Tree Specification] |
Date | Fri, 19 Mar 2010 16:14:34 +0000 (GMT) |
--- On Fri, 19/3/10, Dominik Becker wrote: > I would like to use the seqlogit package to model educational > transitions. <snip> > My intention was to model the decision tree in the > following way: > > tree (0 : 1 2 3 4 5 , 1 2 : 3 4 5 , 1 : 2 , 3 > : 4 : 5) > > However, Stata reports the error "No path in the tree leads > to values 5. > r(198);" Dominik Becker found a bug in the way -seqlogit- parsed complex decision trees specified in the -tree()- option. This has been fixed and thanks to Kit Baum this bug fix is now available from SSC. To install type in Stata: -ssc install seqlogit, replace-. Many thanks to both Dominik Becker and Kit Baum. -seqlogit- fits a sequential logit model. This model estimates the effect of the explanatory variables on the probabilities of passing a set of transitions. For example, given a highest achieved level of education the researcher studies the effect of socio-economic status on the probabilities of passing the different transitions between levels of education. The user can specify any decision tree as long as each level of the dependent variable (e.g. level of education) can be reached through one and only one sequence of transitions. If these levels have values (e.g. years of education) than this model also implies an effect of the explanatory variable on the highest achieved level. The program -seqlogitdecomp-, which is part of the -seqlogit- package, displays a graph that decomposes this effect on the highest achieved level into parts that measure the effect on the probabilities of passing the transitions and the importance of each transition. The -seqlogit- package also allows one to study the potential impact of unobserved heterogeneity, by enabling one to estimate the effects that would occur in various scenarios regarding unobserved variables. -- Maarten -------------------------- Maarten L. Buis Institut fuer Soziologie Universitaet Tuebingen Wilhelmstrasse 36 72074 Tuebingen Germany http://www.maartenbuis.nl -------------------------- * * 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/