Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Irwin T.S. Wang" <irwintswang@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: Han and Hausman (1990) Nonparametric Baseline Hazard |
Date | Thu, 9 Dec 2010 08:19:26 -0500 |
Stephen, Thank you so much for your information. I also read up the material you pointed to and they are very helpful since I have to account for competing risks later on in my analysis. All in all, thanks so much. T.S. Wang On Wed, Dec 8, 2010 at 4:55 AM, Stephen P. Jenkins <stephenj@essex.ac.uk> wrote: > > =============================== > Date: Tue, 7 Dec 2010 08:00:03 -0500 > On Dec 6, 2010, at 8:56 PM, Irwin T.S. Wang wrote: > > Dear All, > > Has anyone tried estimating the nonparametric baseline hazard > discussed in Han and Hausman (1990) in Stata? I browsed through the > archive but did not find relevant discussion. Any thoughts will be > much appreciated. > > Han, A. and J.A. Hausman. 1990. Flexible Parametric Estimation of > Duration and Competing Risk Models. Journal of Applied Econometrics > 5(1):1-28. > > Thanks, > > T.S. Wang > ================================ > > My understanding is that H & H provided a way of estimating the > discrete time proportional hazard regression model using an 'ordered > logit' type regression applied to reorganised data. They also extend > that approach to fit a model with Gamma distributed frailty. > > However you can estimate the same model in other ways. Without > frailty, the model can be estimated using -cloglog-, and the Gamma > frailty version of the model can be fitted using -pgmhaz8- (on SSC). > See the 'unobserved heterogeneity' material in the MS and Lessons at > the URL below. > > I conjecture that the H&H approach is not commonly used is because > they never explicitly provide the expression for the likelihood > contribution for right-censored spells in their article. So, work by > e.g. Meyer (Econometrica 1990) has become more commonly followed. See > the references cited in the previous paragraph. > > My discussion in the previous paragraphs refers to the single > destination case -- not competing risks. For some discussion of > competing risk models in the discrete time (grouped survival time) > case, see the materials cited in para 2 > > Stephen > ------------------------------------------------------------- > Professor Stephen P. Jenkins <stephenj@essex.ac.uk> > Institute for Social and Economic Research > University of Essex, Colchester CO4 3SQ, U.K. > Tel: +44 1206 873374. Fax: +44 1206 873151. > http://www.iser.essex.ac.uk > Survival Analysis using Stata: > http://www.iser.essex.ac.uk/survival-analysis > Downloadable papers and software: http://ideas.repec.org/e/pje7.html > > > > * > * 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/