Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: hshaz with multiple spells data

From   <>
To   <>
Subject   st: hshaz with multiple spells data
Date   Mon, 19 Dec 2011 09:51:23 -0000


Date: Sun, 18 Dec 2011 11:11:50 +0100 (CET)
From: Judit VALL CASTELLO <>
Subject: st: hshaz with multiple spells data

Dear stata-users,

I am trying to estimate a multiple spells duration model with the hshaz and the pgmhaz8 commands. Is there any option that I need to use in order to tell stata that my data has multiple spells or the command and the options are the same than a single spell model? The model with single spells converges but the model with multiple spells with the same command does not.

Thanks for your help!

Judit Vall Castello
Visiting Professor
Universitat de Girona


The likelihood for -pgmhaz8- applies for the case of single spell per observation unit; the module should not be used for multiple spell data. Inter alia, see:
Ondrich, J. and Rhody, S. (1999), ?Multiple spells in the Prentice-Gloeckler-Meyer likelihood with unobserved heterogeneity?, Economics Letters 63, 139-144.

I think the same comment refers to -hshaz-, but do not have the time to check right now.

You could instead use -xtcloglog- for a proportional hazards model with Normal (Gaussian) unobserved heterogeneity. This works in the multiple spell case because the frailty is integrated out numerically, rather than analytically.

Professor Stephen P. Jenkins <>
Department of Social Policy and STICERD
London School of Economics and Political Science
Houghton Street, London WC2A 2AE, UK
Tel: +44(0)20 7955 6527
Changing Fortunes: Income Mobility and Poverty Dynamics in Britain, OUP 2011,
Survival Analysis Using Stata:
Downloadable papers and software:

Please access the attached hyperlink for an important electronic communications disclaimer:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index