lclogit: A Stata Module for Estimating Latent Class Conditional Logit Models via the Expectation-Maximization Algorithm
UNSW Australian School of Business Research Paper No. 2012 ECON 49
16 Pages Posted: 12 Nov 2012 Last revised: 6 Oct 2013
Date Written: November 11, 2012
This paper describes lclogit, a Stata module for estimating discrete mixture or latent class logit models via the EM algorithm. The software components can be downloaded by typing -net install http://www.stata-journal.com/software/sj13-3/st0312- at the Stata command prompt. The EM algorithm allows overcoming some of the computational difficulties that normally arise when fitting such models with many latent classes. The final version of this article is available in the Stata Journal, Volume 13 Number 3: pp. 625-639.
Keywords: lclogit, latent class model, EM algorithm, mixed logit
JEL Classification: C35, C61, C87
Suggested Citation: Suggested Citation