Latent transition analysis: inference and estimation.

by: H Chung, ST Lanza, E Loken

Statistics in medicine, Vol. 27, No. 11. (20 May 2008), pp. 1834-1854.

View FullText article

DOI, Pubmed, Hubmed

Reviews [Write a review of this article]

There are no reviews of this article

Find related articles from these CiteULike users

jyuh

Find related articles with these CiteULike tags

latent

Abstract

Parameters for latent transition analysis (LTA) are easily estimated by maximum likelihood (ML) or Bayesian method via Markov chain Monte Carlo (MCMC). However, unusual features in the likelihood can cause difficulties in ML and Bayesian inference and estimation, especially with small samples. In this study we explore several problems in drawing inference for LTA in the context of a simulation study and a substance use example. We argue that when conventional ML and Bayesian estimates behave erratically, problems often may be alleviated with a small amount of prior input for LTA with small samples. This paper proposes a dynamic data-dependent prior for LTA with small samples and compares the performance of the estimation methods with the proposed prior in drawing inference.

@article{citeulike:2897547, abstract = {Parameters for latent transition analysis (LTA) are easily estimated by maximum likelihood (ML) or Bayesian method via Markov chain Monte Carlo (MCMC). However, unusual features in the likelihood can cause difficulties in ML and Bayesian inference and estimation, especially with small samples. In this study we explore several problems in drawing inference for LTA in the context of a simulation study and a substance use example. We argue that when conventional ML and Bayesian estimates behave erratically, problems often may be alleviated with a small amount of prior input for LTA with small samples. This paper proposes a dynamic data-dependent prior for LTA with small samples and compares the performance of the estimation methods with the proposed prior in drawing inference.}, address = {Department of Epidemiology, Michigan State University, East Lansing, MI 48824, U.S.A. hchung@epi.msu.edu}, author = {Chung, H. and Lanza, S. T. and Loken, E. }, citeulike-article-id = {2897547}, doi = {http://dx.doi.org/10.1002/sim.3130}, issn = {0277-6715}, journal = {Statistics in medicine}, keywords = {latent}, month = {May}, number = {11}, pages = {1834--1854}, posted-at = {2008-06-16 05:35:22}, priority = {2}, title = {Latent transition analysis: inference and estimation.}, url = {http://dx.doi.org/10.1002/sim.3130}, volume = {27}, year = {2008} }

RIS record

TY - JOUR ID - citeulike:2897547 L3 - citeulike-article-id:2897547 TI - Latent transition analysis: inference and estimation. JF - Statistics in medicine VL - 27 IS - 11 SP - 1834 EP - 1854 AD - Department of Epidemiology, Michigan State University, East Lansing, MI 48824, U.S.A. hchung@epi.msu.edu SN - 0277-6715 N2 - Parameters for latent transition analysis (LTA) are easily estimated by maximum likelihood (ML) or Bayesian method via Markov chain Monte Carlo (MCMC). However, unusual features in the likelihood can cause difficulties in ML and Bayesian inference and estimation, especially with small samples. In this study we explore several problems in drawing inference for LTA in the context of a simulation study and a substance use example. We argue that when conventional ML and Bayesian estimates behave erratically, problems often may be alleviated with a small amount of prior input for LTA with small samples. This paper proposes a dynamic data-dependent prior for LTA with small samples and compares the performance of the estimation methods with the proposed prior in drawing inference. KW - latent AU - Chung, H AU - Lanza, ST AU - Loken, E PY - 2008/05/20/ UR - http://dx.doi.org/10.1002/sim.3130 ER -

RIS BibTeX