A Markov Model for Analyzing Polytomous Outcome Data

M Ataharul Islam, Rafiqul I. Chowdhury, Karan P. Singh

Abstract


This paper highlights the estimation and test procedures for multi-state Markov models with covariate dependences in higher orders. Logistic link functions are used to analyze the transition probabilities of three or more states of a Markov model emerging from a longitudinal study. For illustration purpose the models are used for analysis of panel data on Health and Retirement Study conducted in USA during 1992-2002. The applications use self reported data on perceived emotional health at each round of the nationwide survey conducted among the elderly people. Useful and detailed results on the change in the perceived emotional health status among the elderly people are obtained.


Keywords


Markov Models; Covariate Dependence; Logistic Regression; Multiple States; Higher Order; Emotional Health.

Full Text:

PDF


DOI: http://dx.doi.org/10.18187/pjsor.v8i3.530

Refbacks

  • There are currently no refbacks.




Copyright (c)

Title

A Markov Model for Analyzing Polytomous Outcome Data

Keywords

Markov Models; Covariate Dependence; Logistic Regression; Multiple States; Higher Order; Emotional Health.

Description

This paper highlights the estimation and test procedures for multi-state Markov models with covariate dependences in higher orders. Logistic link functions are used to analyze the transition probabilities of three or more states of a Markov model emerging from a longitudinal study. For illustration purpose the models are used for analysis of panel data on Health and Retirement Study conducted in USA during 1992-2002. The applications use self reported data on perceived emotional health at each round of the nationwide survey conducted among the elderly people. Useful and detailed results on the change in the perceived emotional health status among the elderly people are obtained.


Date

2012-07-01

Identifier


Source

Pakistan Journal of Statistics and Operation Research; Vol 8. No. 3, 2012



Print ISSN: 1816-2711 | Electronic ISSN: 2220-5810