Ioannis Ntzoufras

Publications Page

Bayesian Model Comparison for the

Order Restricted RC Association Model.

   Psychometrika, 74, 561–587, 2009.

SYNOPSIS

Association models constitute an attractive alternative to the usual log-linear models for modeling the dependence between classification variables. They impose special structure on the underlying association by assigning scores on the levels of each classification variable, which can be fixed or parametric. Under the general row-column (RC) association model, both row and column scores are unknown parameters without any restriction concerning their ordinality. However, when the classification variables are ordinal, order restrictions on the scores arise naturally. Under such restrictions, we adopt an alternative parameterization and we infer for the equality of subsequent scores using the Bayesian approach. In order to achieve that, we have constructed a reversible jump Markov chain Monte Carlo algorithm for moving across models of different dimension and estimate accurately the posterior model probabilities which can be used either for model comparison or for model averaging.  The proposed methodology is illustrated using two datasets.

Keywords:  Contingency tables, ordinal variables, Reversible jump MCMC algorithm, Equality of Odds, Bayesian model averaging.

 

 Download:

• First  version of the paper [First version: 10 July 2007]

• Transparencies from presentations at

  • Lancaster University, 26/6/2007 [pdf]

  • Athens-Pavia Meeting on Statistics, 6/6/2008 [short & long version].

• Latest version of the paper is available via Psychometrika web-site (Heal-link).

---

All Contents Copyright.
Last revised: 25/11/2011