Resumen de Bayesian analysis of mached pairs with covariates

Elías Moreno Bas , Francisco de Asís Torres Ruiz , Francisco Javier Girón González-Torre

• In a recent paper (Girón et al, 2001) a Bayesian statistical analysis of data that appears when measurements are made on the same individuals before and after treatment is applied, has been developed.

  This paper extends the above analysis to the situation in which part of the variability of the data is explained by a given set of covariates. The effectiveness or not of the treatment is analyzed from a default model choice perspective. The analysis involves variable selection problems in regression models.

  Intrinsic prior distributions (Berger and Pericchi, 1996; Moreno at al, 1998) and their associated Bayes factors and posterior probabilities of the models involved are given.