The aim of the methodological work is to broaden the scope of Bayesian modelling, so as to provided a more useful and realistic treatment for applied problems. We can categorize this work in three main groups: 1. Inference robustness: it consists in assessing the influence of modelling assumptions on the actual inference.
2. Modelling and inference through non-standard distributions: I introduce new classes of distributions that can be more adequate for modelling certain data sets (displaying e. g. skewness and/or fat tails). 3. Reference priors: it deals with finding a prior distribution which adequately captures a lack of prior information. I would like to find this reference prior for stochastic frontier models.
In terms of econometric applications, I would like to investigate theoretical and practical aspects of stochastic frontier models, as well as to treat financial data using nonstandard distributions as mentioned above. In environmental science, I am trying to develop a statistical model for estimating fish captures.