AccessMyLibrary provides FREE access to over 30 million articles from top publications available through your library.
Create a link to this page
Copy and paste this link tag into your Web page or blog:
Gelman's article is a thought-provoking mix of opinions and creative methodology. I agree with Gelman that the disjunction of models and exploratory data analysis (EDA) in mainstream statistics is unsound. The assumption that models belong exclusively to confirmatory data analysis (CDA) is easily shown to be wrong in practice. Many uses of models, in particular model selection, are exploratory, and they usually forfeit the possibility of strict CDA. Gelman's suggestion that EDA be based on mainstream modeling practices is long overdue.
1. DOES BAYES HAVE AN EDGE?
Gelman is a Bayesian, but he gives a nod to non-Bayesian alternatives of creating reference ...