Quick search:
This is a session of Topic 3: Statistics education at the post-secondary level             Full topic list

(Monday 3rd, 10:30-12:30)

Teaching Bayesian Statistics



In the wake of an exponential increase of important Bayesian applications, the last years have seen an increasing number of suggestions to change both the teaching and the practice of statistical science and to to complement frequentist procedures, based on the average performance over repeated sampling, with Bayesian procedures which focus on optimal performance conditional on the data actually observed.

Bayesian methods are also known to guarantee consistency, and to bypass common problems of classical inference such as the elimination of nuisance parameters, the incorporation of known restrictions on the parameter values, or the appropriate modelling of hierarchical structures. The aim of this session is to explore the possibilities for including the teaching of Bayesian analysis methods in both undergraduate and postgraduate statistical training.


PaperTitlePresenter(s) / Author(s)
3I1Teaching independence and exchangeabilityLisbeth Cordani (Brazil)
Sergio Wechsler
3I2A Bayesian Mathematical Statistics PrimerJose M Bernardo (Spain)
3I3Unpredictability, probability updating and the Three Prisoners ParadoxRosangela Helena Loschi (Brazil)
Pilar Iglesias
Sergio Wechsler
3I4Standard statistical concepts: can they produce incoherence?Carlos Alberto de Bragan├ža Pereira (Brazil)