136-505 Philosophical Problems in Statistics

Usually admission to the postgraduate diploma or fourth-year honours, or a postgraduate coursework program.

This subject will examine the claim that statistics used in the social sciences are often seriously flawed. It will examine the history of 20th century statistics, with an emphasis on the Fisher/Neyman-Pearson debates and the attempts over the last 30 years to reform social science statistics. It will examine the claim that faulty statistics severely impede the growth of the social sciences. Finally, it will examine the claim that many outstanding philosophical questions about science can be answered by a sophisticated application of Bayes' theorem.

Written work totalling 5000 words.