3. In Example 4.3 (Hald data), consider the model {x1, x2, x3, x5} as a potential third model, with the models {x1, x2, x3} and {x1, x2, x5} constituting models 1 and 2. Use a trial run in order to assess its standard priors and pseudo-priors when model 1 or model 2 is selected.

Similarly set up pseudo-priors for models 1 and 2 when model 3 is selected. With equal prior probabilities what is the most likely posterior model?