Contribution to the discussion of A. Zellner and A. Siow “Posterior odds ratio for selected regression hypotheses”