If is fixed to 0. You check that E(yi | ) = var(yi | ) = 0 What is the distribution of yi | then? If, for example, we choose the prior p() 1/2, the posterior distribution is p( |) = gamma( | X yi + 1/2, n0) and therefore E( |) = P yi + 1/2 n0 and var( |) = P yi + 1/2 [n0]2 In the example n = 5. What can you say about the influence of 0 in the posterior inference