Let X,..., Xn be i.i.d. N(u, o2) variables where and o2 are both unknown. Consider the prior (, o) x 1/0. Show that the posterior predictive distribution of a future observation Xn+1 is a t distribution with n-1 d.f., location X and scale (1+1/n)/2s where s = = n Ei=1(Xi-Xn). n-1