Suppose we have two sets of data, labelled 1 and 2, and a normal linear regression model holds in each. Consider the model that results from "pooling" the two sets of data, i.e. -[ ][& where the ; are stochastic disturbances with E(e ;[X;) = 0 and V(e; (X;) = o;In;. j = 1, 2. Show that the least squares estimate of B formed by pooling data sets 1 and 2 is equivalent to a Bayesian analysis, in which the information about B in one data set is treated as prior information