A set of n subjects give data on a response variable y and two regressors, x1 and x2 . The data are fitted to Model 1: y = Bo + B1X1 + B2x2 + & Model 2: y = do+ al(x1 - X1) + a2(X2 - X2) +n where Xx is the simple average of xx values from this set of data, k = 1, 2; & and n are random errors with mean zero. Discuss with mathematical arguments whether the ordinary least-squares estimator of B; under Model 1 is equal to the ordinary least-squares estimator of a; under Model 2, for i = 0, 1, 2. State the assumptions in your discussion. Note that the regressors can be either random or non- random