Reconsider the data from the injection molding experiment in Table 12.13.

(a) Fit the model y = 0 + AXA + BXB + A×BXAXB in the adjustment factors using the method of least squares.

(b) Simultaneously estimate the variance of the residuals at each level of the dispersion factor, C, and compute the weighted least squares estimates of the coecients in the model in

(a) using proc mixed as shown in Section 12.6.1.