Consider a classical linear regression model y; = Bo+B1x1 + B2x12 + B3x13 + u. For a sample of 104 observations, regressing y on x1, x2, x3 and a constant gives 70 as the sum of squared residuals. We would like to test at the 5% significance level the null hypothesis B1+2 + 3 = 1 against the alternative 1 + 2 + 3 1 by comparing the SSR from restricted and unrestricted models. a) Write down a restricted regression model. What is your new dependent variable? What are your new regressors? b) Suppose that the model in (a) yields SSR = 80. Test the null hypothesis.