Question 3 [2+1+1+2+2+1+2+1+1=13] (Chapter 5, Section 5.2) The Question 3 data (attached) show the numerical values of a predictor y and three covariates 1, T2 and 23. Assume that a valid linear regres- sion Ely|z1, 12, 23] = Bo+ Bill + B212 + Bal3 is in force. 3.1 Write down the least squares estimates of the four regression coefficients. Bo = ; B1 = ; B2 = ; Ba = 3.2 Test the null hypothesis Ho : B1 = $2, 83 = 0 at the 10% level of sig- nificance by writing your answers to (a) through (f) below. Follow the same METHOD as the Partial F-test, but make allowance for the fact that the null hypothesis is compound, not simple. This is equivalent to, but simpler to implement, than Model Reduction Method 2 on slide 45 of the Chapter 5 slides and 46:54 min. into Lecture 14. (a) RSS full = (b) dffull = (c) RSSmull = (d) dfull = (e) Write an expression for your test statistic in terms of the symbols in parts (a) through (d). (f) The numerical value of the test statistic is (g) State the critical value for the hypothesis test OR state the p-value associated with (f). (h) State your decision regarding the validity of Ho and justify it with ref- erence to part (g)