True, False or Uncertain: Justify your answer.(6 Marks)14*5-20 Marks(a) Suppose the estimate of A, in the regression model Y -A+BX +Ax, +,X +e, is notsignificantly different from zero at the 5% level of significance. The 95% confidence intervalfor , will be wider than what it should be(b) Consider the following models Model LY A+AX, BX,+BX+, and Model II: Y =q+a.X. ++ Conclusion: SSE(Model )s SSE(Model I).(c) Consider the linear regression model: y =A+A +, Where =2x Conclusion: The least square estimates of A and B, are (i unbiased, () efficient and all coment(d) The average errors from the following model will always be zero: y, A +Bx, +e.(e) If the For the Linear regression model is 0.64, then the correlation coefficient between thevariables x and y is -0. 8

Question 5: True, False or Uncertain: Justify your answer. [4*5-20 Marks] (a) Suppose the estimate of , in the regression model Y, = B, + B,X2, + B,Xy, + B,X,, + ey, is not significantly different from zero at the 5% level of significance. The 95% confidence interval for B, will be wider than what it should be. (b) Consider the following models: Model I: Y, = B, + B2X2, + B,X,, + B,X,, + ey, and Model II: Y, = a, +a, X2, + +ez Conclusion: SSE(Model I) s SSE(Model II). (c) Consider the linear regression model: y, = B, + B,x,, + B,X,, +ey, where Xy, = 2x2, + ez. Conclusion: The least square estimates of B, and B, are (i) unbiased, (ii) efficient and (iii) consistent. (d) The average errors from the following model will always be zero: y, = B,X,, + Bzxz + er. (e) If the R for the linear regression model is 0.64, then the correlation coefficient between the variables x and y is -0.8