For Example 11.2 (page 312) on y = mental impairment, x1 = life events, and x2 = SES, the multiple regression model has output Coef. Std. Error t Sig.

(Constant) 28.230 2.174 12.984 .000 LIFE .103 .032 3.177 .003 SES -.097 .029 -3.351 .002 and the model allowing interaction has output Coef. Std. Error t Sig

(Constant) 26.036649 3.948826 6.594 0.0001 LIFE 0.155865 0.085338 1.826 0.0761 SES -0.060493 0.062675 -0.965 0.3409 LIFE*SES -0.000866 0.001297 -0.668 0.5087 SES had a P-value of 0.011 in the bivariate model containing only that explanatory variable, and LIFE had a Pvalue of 0.018 in the bivariate model containing only that explanatory variable. Select explanatory variables from the set x1, x2, x3 = x1x2, with α = 0.05,

(a) Using backward elimination.

(b) Using forward selection.