(Consider the two-explanatory-variable linear-regression model, with variables written as vectors in mean deviation form (as in Section 10.4): y( ¼ B1x(

1 þ B2x(

2 þ

e. Let xð1Þ

and yð1Þ represent the residual vectors from the regression (i.e., orthogonal projection) of x(

1 and y(, respectively, on x(

2. Drawing the three-dimensional diagram of the subspace spanned by x(

1, x(

2; and y(, prove geometrically that the coefficient for the orthogonal projection of yð1Þ

onto xð1Þ is B1.