Consider an OLS (ordinary least squares) regression of variable Y on variable X. Let b be the coefficient on X, let a be the intercept and let e be the zero-mean error term. That is, we assume the relationship between Y and X follows:

Y = a + bX + e

Which of the following is a mathematically correct statement in the context of the above?

Note: SD[.] means standard deviation, COV[.,.] means covariance, VAR[.] means variance, E[.] means expectation and ^ is the exponentiation operator.

Select the true statement:

Select one alternative:

• Y = a + b(X)
• E[Y] = E[a] + E[b](X) + E[e]
• X = (Y-a)/b
• VAR[Y] = a + b(VAR[X])
• SD[X] = (COV[X,Y]/b)^(1/2)