Homework 1 (4 Points), due at the beginning of class on September 12, 2016. You can help each other! • Q1 (2 point) Consider a simple regression Y = E(Y |X) + e where E(Y |X) is the population regression function, and e is the error term. Please prove that E(e|X) = 0 (1) E(e) = 0 (2) E(eX) = 0 (3) cov(e, X) = 0 (4) Hint: E(e) = E[E(e|X)]; E(eX) = E[E(eX|X)] = E[XE(e|X)] • Q2 (1 point) Please use grade data posted in my webpage, and use stata to find E(exam2) and E(exam2|section). Verify that the law of iterated expectation holds (i.e., E(exam2) = E[E(exam2|section)]). Show me your log file or stata results. Hint, the stata command to find the distribution for section is tab section and the command to find the conditional mean is sort section by section: sum exam2 • Q3 (1 point) Continue Q2. Please use stata to find the best linear predictor for exam2 based on exam1 (i.e., use exam1 to predict exam2). Hint, the stata command to find variance and covariance is cor exam2 exam1, cov To learn more about the command cor, type help cor in the command window of stata. 1