Problem 1 (Wooldridge B.10): Suppose that at a large university, college grade point averages, πΊππ΄, and SAT score, ππ΄π, are related by E[πΊππ΄|ππ΄π] = 0.70 + 0.002ππ΄π. (i) (ii) (iii) Find the expected πΊππ΄ when ππ΄π = 800. Find E[πΊππ΄|ππ΄π = 1400]. Comment on the difference E[GPA|SAT] = 0.70 + 0.002(800) = 2.3 E[GPA|SAT] = 0.70 + 0.002(1400) = 3.5 Comment: The expected GPA of a greater SAT score is much better compared to the expected GPA of a lower SAT score. If the average ππ΄π in the university is 1100, what is the average πΊππ΄? E[GPA|SAT] = 0.70 + 0.002(1100) = 2.9 If a student’s SAT score is 1100, does this mean he or she will have the GPA found in (ii)? Explain No. he or she will get a GPA above or below 2.9 Problem 2: Let {(π₯π , π¦π) : π = 1, 2, ..., π} be a sample. πΏππ‘ π₯ = π ππ=1 π₯π (i) πππ π¦ = π 1 ∑ππ=1 π¦π Show that π π ∑ π₯π − ∑ π₯ π π=1 ∑ π₯π − ππ₯ π=1 π π ∑ π₯π − ∑ π₯π ππ=1 (ii) Proven. Using the result in part (i), show that π π ∑(π₯π − π₯ )2 = ∑ π₯(π₯π − π₯) π=1 π=1 πΏπ»π 2 π ∑(π₯π − π₯)(π₯π − π₯) π=1 π π ∑ π₯π2 − 2π₯ ∑ π₯π − ππ₯2 π=0 π=0 ππ₯2 ππ₯2 π π ∑ π₯π2 − ∑ π₯π π₯ π=0 π=0 π ∑ π₯π(π₯π − π₯) π=0 (iii) Show that π π π ∑(π₯π − π₯)(π¦π − π¦) = ∑ π₯π(π¦π − π¦) = ∑ π¦π(π₯π − π₯) π=0 π=0 LHS π ∑(π₯π − π₯)(π¦π − π¦) π=0 π ∑(π₯ππ¦π − π₯ππ¦ − π₯π¦π − π¦π + π₯π¦) π=0 π π=0 ∑[(π₯π(π¦π − π¦) − π₯(π¦π − π¦)] π=0 π ∑ π₯π(π¦π − π¦) π=0 Problem 3: Let π and π be random variables. Hint: this problem is very similar to the previous one. (i) (ii) (iii) Show that E[(π − E[π])] = 0 LHS E[(X – E[X])] E[(X – X)] E[0] 0 Show that Var(π) = E[π(π − E[π])] LHS Var(π) = 0 RHS E[π(π − E[π])] E[X(X-X)] E[X(0)] E[0] 0 LHS = RHS Show that Cov(π, π) = E[π(π − E[π])] = E[(π − E[π])π] Cov(X,Y) = 0 E[π(π − E[π])] E[X(Y – Y)] E[X(0)] E[0] 0 E[(π − E[π])π] E[(X – X)Y] E[(0)Y] E[0] 0