Suppose you want to estimate the effect that number of years of schooling have on a person's earnings. Let X be the number of years of schooling a person has received and Y be this person's monthly income. The model you want to test would be: Y₁ = Po + B₁X; +ui, where u; is the error term, which could include factors like the ability of the person, ranking of the schools the person attended, income levels of the person's parents, etc. Suppose you decide to use a computerized random number generator that uses no information about the subject to select the individuals to include in your sample. Which of the following statements are true? (Check all that apply.) A. If X and Y have kurtosis values of 5 and 7 respectively, large outliers in your data are unlikely. B. If X; and u; are uncorrelated, the conditional mean of u; given X; will be zero. | C. The conditional mean of u; given X; will be zero in this case. D. This sampling scheme produces i.i.d. observations on (X;, Y;). Suppose you decide to collect data for the same set of individuals over a period of 30 years to estimate the relationship between X and Y. Which of the following statements is true in this case? ○ A. The sampling scheme will no longer produce i.i.d. observations on (X;, Y;) as opposed to the previous case. O OB. The conditional mean of u; given X; will continue to be zero as in the previous case. C. The sampling scheme will continue to produce i.i.d. observations on (X;, Y;) as in the previous case. D. The conditional mean of u; given X; will no longer be zero as opposed to the previous case.