2. (10')Choose the one best answer, and briey explain your reasoning. Suppose the following population model meets the Gauss-Markov assumptions: pct of students passing a reading exam = le + 15'] ]n(expenditures) + ,6; ln(enrolln1ent) + g pct of students living in poverty + u, but we do not have data on the percentage of students living in poverty and plan to use the percentage of students eligible for the federally funded school lunch program as a proxy. Our estimates of ,8] and )92 will be consistent if a. pot of students living in poverty is closely related to pet of students eligible for the lunch program b. u is uncorrelated with pct of students eligible for the lunch program c. the difference between pct of students living in poverty and pet of students eligible for the lunch program is unrelated to h1(expenditures), ln(enrollment) and pct of students eligible for the lunch program d. all of the above must be true