2. The following presents results of regression of test score (TestScore, measured in points) on the Student-Teacher Ratio (STR) and student characteristic control variables using the dataset of California elementary school districts: HiEL = 1 if the percentage of English learners is larger than 10% and HiEl = 0 otherwise Lunch = % Eligible for subsidized lunch Income = annual per capita income in district, measured in thousands of dollars. You had estimated the following regression model: TestScore = Bo+ 31 STR+ 82 STR" + B; HiEL + BA Lunch + B; In(Income) + u TestScore = 252 + 64.15 STR - 3.1 STR' - 5.47 HiEL - 0.42 Lunch + 11.75 In(Income) (163.6) (26.39) (1.24) (1.03) (0.029) (1.78) (a) (6 points) What is the predicted TestScore for a school district with STR = 20, HiEL = 0, Lunch = 1, and In(Income) = 2.7? (b) (8 points) If STR decreases from 20 to 18, then what is the predicted effect on Test Score