A justification for tutoring programs is that they improve student quality. Suppose that you are asked to evaluate whether more tutoring makes better quality students. However, rather than having data on individual students, you have access to data on schools in Florida. For each school, you have information on hours of tutoring per student and the number of courses passed with a B or better per student (output).

i)Carefully state the ceteris paribus thought experiment underlying this policy question.

ii)Does it seem likely that a school's decision to tutor its students will be independent of student characteristics? What are some of those measurable and unmeasurable student characteristics?

iii)Name a factor other than student characteristics that can affect the quality of a student.

iv)If you find a positive correlation between student quality and tutoring, would you have convincingly established that tutoring makes better quality students? Explain.