Suppose that in general the size of a student's class is related to the student's parents' annual income. Does this mean that parental income would be a good instrumental variable for class, size? Why or why not?
Question 29
Suppose that in general the size of a student's class is related to the student's parents' annual income. Does this mean that parental income would be a good instrumental variable for class, size?
Why or why not?
Question 30
Now imagine our random sample of data is on students living in the Newcastle reeion of NSW. Unfortunately a major earthquake hits the Newcastle region early in the school year, Some Newcastle schools are forced to close for the rest of the school year to undertake repairs. Those closest to the earthquake epicentre are the most severely affected and are more likely to need to close for repairs. Some schools need to "double up" on the mumber of students in elasses to absorb those students whose schools have been closed for repairs. This temporarily increases class sizes at these schools.
Conveniently for our study, the test score data is collected at the end of the school year in which the earthquake occured.
Carefully explain in detail how you would use this information to construct an instrumental variable for class, size.
Step by step
Solved in 3 steps
