WAM = β0 + β1Attend + u Where: WAMis a student's weight average mark Attendis the proportion of lectures attended by a student in an academic year
Consider a single variable model to estimate the effect of in-person lecture attendance on university students' WAMs (weighted average marks):
(C1) WAM = β0 + β1Attend + u
Where:
- WAMis a student's weight average mark
- Attendis the proportion of lectures attended by a student in an academic year
Using the information above, answer the following 3 questions.
[i] Explain why Attend might be endogenous in Model (C1). What does this suggest about E [u ∣ Attend]?
[ii] Suppose that student attendance is related to a student's conscientiousness. Does this mean that conscientiousness would be a good instrumental variable for Attend? Why or why not? Explain your reasoning.
[iii] Unfortunately a major strike by railway works disrupts transport links to 3 of the 7 universities in a city for a period of 2 weeks. This temporarily means that students who are enrolled at those 3 affected universities cannot attend lectures in person. Carefully explain how you would use this information to construct an instrumental variable for Attend.
Step by step
Solved in 3 steps