Question

Suppose that there is a running variable (such as test scores) that has a a sharp cutoff such as 3.7 GPA). If you are above that cutoff, you receive a scholarship. You are interested in studying the effects of receiving a scholarship on some health outcomes. Given this description, which of the following is the most appropriate technique for your study?

A) Regression discontinuity (RDD)

B) Instrumental variables (IV)

C) Difference-in-differences (DiD)

Answer 1

The correct option is B- instrument variables.

Instrument variables method is used when we are regressing two variables but one of them is affected by a third variable. For example, Let’s say we had two correlated variables that we wanted to regress: X and Y. Their correlation might be described by a third variable Z, which is associated with X in some way. This third variable is associated with Y only through its relation with X.

In our case, we want to find the effects of receiving a scholarship on health outcomes. But there is a third variable, GPA, on which getting scholarship depends. The GPA variable is not directly related to health outcomes but affects it through scholarship. In this case, we should use instrument variables method.