The fact that the comparison in means estimates vary by education group is something we will call 'treatment heterogenie

[answerhappygod]

The fact that the comparison in means estimates vary by
education group is something we will call 'treatment
heterogeniety'. It need not be the case that all students benefit
from going to catholic high schools in the exact same way. Suppose
that catholic high schools are higher quality than public schools.
Maybe because they are located in more well-to-do neighborhoods and
get better funding. It could be the case that the quality
matters more for students with less well educated parents compared
to more educated parents -- maybe because more educated parents can
supplement poor quality schools by helping their children at home,
or hiring extra tutoring. Whatever the case may be, we can
test for it.

The Fact That The Comparison In Means Estimates Vary By Education Group Is Something We Will Call Treatment Heterogenie 1 (30.23 KiB) Viewed 33 times

## G The fact that the comparison in means estimates vary by education group is something we will call 'treatment heterogeniety'. It need not be the case that all students benefit from going to catholic high schools in the exact same way. Suppose that catholic high schools are higher quality than public schools. Maybe because they are located in more well-to-do neighborhoods and get better funding. It could be the case that the quality matters more for students with less well educated parents compared to more educated parents maybe because more educated parents can supplement poor quality schools by helping their children at home, or hiring extra tutoring. whatever the case may be, we can test for it. In the code chunk below, estimate a regression that tests whether students with higher ‘min_ed' parents benefit more or less than students with less minded parents. use the continuous measure mided here and base your test on an extension of the model (1). what do you find? Interpret your results. Produce a figure to visualize this result.