Lab 05-Q1d

[malmufre]

Hello, I am having an issue interpreting Question 1d: Compare the mean 8th grade math GPA for students that scored above and below 60 on the 7th grade standardized exam. Is there a statistically significant difference? Note that this data all comes from the pre-treatment period. (5 + 5 points)

I have answered it as the following:

plot(data$math_7,data$gpa_8)
data$Treatment1 <- ifelse( data$math_7 < 60, 1, 0 )
mean( data$gpa_8[ data$Treatment1==1 ] )
mean( data$gpa_8[ data$Treatment1==0 ] )

I am unsure if I am approaching this correctly as the results are: [1] 1.460703 [1] 2.86946

I assumed that I would be getting a higher number for the treatment group. Thank you!

[lecy]

Sorry, ignore that.

"Note that this data all comes from the pre-treatment period."

What are expectations about which group has higher scores in the pre-treatment period?

[malmufre]

The expectation is that the group that has a higher score on the standardized math exams at the end of 7th-grade, is expected to have a higher GPA in 8th grade. That is expected in the pre-treatment period. I think that is why gpa_8 for the treatment group is lower than gpa_8 for the control group. Is that correct?

[lecy]

Treatment group is often lower in both pre and post treatment periods. They are in the treatment because their performance is below average.

This is why the RDD uses a special counter factual, otherwise the comparison of the two groups in the post-treatment period would be misleading.

The important stat is not final score of kids in the treatment (performance will still be worse than control group), it’s how much they improved.