Data Analysis for Impact Evaluations – Quiz 5

1. Which of the following are NOT reasons for stratifying your treatment assignment? (select all that apply)

2. If you run the command “areg…, absorb(strataID)”, what is this equivalent to doing in your regression?

Regress Hindi Year 3 scores on treatment and Hindi baseline scores, and cluster at the village-level. Now run this regression again and include village pair fixed effects.

3. Which of the following statements is true about the differences between the first regression and the second regression? (select all that apply)

4. What is the correct interpretation of the coefficient on hindi_bl in these regressions?