ECON 151 Lecture Notes - Lecture 43: Quarterly Journal Of Economics, Maimonides
Document Summary
Random assignment of students in small size classes or regular size classes. Treatment lasts from k to g3 then back to regular class size. Then tracked until college: sat and act results were collected. Hard to establish purely causal relationship between class size and performance. Teachers love small classes so likely that best teachers ask for the smallest class sizes (older teachers more likely to get smaller classes) So hard to convince schools to do this so experiment only done only once schools have doubts. Percentage points = very big effect= 30% of sd. Results tend to reduce (fade-out of program) after grade 3 (end of intervention) Black students and free lunch students benefit more than full sample which means those students are particularly more affected by program. Increase probably to go to college if you have smaller primary school class sizes (long lasting effects on performance)