In the balsakhi example, for a (non-randomized) simple difference method to produce a valid measure of impact, the key assumption is that
1. It does not matter whether the comparison kids are in school or out of school
2. Other than receiving the program, students are similar along observable and unobservable characteristics
3. Students learn anyway if they are in school, but not if they are out of school
4. Student test scores would not have changed over time absent the program
5. None of the above