News for hackers. Stuff that matters.
Re:D.C. (School) Vouchers: Better Results at a Quarter the Cost by snewe on Saturday April 04, 2009 @03:07PM (#546556)
Read the article:

"The evaluation is based on a randomized controlled trial design that compares the outcomes of eligible applicants randomly assigned to receive (treatment group) or not receive (control group) a scholarship through a series of lotteries."

Now, given lottery status parents might help their kids with HW or otherwise (b/c they don't want to waste an opportunity). I am still reading on how they control for that, but I suspect they have a TON of observables on the parents, income, location, etc. School voucher studies these days -- particularly those read and "approved" by those in the list of acknowledgements -- will go nowhere without either randomization or a suitable instrumental variable that correlates with the voucher, but not with educational outcomes. No need to throw around econometric criticisms until you read at least part of the article.

There are no comments on this comment yet.