Random Assignment in Schools Posted on February 15, 2006 Robert C. Pianta (bio) details some of the challenges of proposing random assignment studies with schools and offers some solutions.

Doing intervention studies with teachers or schools more generally is pretty tricky. It's something those of us involved in research in education science are increasingly being asked to do in the studies we do conduct.

In fact an awful lot of funding from the Department of Education is now being directly targeted to randomized controlled trials rather than more naturalistic kind of studies. I think researchers who do want to do work in schools are going to have to begin to deal with some of the challenges of random assignment. The partnership issues we've talked about are critical. You have to have the school completely on board, and the schools need to understand the reason for random assignment. We have found schools to be most willing to be randomly assigned and teachers to be willing to be randomly assigned if they understand it and if you're not withholding treatment down the line.

It's important to make plans for the kind of intervention that you want to do and create as potent an intervention as possible so that people will agree to be assigned into that, with a strong rationale for why you're doing it, and then also have available either a delayed trial of that so that the delay group serves as a control group for a certain period of time or a somewhat less potent resource that's available to teachers that they would find helpful but would probably not require as much work as the original, more potent intervention, so sometimes they like being assigned into that less potent group, but at the same time provides them with resources.

« Back to Article