Freshman class

Testing the tests

Henry Braun, Boisi professor of education and public policy

Ph.D. Stanford University

Specialization: Testing, education policy, statistical modeling

Representative publication: "Reconsidering the Impact of High-Stakes Testing," Education Policy Analysis Archives

I have always had an interest in using mathematical models and data analysis to help understand some aspect of our world. Being a statistician has allowed me to collaborate with a variety of researchers in different fields.

Over the last five years a good deal of my attention has focused on how test data are used—or misused—as instruments of education policy. In particular, I have studied how results of state testing programs are being used to make evaluative judgments about schools and teachers. Although it seems intuitively reasonable that students attending better schools and with better teachers would demonstrate greater gains on tests, a closer look reveals many complexities, methodological, political, and even philosophical. Right now I am trying to develop guidelines that policy-makers can use as they make decisions about how to hold educators responsible for the learning trajectories of their students. With the reauthorization of No-Child-Left-Behind legislation looming, there is a real sense of urgency about providing support that is statistically correct and understandable to non–technical readers.