Ann E. Harman, director of research and information for the National Board for Professional Teaching Standards, is coordinating the NBPTS's efforts to beef up scholarly research about the certification process and its effects. She recently spoke with the Harvard Education Letter about that work.

HEL: The Board has asked well-known researcher William Sanders to examine the effectiveness of its teachers using his "value-added" analysis of standardized test scores. What's the advantage of using that model?

HARMAN: The Sanders model looks at the growth in each student's performance. When you just measure average scores in a class, you are inviting teachers to focus on kids who are just below the standard because the best chance a teacher has of demonstrating her competency is to get as many kids as possible above average. Sanders is saying that a teacher who does that is not really doing her job because some kids make no progress at all. She loses sight of the fact that she's teaching not just a class but individuals in a class. You can have very high-scoring kids who learn nothing in a particular year. Teachers should be required to focus on them as much as on struggling kids. We need to use this assessment to get beyond the numbers game and enlarge the conversation about how to meet the needs of all our students. If not, we're losing sight of what's really important, which is the quality of instruction and quality of student learning. Until now, standardized tests are all we've given the public to evaluate student learning, so to get the conversation moving forward we first have to address that issue.