by Helen F. Ladd, Sanford Institute of Public Policy, Duke University
and Tim R. Sass, Department of Economics, Florida State University
In two recent papers with various co-authors (Clotfelter, Ladd and Vigdor (2007)) and (Harris and Sass (2007)), we reach somewhat different conclusions about the efficacy of the certification process established by the National Board for Professional Teaching Standards (NBPTS). In North Carolina, Clottfelter, Ladd and Vigdor find that National Board Certified Teachers (NBCTs) are more effective than the average teacher in the elementary grades prior to going through the certification process, but their advantage over other teachers does not significantly change during and after the certification process. These results suggest that the NBPTS process identifies superior teachers in North Carolina, but the process itself does not enhance teacher quality. The analysis for North Carolina is limited to the elementary grades.
"The divergent findings with respect to NBPTS certification in Florida and North Carolina are puzzling and have led us to try to determine the source of the differences. As a first step we have estimated identical econometric models for both states using the same variables and time period to ensure that the cross-state differences are not due to differences in methodology, variable selection or timing. Our preliminary results indicate that even in this 'apples-to-apples' comparison, the cross-state differences remain."
In Florida, the results are more mixed. Harris and Sass find that the ability of the NBPTS process to identify better-than-average teachers varies with the grade level, subject matter and test instrument used to evaluate student achievement. Using Florida's curriculum-based "Sunshine State Standards" exam, they find that NBPTS certification identifies teachers who are superior prior to certification in middle school math and reading, both not in the elementary or high school grades. Using the Stanford-9 exam as a measure of student performance, they find that NBCTs in Florida are more effective than the average teacher in middle and high school math prior to going through the certification process, but future NBCTs are statistically no more effective in teaching elementary math or reading at any level. For neither Florida exam is there consistent evidence that the certification process itself makes teachers more effective, and in many instances NBCTs appear to be less effective in raising test scores during and after going through the certification process compared to the period before they applied.
The divergent findings with respect to NBPTS certification in Florida and North Carolina are puzzling and have led us to try to determine the source of the differences. As a first step we have estimated identical econometric models for both states using the same variables and time period to ensure that the cross-state differences are not due to differences in methodology, variable selection or timing. Our preliminary results indicate that even in this "apples-to-apples" comparison, the cross-state differences remain. We continue to find robust evidence that NBPTS certification provides a signal of effectiveness in North Carolina, but the process itself does not enhance teacher quality. In Florida, the findings continue to be mixed. Using the same model as in North Carolina and Florida's Sunshine State Standards exam, teachers who eventually become NBCTs are found to be more effective than the average teacher in promoting achievement in elementary reading, middle school reading and middle school math. Using the Stanford-9 to measure achievement, future NBCTs outperform the average teacher in boosting student achievement in both math and reading at the elementary and middle school levels, but not in high school.
We continue to find that after going through the NBPTS certification process, Florida NBCTs are no more effective than before they applied and in some cases appear to be significantly less effective after achieving certification. We continue to explore alternative explanations for the observed differences in the relative effectiveness of NBCTs in North Carolina and Florida. The differences may simply be due to differences in the composition of the achievement exams in the two states, with the North Carolina exam being better aligned with the teaching dimensions in which NBCTs excel. Alternatively, there may be differences in the characteristics of teachers who seek NBPTS certification in the two states.
Despite the observed differences between Florida and North Carolina, we can be confident in a number of findings. First, at least for Florida and North Carolina, there is no evidence that the NBPTS process itself enhances the ability of teachers to promote student achievement. Second, for North Carolina and for at least some grades and subjects in Florida, NBPTS certification does serve to identify teachers with better-than average effectiveness. Third, even when differentials between future NBCTs and the average teacher who never becomes certified emerge, the differentials are typically modest in both states, in the range of 4-6 percent of a standard deviation in achievement. These small effects, however, are over and above the effects of other characteristics of NBCT teachers, such as their experience, that are also associated with higher student achievement.
Helen F. Ladd is Edgar T. Thompson Professor of Public Policy Studies at Duke University. Tim R. Sass is a professor of economics at Florida State University.
References
Clotfelter, Charles T., Helen F. Ladd and Jacob L. Vigdor. 2007. "How and Why Do Teacher Credentials Matter for Student Achievement?" CALDER working paper no. 2.
Harris, Douglas N. and Tim R. Sass. 2007. "The Effects of NBPTS-Certified Teachers on Student Achievement." CALDER working paper no. 4.
Read additional commentary by Daniel F. McCaffrey and Mary Dilworth.
