Measurement Matters: Perspectives on Education Policy from an Economist and School Board Member

摘要

Standardized tests generate a litany of concerns about the tests themselves, the subject material coverage, and the measures derived from them. While I have sidestepped many of the broader issues in this paper, I have emphasized that our ability to draw strong conclusions from test results is limited by measurement issues that we are unlikely to overcome soon, if ever. However, I do believe that we can use tests as a trigger for further investigation. Teachers and schools with both low average test scores and low estimated value-added are likely to be weak and, given limited resources, this group is the obvious place to target. Since tests are only very imperfect measures of the desired outcomes of education, we should avoid any automatic consequences for teachers or schools based solely on measures derived from tests. However, in the case of teachers, we know that in most cases, qualitative investigation by principals will confirm the statistical analysis. I expect that the same will prove true if qualified reviewers inspect schools.