Over the past few years, the Common Core State Standards have been embroiled in a proxy war over the role of the federal government in education. To those most protective of state and local prerogatives, 鈥渃ommon鈥� became a synonym for 鈥渇ederal.鈥� Perhaps now that the Every Student Succeeds Act has settled that fight by curtailing the federal role, and the Common Core State Standards are now just the state standards, policymakers can recognize that the common standards and assessments are not antithetical to states鈥� rights after all. On the contrary: With the common standards and assessments, state leaders will be in a much better position to learn from their policy differences.
Last year, the Center for Education Policy Research at Harvard University surveyed a representative sample of approximately 1,500 teachers across five states (Delaware, Maryland, Massachusetts, Nevada, and New Mexico) to in preparation for the new assessments from the Partnership for Assessment of Readiness for College and Careers and the Smarter Balanced Assessment Consortium.
My colleagues at the center (where I鈥檓 the faculty director) and I were surprised by the magnitude of the changes they described. Eighty-two percent of math teachers and 72 percent of English teachers reported changing at least half their instructional materials to align with the new standards. Moreover, the changes they described seemed to reflect the goals of the new standards, focusing more deeply on the concepts behind arithmetic and fractions in elementary grades and emphasizing close reading and persuasive writing in English.
Four out of five math teachers reported increasing their emphasis on students鈥� conceptual understanding and the application of math skills in problem-solving. Four out of five English teachers reported that they had increased the amount of nonfiction-reading assignments, as well as the number of writing assignments in which students needed to cite evidence from texts.
We measured the student achievement in each teacher鈥檚 classroom and, after accounting for students鈥� demographics and their achievement on the previous state assessments, we investigated whether there were particular implementation strategies that were associated with greater student success.
In mathematics, we identified three factors that seemed to boost student achievement: more days of teacher professional development on the new standards; classroom observations of teachers followed by explicit feedback on alignment with the new standards; and the inclusion of common-core-aligned student outcomes in teacher evaluations.
In English, we found fewer clues to successful implementation. However, we did learn that the new tests were more sensitive to instructional differences between teachers than the previous state tests. While the magnitude of variation in teachers鈥� impacts on students鈥� math performance was largely the same as on the previous assessments, the variation in teachers鈥� impacts on students鈥� performance on the English assessments grew substantially. This was most pronounced in middle school grades, where it grew by 50 percent.
Common assessments need not imply a common policy agenda.
The previous state English assessments were essentially multiple-choice tests of reading comprehension and included few, if any, writing prompts, which are more costly to score. As students transitioned to middle school, their teachers鈥� success or failure in developing students鈥� writing ability was largely ignored.
Our analysis suggested that most of the increased variation in teachers鈥� impacts occurred on the writing portion of the new exams. By including measures of student writing, the new tests ensure that teachers have an incentive to develop students鈥� writing skills, which were too often overlooked in the No Child Left Behind era.
As our study illustrates, shared standards and assessments create shared learning opportunities. The states that remain in one of the consortia, PARCC or SBAC, are just waking up to the fact that they can pool their resources, survey a sample of teachers across multiple states, and hear from teachers about the supports they are receiving and the obstacles they may be facing. By linking individual teachers to students and their scores in state databases, and by accounting for students鈥� characteristics and prior achievement, states can also start learning about which types of teacher training, or which textbooks and educational software tools, are making the biggest difference for student achievement.
For instance, think about how an annual study of this kind could fundamentally change the textbook market. In the past, textbook publishers had an incentive to cover the maximum number of standards across as many states as possible. As a result, U.S. textbooks were a mile wide and an inch deep. Now states can share the cost of judging publishers on the quality of their results by studying the comparative achievement gains of students who use their textbooks. Moreover, by ensuring that poor results reflect poorly on a publisher鈥檚 offerings, states could strengthen publishers鈥� incentives to focus on the quality of their materials and the training they provide teachers鈥攏ot just on their marketing.
Common assessments need not imply a common policy agenda. In fact, common assessments are more valuable when states have differing policies that could lead to differing results. The PARCC and SBAC states can now share the cost of checking in with their teachers to see how implementation is going and to identify which policies, which textbooks and materials, and which types of training seem to be making a difference.
Our reliance on state and local leadership has a lot of advantages鈥擴.S. Supreme Court Justice Louis Brandeis famously referred to states as . But, until now, we鈥檝e struggled to interpret what those different laboratories were producing, since they each used different standards and different tests for measuring student performance.
The National Assessment of Educational Progress was never intended to measure the achievement gains of individual students or to associate those gains with particular interventions. As a result, it鈥檚 been difficult to isolate the effect of state differences in policy on student achievement from the effect of state differences in income or parental education, or even variation in state standards. The new federal law, ESSA, guarantees that states will be free to experiment with different approaches to teacher evaluation or teacher training. But, unless there is a shared assessment, it will be harder for any of us to see which of their experiments are working.
