Standards and Assessments

Page 5 of 6

2. On developing sophisticated assessments and accountability mechanisms

The development of new, high-quality classroom assessments and accountability mechanisms, linked to common standards, is an important priority—indeed, a necessity if common standards are to achieve their maximum effect for improving math and science education for all American students. Assessments aligned with common standards will also be essential to the creation of useful, accurate measurements of teacher, school, district, and state performance.

As Stanford University professor and assessment expert Edward Haertel wrote in a paper presented to the Commission, “assessment is woven into the fabric of educational practice in the United States. Individual assessments help determine the classifications of students as gifted, learning disabled, English Learners, or ADHD. The quizzes, unit tests, and final exams that teachers create or choose help determine the pacing of classroom instruction, instructional grouping, and marks and grades, as well as informing students about expectations for learning and about their success in meeting those expectations. Advanced Placement and International Baccalaureate tests define ambitious curricula for respected high school courses. The SAT and the ACT are central to the sorting and selecting process at the point of college admissions. High school exit examinations are viewed as a form of quality assurance, but also stand as significant barriers to graduation for substantial numbers of students. State testing systems mandated under the No Child Left Behind Act of 2001 (NCLB) define school-level success or failure, and a range of sanctions are imposed if scores repeatedly fall short of targeted levels⁵¹.”

The development of new, high-quality classroom assessments and accountability mechanisms, linked to common standards, is a necessity if standards are to achieve their maximum effect.

Of these, the most important for raising mathematics and science achievement for all American students are classroom assessments and assessments for accountability. If well crafted and administered appropriately, classroom assessments can provide information about student learning and help teachers improve instruction. If well aligned with standards or other clear statements of expectation, assessments for accountability can provide information about how students, teachers, schools, or even states and nations are performing and whether or not students are learning prescribed curriculum; that information, in turn, can help shape improvements to instruction and to educational practice and policy. Experience has shown that it is not easy to get assessments right: assessments are frequently used for purposes they were not designed for, and rote preparation for “high-stakes” tests displaces or distorts other learning goals. In science, for example, the need to obtain reliable results from tests that are easy and inexpensive to administer has driven assessments—and instruction—toward the first strand described in Taking Science to School (“know, use, and interpret scientific explanations of the natural world”) and away from the other three more complex and difficult-to-assess competencies.

The Commission believes that better assessments will be crucial to guide and reinforce improvements to mathematics and science instruction in American schools and colleges, and that those assessments should be closely linked to the new fewer, clearer, and higher standards. In addition, as Haertel has argued, it will be essential to improve and clarify “the rules by which [assessments] are used or interpreted,” which may require “decoupling the multiple purposes for which some tests are used” and making appropriate changes in federally mandated accountability systems. Following Haertel’s recommendations, the Commission therefore urges development and implementation of five interconnected types of classroom-level and accountability assessment:

• Portfolio-based school accountability, which should incorporate student- or classroom-level math and science portfolios

• Performance assessment component for school accountability, which should include matrix-sampled school-level performance assessments

• Classroom assessment for learning, using improved curriculum-embedded formative assessments

• Better high-stakes tests that are closely linked to new mathematics and science standards

• Better decision rules for evaluating school-level assessment results

Developing these systems and putting them into place would need to be phased with care, with supports provided to districts, schools, teachers, and parents and communities about what each component is intended to measure and how it operates.

The Commission also believes that new assessments should be informed by and calibrated against the most reliable international measurement systems in mathematics and science—the Programme for International Student Assessment (PISA), which periodically assesses the skills and knowledge of 15 year olds in mathematics, science, reading, and problem solving and measures changes in student performance, and Trends in International Mathematics and Science Study (TIMSS), which periodically measures the performance of fourth and eighth graders—and the skills and knowledge those systems assess. Wider use of internationally benchmarked assessments would give states and the federal government a more meaningful picture of student and school performance and would inform district and state efforts to improve American schools. The science framework for the 2009 National Assessment of Educational Progress also lays out a well-regarded, comprehensive approach to assessing content knowledge, its application, and students’ command of the process and practice of science for students in grades 4, 8, and 12. The new NAEP framework should inform the development of new common state and classroom assessments.