Tomorrow's Tests

Commentary | September 22, 2009

In late July, in what he called the "equivalent of education reform's moon shot," Secretary of Education Duncan announced the priorities for the administration's $4.35 billion "Race to the Top" fund to spur education improvements. The secretary wisely prods states to expand public charter school options, improve the quality of teaching, and address failing schools. But, unless his plans for improving our underlying navigational instruments—the tests that generate the data to determine which students, teachers, and schools are succeeding—are just as bold, the "moon shot" will surely fail.

Given to millions of students each year, states' tests are primarily multiple-choice measures of basic skills in reading and math. They are useful for showing if and how many students are meeting the proficiency goals of the federal No Child Left Behind Act and state accountability systems. But they are limited in use and value. Today's tests do not assess all that they should, nor do they tell us very much about how to help students do better.

The nation needs an assessment system with a broader purpose, one that is designed not only to measure student performance, but also to support and improve the daily instructional work of teachers. To do this, tests must measure more than whether a student answered a test question right or wrong, but also why a student scored a certain way and what to do about it.

The best teachers already know what the latest research on learning tells us—that knowledge and skills go hand in hand, that so-called "higher order" skills cannot be learned without knowledge and that knowledge will not be sustained if students do not have the skills to apply and use it. Tomorrow's better tests, then, will be designed to measure knowledge and the application of knowledge at the same time. And perhaps most importantly, these tests will show us not only what students are learning but also how they are learning and why they may or may not be gaining particular skills or knowledge.

The technology to create better tests already exists. Using simulations, we can present complex, multi-step problems for students to solve, and we can collect detailed information about an individual student's approach to problem solving. These performance-based assessments, which are designed to mirror more complex, or real-world situations, are increasingly common in the military, medical education, and international educational benchmarks such as the Programme for International Student Assessment test. Like modern medical testing, they provide much more specific and reliable data—data that both strengthens accountability and helps improve performance.

Both President Obama and Secretary Duncan agree. The president, in his education address to the Hispanic Chamber of Commerce in March, called for "assessments that don't simply measure whether students can fill in a bubble on a test." And, the secretary has set aside $350 million of his "Race to the Top" fund to develop these assessments. But to date, plans for these funds are troublingly vague.

Dispersing funds to states with crossed fingers for more testing innovation will not suffice. Nor will a national test built on the foundations of our current inadequate testing systems—the anticipated next step in assessment if states embrace plans for common educational standards.

It's time for the federal government to lead a full-scale transformation of testing in American schools. Like an Apollo program for education assessment, the nation needs a long-term and cohesive strategy for applying decades of research on learning and measurement, experimenting with major advancements in technology, and ultimately collecting the kind of data that will be meaningful to schools, teachers, students and parents. This requires putting ideas into practice by investing in the rapid piloting and evaluation of  promising new assessment tools and approaches. It isn't rocket science. Start with small groups of schools that are ready and willing to go beyond the current low-level testing measures. Give them the freedom and funding to implement a variety of new and better ways to assess and improve student learning. Learn from them, scale what works and over time build a system of testing that measures what really matters for school improvement and student learning.