The Next Generation of Assessments Can—and Must—Be Better
As educators, we know that today's students will enter a workforce in which they will have to not only acquire information, but also analyze, synthesize, and apply it to address new problems, design solutions, collaborate effectively, and communicate persuasively. Few, if any, previous generations have been asked to become such nimble thinkers.
Educators accept the responsibility to prepare our students for this new and complex world. We also know that in our current high-stakes context, what is tested increasingly defines what gets taught. Unfortunately, in the United States, the 21st century skills our students need have gotten short shrift because our current multiple-choice tests do not test or encourage students' use of these skills. In fact, a recent RAND Corporation study found that on current state tests, only about 2 percent of math items and 20 percent of English language arts items tap higher-order skills.
This represents a step backward from where many states were in the 1990s. Over the last decade, the nature of federal requirements under the No Child Left Behind Act (NCLB)—including the expansion of grade-level testing—caused states to abandon the performance assessments many had launched in favor of cheaper, machine-scored tests.
Despite the fact that these tests have enormous consequences for instruction and school decision making, the average spending on tests required by NCLB is only $25 per pupil—less than one-quarter of 1 percent of overall education costs. This is considerably less than high-achieving countries spend on more robust assessments (or what wealthy parents spend on SAT, AP, and IB exams) and far less than any of us spends on routine checkups for our car each year.
But there are also other costs we fail to consider. Because assessments no longer ask students to use technology or to engage in research, extended problem-solving, using technology, writing, or oral communications, there are few incentives to emphasize these skills even though they are the most in demand in today's economy. If our goal is to improve the quality of teaching and learning, our current testing behavior is penny wise and pound foolish.
Unless major changes are made in the nature of assessments, we fear that the most important aspects of the Common Core State Standards will not be evaluated or adequately taught. That's why our recent report, Developing Assessments of Deeper Learning, investigates the costs and benefits of high-quality assessments that promote 21st century skills.
We found that nearly all districts are spending considerable money on test prep, interim and benchmark tests, data systems, and staff time to manage and analyze tests. These outlays bring the total cost of current testing practices to more than $50 per pupil nationwide, with some states spending over $100 per pupil. Yet these investments do not improve the quality of learning because they are focused on boosting scores on low-level tests.
Many assume high-quality assessments that include open-ended items and tasks will cost even more, but we found that this need not be the case. Cost analyses show that several factors can make higher-quality assessments feasible and affordable: Costs are dramatically reduced when states act together in consortia that unlock economies of scale. Online delivery and efficient scoring of open-ended tasks by computers (where the technology allows) and by teachers who are paid professional development stipends reduce costs also.
Furthermore, involving teachers in developing and scoring assessments produces the double benefit of improved instruction and more efficient use of resources. As former Kentucky school chief Gene Wilhoit recently observed, "Teachers get great professional development when they score open-ended tasks. [Having teachers score writing portfolios and mathematics tasks in Kentucky] resulted in a more knowledgeable workforce, and I'm convinced it helped improve reading and writing performance by students."
With these innovations—and the integration of formative and interim tools into coherent systems like those being developed by PARCC (Partnership for Assessment of Readiness for College and Careers) and the Smarter Balanced Assessment Consortium—states and districts could spend considerably less than they do now for assessments that are considerably more robust in evaluating college- and career-ready skills.
We believe that with intelligent investments, assessments of deeper learning can become feasible and affordable and that now is the time to implement assessments that empower teachers to focus on the skills students really need to learn.
Linda Darling-Hammond is the Charles E. Ducommun Professor of Education at the Stanford University School of Education and codirector of the Stanford Center for Opportunity Policy in Education.
Frank Adamson is a policy and research analyst for the Stanford Center for Opportunity Policy in Education.
ASCD Express, Vol. 8, No. 18. Copyright 2013 by ASCD. All rights reserved. Visit www.ascd.org/ascdexpress.