President Obama had been in office only a couple of months when he sought to turn one of the most powerful forces in American education to the cause of school improvement. I am calling on our nation's governors and state education chiefs to develop standards and assessments that don't simply measure whether students can fill in a bubble on a test, he declared, but whether they possess 21st-century skills like problem solving and critical thinking, entrepreneurship and creativity. Standardized multiple-choice tests increasingly dominate what and how teachers teach in American classrooms, and the President lent his voice to a growing chorus of critics of the mostly low-level measures of student and school performance. The Administration has joined the National Governors Association, the Council of Chief State School Officers, and a host of education reformers in calling for new tests that rely more heavily on tasks ranging from short-answer questions to extended work like writing essays and conducting laboratory investigations that require students to develop their own responses rather than select from among answers presented to them in test booklets, to apply knowledge they've acquired rather than merely reproduce it. The U.S. Department of Education is funding two national testing consortia to build such tests, and it has called for more performancebased tests in its blueprint for reauthorizing the federal No Child Left Behind Act, a law that mandates extensive standardized testing. This is a daunting agenda. Performance questions are more expensive, more challenging, and more time-consuming to create, administer, and score. And when policy makers first embraced performance tests two decades ago, the experiment didn't go very well. But researchers learned a lot from the experience, and, more recently, performance measures have been deployed on a large scale in many countries with high-achieving education systems. Making the Case The case for more performance measures is compelling. By their nature, the multiplechoice-centric tests that states have used to comply with NCLB aren't well-suited to judging students' abilities to express points of view, marshal evidence, and display other advanced skills on which today's employers and school reformers have placed a high premium. Wellconstructed performance measures, in contrast, go beyond measuring the information students have memorized to their ability to analyze and synthesize what they've learned. By tapping into students' advanced thinking skills, performance assessments yield a more complete picture of students' strengths and weaknesses. The prevalence of testing under NCLB that gauges such low-level skills as recalling and restating information led many teachers to make such skills the focus of their instruction. Performance measures have the potential to put teaching on a higher trajectory. Studies in Maryland, Kentucky, and other states that have experimented with performance measures in the past and in Singapore, Australia, England, and other countries that are using performance assessments today report that teachers assign more writing and other higher-level tasks when standardized tests measure such skills. It's not a coincidence, perhaps, that nations that have introduced performance assessments most aggressively have turned in some of the best results on international measures of advanced skills in recent years. And by potentially giving teachers a role in scoring essays and other performance measures--the way the Advanced Placement (AP), International Baccalaureate (IB), and PISA (Programme for International Student Assessment) programs do today--the new assessments would help teachers become further invested in teaching advanced skills and provide a valuable professional development opportunity to engage with the quality of student work. The new tests thus would tie accountability much more closely to the improvement of classroom instruction than NCLB has done. …