Standardized tests don’t get a lot of love. Kids don’t like to take them. Teachers resent the time they take away from instruction. Critics say they reduce knowing to the ability to parrot the right response at the expense of critical thinking and problem solving. And the wonks say that if those high-level skills don’t get measured, they won’t get taught. Indeed many charge that the emphasis on standardized testing will be the ruin of public education, most recently in this widely circulating petition launched by an anti-testing coalition.
But what if standardized tests asked test-takers to do more than fill in the bubble next to the correct factoid? What if they challenged students to analyze, synthesize and solve real-world problems? What if it were impossible to distinguish a standardized test from a well-constructed, interactive lesson? What if they were fun?
Truth is, the field of testing has changed a lot since those of us over a certain age were in school. In those days, standardized tests mostly asked test-takers to recognize specific facts or answers from a list of multiple choices, which ended up testing memory more than analysis. But today’s standardized tests are different. They often include open-ended items that call on students to think critically and creatively. Even the much maligned multiple-choice question has undergone a transformation as test designers have developed ways to structure items that require more analysis from test-takers. Just ask any 11th grader who recently took the SAT. While the quality and richness of the newer assessments vary, for the most part, these aren’t your father’s standardized tests anymore.
Even so, most current state tests would not be mistaken for a hands-on learning experience. The problem isn’t so much standardization, but the cost to administer and score multi-step, open-ended items. The introduction of new technology and advancements in the field, however, promise to deliver a new generation of standardized testing that requires students to show what they know by applying it to something new and unfamiliar. In short, the new tests will blur the line between assessment and good instruction.
Yesterday, NCES launched an interactive science tasks website that features several hands-on and computer-based investigations that do precisely that. Each task sets up a problem that students are asked to investigate through simulated experiments and then explain their findings. I could describe these, but I encourage you to try them out for yourselves, especially the computer-based tasks. These are so much like the kind of classroom assignments we would want to see all students recieve, they hardly seem like a standardized test at all. Yet they were administered and scored as part of the 2009 NAEP-science assessment.
Widescale assessments like this are only beginning. The two state consortia developing assessments for the common core state standards are working to develop computer-based testing that measures both students’ knowledge and skills through authentic and problem-based tasks. They are further collaborating on protocols for scoring such problems by computer as a means to keep the costs down. We have been keeping track of progress in common core implementation and policies, including assessment examples as they come available.
One of the often heard criticisms of standardized tests is that they encourage “teaching to the test,” implying that this is a bad thing. Interesting, though, that few see this as a problem with Advanced Placement courses, which in fact are designed to prepare students for the AP exam. The difference, of course, is that AP is commonly viewed as a test worth teaching to.
Current standardized tests are big improvements over what they used to be, but they are only part way to where they can be. This next generation of assessments is already showing the potential to both model good instruction and measure learning. As such, their influence over what happens in the classroom should be welcome, not something to dread. – Patte Barth