Lamar Alexander, the new head of the Senate Health, Education, Labor and Pensions Committee, has set off the long-delayed reauthorization process for ESEA. Testing is taking center stage in the debate.
There is no shortage of critics (and no real defenders) of the quality of current tests and how they are used for accountability. There are plenty of reasons to keep working on making tests more valid and reliable, and reducing incentives for teaching to the test by requiring that test results be interpreted in light of other facts about school performance.
But we can’t lose sight of what tests alone can do, and what would be lost if we made drastic cutbacks that, for example, made it impossible for anyone to track an individual student’s growth every year.
Here is a look at what we would give up. Matt Chingos and his colleagues at the Brookings Institution have made excellent points about one thing we’d lose: the ability to fairly measure how much individual schools contribute to student learning gains. But we’d lose much more than that.
First, reducing testing to occasional snapshots of average student performance will make it even harder to identify and target resources and support for those students who are falling behind, as well as accurately identify reasons for high or low school performance. Without the ability to track student growth, it would be difficult to assess whether high or low test scores are a function of a student’s prior learning experiences or family background. Years could go by before anyone knew that a student was falling behind, or that the performance of a school was suddenly declining. Nobody—not parents, school staff, or district leaders—can fix a problem quickly if they don’t know about it.
Second, we lose any ability to measure how much a given teacher is contributing to student content mastery. There is reason for controversy about whether such measures should be used by states and districts to set teacher pay. But value-added measures are irreplaceable resources for school leaders who are serious about evaluating teacher effectiveness and supporting professional development.
Third, parents will lose information they need to make smart school choices. School choice is increasingly available to families across the country. Research shows that parents pay close attention to the performance of schools when making choices. Though parents can refer to school “snapshots,” student assessment data and trends provide families with valuable information when making choices.
Finally, we will lose most of the analytic capacity that allows us to learn how to improve our education system. Scholars’ access to annual testing data has shown us that a great teacher is by far the greatest influence on student learning—and has proven that the effects of good or poor teaching can persist throughout students’ lives. Without annual testing we would not be able to identify which types of charter schools are succeeding and should be replicated, or to hold charter schools and traditional public schools accountable for the same measures of student growth. We would not know how various pre-service training, certification programs, and professional development strategies vary in their effectiveness.
A consensus that we can find better ways to track student performance is clear. Teachers and parents have legitimate complaints, for example, about how disruptive large blocks of testing periods can be. Exactly what will do better is far less clear.
Testing can have a lighter touch if states were to move to random sampling or alternate-year testing or some combination of the two. In each case, some of the benefits of annual testing would still be lost.
Instead of throwing out annual testing, it is time to challenge the testing industry to design less intrusive ways of regular, meaningful testing. Why, with all of the technology we have for test administration and rapid response, do we still administer tests as a multi-hour annual event? Can we make use of the technology schools already have to test more briefly (using just a handful of questions at a time), but more frequently so that teachers can track student growth in real time? This also reduces the stakes associated with any one testing episode so that an anomalous low score due to illness or other distractions could be discounted.
Scholars and elected officials need to address the legitimate issues that teachers and parents raise about the current testing regime and find solutions. By throwing out annual testing, we risk blinding ourselves to invaluable information and providing opportunities for corrective action that can only come from tracking student progress every year.
