“Well y’all, we have a problem. Somebody lost our meter sticks.” I taught an integrated algebra and physics course for five years at a high school in the Bronx, and each year I taught a lesson that started with this white lie. The crux of the lesson was that we needed to know the height of our classroom tables, but since we had no meter sticks we would have to figure out a new way to do it. Students spent the period developing and implementing a plan to calculate the height. At the end of the period, one of my colleagues would show up with an armload of meter sticks. “What!” the students would shout. “You made us do mad work when we could have just measured!”
The lesson objective was about problem-solving skills and using mathematical relationships to determine quantities, but one of the biggest take-aways was always that having a meter stick to find lengths is a big time saver. No student ever used the “mad work” method to find a length after they had the right tool.
While this lesson was easy for my 9th grade students to understand, federal education policymakers have spent the last decade and a half failing to figure it out. The No Child Left Behind Act requires schools to spend huge amounts of time, money, and energy testing every student annually for much of a child’s K-12 career in order to collect a set of data that is not particularly useful. By contrast, the “gold standard” for educational testing, the National Assessment of Educational Progress, or NAEP, collects more meaningful data at a fraction of the financial and temporal cost.
For evidence of NAEP’s value we need look no further than the number of citations the data receive in policy conversations. Civil rights leaders, education scholars, and other thought leaders routinely cite NAEP data instead of state standardized test scores. In fact, many advocates of the current testing regime argue that the comparability of state test results to NAEP scores show the state tests’ validity. Those individuals may not have an issue with using two tests to find out a single piece of information, but all advocates of good government should question the wisdom of such a policy.
The current public conversation about testing revolves around two extreme positions, both of which are fundamentally flawed. On the one side are those who argue we should have no standardized testing. On the other are those who insist that we must have annual testing and even that we ought to ramp up the implications of those tests. But we can transcend the two camps to find a middle ground that acknowledges the value of data collected through standardized testing and the need to move beyond the idea of testing every student every year by means of stock assessments that are of questionable relevance to student learning.
The promise of No Child Left Behind lay in its requirement that schools collect subgroup achievement data. I know from experience that a school that works well for some students may not work well for all students. I attended a high school where the majority of students scored far above average on state tests and the SAT. Dozens of students from my graduating class attended and excelled at selective universities; however, about half of the African-American males who started with me in 9th grade dropped out before graduating. We can disagree regarding how to go about addressing this problem, but we ought not disagree about whether or not this represents a problem. We need hard data on students’ performance that we can analyze to find trends like this one. We cannot allow teachers, schools, or communities to systematically under-serve certain segments of their population with impunity.
The failure of No Child Left Behind is that it gives shallow and inconsistent data while distracting schools from the things we might be doing to develop deep learning and genuine intellectual achievement. My teaching colleagues and I currently work in a system that goes to the other extreme, where “accountability” limits rather than supports the great work we can do.
No Child Left Behind has shown us that state-sponsored testing systems are costly, inefficient, and prone to corruption. They take up “all the air in the room,” particularly at struggling schools, leaving little time for performance assessment, inquiry-based lessons, and the arts. As currently implemented, state standardized tests violate the basic principles of formative assessment that can be used to support learning, and they have shown little reliability in measuring higher-order-thinking skills.
The promise of the Common Core State Standards, in my view, is that they allow us to implement a curriculum that moves away from rote learning toward the kinds of critical-thinking skills young people need to actively and productively contribute to the 21st-century world. But this promise is being undermined by 20th-century assessment tools and politics.
Taking a Sample
As the NAEP shows, we have the mathematical and psychometric tools to assess the deeper conceptual understanding of a sample of students in meaningful ways. If Congress removed the “every student every year” mandate for grades 3-8, we could start crafting assessment systems that provide more meaningful data tomorrow. By using random sampling, we could generate a subset of any student population that would provide insight into how the population as a whole and relevant subgroups were achieving. As with the NAEP data, we could identify large trends within schools and school systems and leverage these lessons as we create policy, without leaving such a large footprint on daily teaching and learning.
Random sampling has the added benefit of creating a layer of complexity that would disincentivize some of the current gaming that has plagued accountability systems. Corrupt administrators and politicians would no longer be able to game the system by manipulating the students who take the test or cutting scores. Without the big test looming for all their students, teachers would feel less pressure to teach to the test. To measure their students’ progress and learning internally, schools would be free to develop performance-based assessment strategies that work for their communities (like we do in the New York Performance Standards Consortium). The billions of dollars currently going to private, for-profit testing companies could be redistributed to schools to promote meaningful reform in teaching and learning. All this and voters could still hold schools accountable based on meaningful data collected in an efficient way.
For the past decade and half, No Child Left Behind has made a lot of people mad. Teachers are mad when we have to narrow our curriculum to prepare students for a limited test. Students and their families, as well as teachers, are mad that states can use these tests—based on a single performance on a single day—to determine so much of their future. Citizens are mad at politicians who manipulate test scores for short-term political gain or funnel public dollars to private testing companies. Educators, students, and families are doing lots of work to satisfy testing mandates while finding little time for enrichment or deeper learning.
As I write, Congress is still debating the testing provisions to be included in the proposed NCLB legislative rewrites, though the annual-testing requirement still seems to the default preference. As a teacher who’s dedicated to giving students the greatest experience possible, I’m disappointed by this. To quote my former students: “We’re doing mad work.” Let’s stop doing “mad work” that makes us mad. Let’s start working smarter to collect meaningful data in an efficient way.