Imagine that you’re a sixth-grade math teacher. It’s the first day of school, and the vast majority of your students arrived multiple years behind where they should be. Your job is to teach them concepts such as understanding percentages and dividing fractions. Both will appear on the sixth-grade state test, but your students never successfully learned much of anything about the basics of fractions in fourth and fifth grade.
What would you do?
Would you try to go back and address their unfinished learning from prior years and run the risk of not having enough time to cover all of the sixth-grade-level material? Or would you march dutifully through grade-level content and try to fill in gaps where you can, knowing that it probably won’t be enough?
Today, New Classrooms Innovation Partners, of which I’m co-founder and CEO, published The Iceberg Problem: How Assessment and Accountability Policies Cause Learning Gaps to Persist Below the Surface….and What To Do About It. In it, we contend that, while college and career readiness needs to be the goal for all students, policies that were oriented around annual grade-level expectations may, at least for middle school math, make it harder for some students to accomplish that objective.
We believe in accountability and the importance of providing students, teachers, parents, administrators, and policymakers with objective information on student performance. We also believe in the critical importance of high expectations and policies that seek to counteract systemic biases that can keep students—and especially students of color—from accessing rigorous, high-quality instruction.
At the same time, we cannot ignore the fact that math is cumulative and that many (if not most) students enter middle school with unfinished learning from prior years. According to the 2017 National Assessment of Educational Progress (NAEP), only 40 percent of fourth-grade students were proficient in math. For black and Hispanic students, the numbers drops to 19 percent and 26 percent, respectively.
Failing to acknowledge this simple truth and assuming teachers can somehow make up for years of unfinished learning in math while also comprehensively covering new grade-level material is not only dishonest, but it sets students up to fall even further behind.
And it may be getting worse. The adoption of more rigorous standards and tougher tests may be helping to close the honesty gap between where students are actually performing and the trajectory they should be on. But it also means that the ground teachers are supposed to make up in a single year is even larger.
Some may argue the solution lies in training teachers to differentiate their instruction. But savvy educators know that, absent fundamentally restructuring the classroom itself, it’s far easier said than done. Refusing to confront this reality and assuming that differentiated teaching strategies are a viable approach to dealing with students’ wildly disparate incoming performance levels only ends up perpetuating them.
The improbability of students catching up during the course of a single year in math class was reinforced by a recent study conducted by the Institute for Education Policy at Johns Hopkins University. Using publicly available state assessment data, researchers analyzed sixth and eighth grade cohort data from 1,651 schools across six states and the District of Columbia. They found that less than 1 percent of schools were able to improve their overall proficiency levels (by any amount) between sixth and eighth grade while also reducing the number of students scoring at a level 1 on their state math assessment.
Recognizing the importance of measuring progress, forty-eight states included growth measures in their accountability plans submitted under ESSA—a worthy and important shift from No Child Left Behind and its focus on grade-level proficiency. Yet what many may not fully realize is that the assessments undergirding these accountability systems still focus almost exclusively on grade-level items. As a result, these instruments are unlikely to detect most learning gains made by students who are far above or below grade level.
All of this might seem interesting only to policy wonks until one realizes that, from a teacher and school perspective, the message is clear: Teach grade-level material. While there may be sound pedagogical reasons to do so in reading, math is different. When middle school students’ unfinished learning stemming from insufficient elementary school math follows them into higher grades, it can stymie their ability to succeed and cause new learning gaps to accumulate.
We’ve seen ample evidence of this phenomenon in our own program, Teach to One: Math. There we employ a combination of live, online, and collaborative learning in ways that enable each student to receive personalized instruction each day. Partner schools end up implementing the model in different ways that provide varying degrees of exposure to grade-level material depending on individual school preferences.
Does prioritizing grade-level exposure in math make a difference? A recent study of Teach to One by MarGrady Research found that, though students across all schools grew an average 20 percentile points over three years on NWEA MAP assessments, those schools that were more focused on overall learning growth (including below-, on-, and above-grade-level skills) saw their students grow by 38 percentile points, while those more focused on annual grade-level proficiency grew 7 percentile points.
More evidence is required before a causal relationship between student progress and assessment and accountability policy can be proven. And the failures of remedial education provide a cautionary tale for what happens when “meet them where they are” ends up serving as the rationale for instruction that never gets students to where they need to be.
At the same time, the presumption embedded within our assessment and accountability systems—namely, the idea that, in math, grade-level content is best for all students at all times—has little evidence behind it. In the 2000s, for example, there was a policy push to place eighth-grade students into algebra who would have otherwise taken pre-algebra. A subsequent Brookings study found that the effort “had unintended and damaging consequences, as some 120,000 middle-schoolers [were] now struggling in advanced classes for which they are woefully unprepared.”
So what should policymakers do?
ESSA requires that all students in grades three through eight take an assessment aligned to grade-level standards. It also requires that results from those assessments be included in a statewide accountability system. Given these parameters, The Iceberg Problem includes several recommendations for states and districts to consider in order to create the space for new approaches to math instruction, assessment, and accountability. These include:
- Using adaptive assessments that measure true learning growth, not just grade-level proficiency. If the goal is ultimately for all students to achieve college and career readiness, educators need the data to know where kids are actually starting from each year. Adaptive assessments do this much better than grade-level tests because the tested items cross multiple grade levels. Cities such as Chicago and Tulsa have actually created their own local accountability systems based on these kinds of tests instead of using the state tests (which students there continue to take).
- Modifying state accountability systems to look at growth over multiple years or emphasize key grade levels. There are many kids who begin each school year well behind where they should be. They might need a different pathway—one that takes more than a single year—to catch back up to proficiency. States can create the space for that to happen within the bounds of ESSA.
- Launching Math Innovation Zones, as Texas has done, which identifies innovative districts and schools, matches them with partners, and provides them with a different accountability system focused on learning growth in service of grade-level proficiency.
At the same time, policymakers—and policy-thinkers—must begin to grapple with what a next-generation assessment and accountability system might look like that preserves the transparency, clarity, and equity guardrails embedded within the current system, while also ensuring that there are instructional incentives aligned to what’s truly best for each student. This is especially critical for math.
Accountability must continue to be a pillar of our efforts to improve our nation’s schools. We hope the Iceberg Problem provides policymakers with a more nuanced understanding of what the current system can and cannot do, so that future efforts can build upon these lessons.