The Hechinger Report is a national nonprofit newsroom that reports on one topic: education. Sign up for our weekly newsletters to get stories like this delivered directly to your inbox.

As students return to school for the 2021-22 year, educators are thinking about how to teach children who have missed months of instruction because of the pandemic. Should they step back or proceed ahead? Should math teachers start fourth grade, for example, by reviewing important third grade topics such as multiplication and fractions? Or should they start fourth graders off at the usual beginning, learning place values through millions and billions, as if third grade hadn’t been disrupted?

Educators have long debated about the best way to catch up kids who have holes in their knowledge. Some argue that it’s critical, particularly in math, to go back and master foundational concepts. “Math is cumulative,” Joel Rose, the chief executive officer of New Classrooms, a nonprofit company that sells math software to schools, once told me. “You just can’t learn linear equations if you don’t know how to multiply.”

That seems logical, but in practice students often get trapped in remedial classes and fall further behind their peers. Even efforts to tailor remedial instruction to each child’s needs through computerized instruction, such as New Classrooms’ Teach-to-One program, haven’t panned out well in independent evaluations.

In the summer of 2020 after schools had been shuttered for about three months, another nonprofit maker of math software for elementary schools called Zearn was thinking about a way to keep kids on grade level. The company borrowed a page from corporate America. If just-in-time supply delivery works for production, might just-in-time review work for learning?

Zearn’s instructional designers tore apart the math curriculum for each grade, identified prerequisite skills for each lesson and wove in ultra-brief review tutorials throughout. For example, before introducing division to third graders, students first received a mini review of multiplication by picturing four rows of five desks. Then students could take a crack at this problem: Ms. Alves puts 21 papers in 7 piles. How many papers are in each pile?

In a four-month test of the new lessons during the 2020-21 pandemic school year, Zearn found that students mastered 27 percent more grade-level topics than students in classrooms that picked up where they left off in spring 2020 and taught (or retaught) many lessons from the previous grade level. More importantly, students using the new curriculum progressed through grade-level math problems more easily than students who were given more remedial instruction. Students who reviewed more were more likely to struggle and get stuck.

“That really surprised us — jaw drop,” said Shalinee Sharma, CEO of Zearn. “I could not fathom that kids would struggle that much more when they were remediated. The whole point of remediating a kid is so that they struggle less.”

To Sharma, the increased progress and the decreased frustration is a sign that just-in-time review is a much better way to catch up than reviewing old material sequentially. But this data analysis isn’t proof that just-in-time review was the key to success. It could be that the teachers who opted for the new curriculum were better teachers and they would have gotten the same results even if they had remediated more. You would need to randomly assign classrooms ahead of time to each approach and see how the students fared.

The concept of just-in-time review isn’t entirely new. Good teachers often remind their students of something they already know before introducing a new but related concept. Tutors sometimes coordinate with classroom teachers to review foundational concepts so that a struggling student can better follow along in class. What makes this new is that instructional designers made some guesses about what all students should review during class time.

Thousands of classrooms elected to try the just-in-time review approach in the fall of 2020. But many classrooms didn’t. Instead, their teachers retaught many lessons from the previous year, spending time, for example, on addition and subtraction of two- and three-digit numbers, a second grade skill, at the beginning of third grade.

Zearn observed that teachers were less likely to believe that Black and Hispanic students and those from low-income families were ready to engage with grade-level work and had them to go through more review lessons than their white and wealthier peers. Students in high-poverty schools were nearly twice as likely to receive remedial lessons as students in low-poverty schools. In schools where most of the students are Black or Hispanic, nearly one in six students was remediated, even when the students had successfully completed grade-level lessons.

It’s important to note that some amount of struggle is desirable when learning. In Zearn’s software, students are encouraged to try again after a wrong answer because the company believes that learning happens as students sort out their misunderstandings and correct their mistakes. Pop-up hints and examples guide struggling children along the way. Children are required to get 100 percent correct on a series of problems, mastering the topic, before moving on to the next lesson. Frequent mistakes are nothing to worry about.

However, sometimes children get stuck. If a child enters the wrong answers over and over again and a series of three computerized hints fails to help the child make any progress on a digital worksheet, a “struggle alert” flares up on the teacher’s computer screen. When these alerts flare, it’s a sign that computerized instruction isn’t going well and the student needs human help.

Behind the scenes, Zearn’s data analysts could see the increase in struggle alerts for kids who were doing more review lessons. That presented a puzzle. Perhaps the children who were getting more remedial instruction might have been more behind than the children who got to try the new curriculum?

Zearn consulted with Steven Levitt, the famed “Freakonomics” economist at the University of Chicago, and his team at the Radical Innovation for Social Change who helped Zearn analyze their 2020-21 data more rigorously. They focused on a smaller group of 50,000 students in 6,000 third to fifth grade classrooms where teachers had been using Zearn lessons before the pandemic. The demographics of the students were strikingly similar between the 2,300 classrooms that opted for the new curriculum and 3,700 classrooms that decided to review old lessons. More importantly, unproductive struggle alerts hadn’t been higher in the classrooms that chose the remedial route before the pandemic hit. The students seemed to be similar.

Yet, between October 2020 and January 2021, classrooms using the new curriculum, which emphasizes grade-level content with just-in-time review, had about five fewer struggle alerts per lesson than classrooms that focused more on remedial instruction. Teachers dealt with 10 as opposed to 15 struggle alerts per lesson. The results were jointly published in May 2021 by Zearn and TNTP, formerly known as The New Teacher Project.

For Black and Hispanic students, the results for just-in-time review seemed even stronger. In classrooms where the majority of students were Black and Hispanic, students completed 50 percent more grade-level lessons with the new curriculum than with a remediation approach. Remedial classrooms with mostly Black and Hispanic students had 18 struggle alerts per lesson, six more struggle alerts than classrooms of similar students who were taught with the new curriculum.

Children didn’t completely catch up to grade-level using the new curriculum but they learned more this way than by methodically reviewing below-grade-level lessons.

I learned about this Zearn study while attending a recent webinar about how to “accelerate” learning. This whole year I’d been hearing educators say, “Accelerate, don’t remediate.” The slogan has a nice ring to it but I had no idea what it meant in practice.

As I learned about this particular example, which was and is marketed to schools as a “learning acceleration scope and sequence,” it was clear to me that it’s actually a form of remediation, where children are promoted to the next level while still getting truncated remedial lessons. And nothing about it is sped up or faster, as the word acceleration implies. Indeed, instructional designers had to make some tough edits and jettisoned two to four weeks of grade-level material that students would have ordinarily learned to make room for the review material. Arguably, students were learning more slowly than they did before the pandemic

Despite my peeve with word usage, the just-in-time remediation approach appears promising and worthy of further study. Implementation will be a challenge. It’s a heavy lift for a teacher to figure out the essential prerequisite skills for each lesson and which grade-level units are less essential and can be cut. By no means does this just-in-time approach require computerized instruction, such as Zearn’s, but it requires teams of veteran educators who can dissect and rebuild lesson plans. They’ll need hours more planning time to do it well.

I hope schools track how they review this year and see how student outcomes differ. It may help us not only recover from the pandemic but also shine a light on how to teach children who are behind in the future. Perhaps it’s a better way to teach everyone, even children who are advanced, with a quick reminder of what they should already know.

This story about remedial education was written by Jill Barshay and produced by The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education. Sign up for the Hechinger newsletter.

The Hechinger Report provides in-depth, fact-based, unbiased reporting on education that is free to all readers. But that doesn't mean it's free to produce. Our work keeps educators and the public informed about pressing issues at schools and on campuses throughout the country. We tell the whole story, even when the details are inconvenient. Help us keep doing that.