74 Interview: Expert Matthew Kraft on How the Right Tutoring Materials & Training Can Help Students Make Gains & Solve Schools’ Staffing Woes

Building a strong “tutoring infrastructure” — curricula, coaching, scheduling — can expand the pool of possible tutors, says the Brown U. professor

Matthew Kraft (Brown University)

Get stories like these delivered straight to your inbox. Sign up for The 74 Newsletter

As schools reckon with the toll of the pandemic, leaders across the country have begun to test out a strategy they hope will help students catch up on missed learning: tutoring.

Either one-on-one or in small groups, researchers say tutoring may be among the best approaches for helping youth quickly recoup lost ground. And armed with $190 billion in federal stimulus spending, the nation’s schools now have the resources to invest in new programs.

But launching an effective tutoring initiative requires more than just financial investment. Schools must answer key questions about how to structure the program: Which students will participate? What curricula will they follow? How long will sessions be? Will they take place during school or outside of it? Who will work as the tutors?

Matthew Kraft is an associate professor of education at Brown University and a leading voice on tutoring as an intervention to accelerate students’ learning. In 2021, he published a paper laying out a blueprint for how schools might effectively scale tutoring programs to help youth catch up after COVID. Finding the right curriculum and structure can lay the foundation for strong results — and can even ease staffing woes, he explained.

“The stronger the tutoring infrastructure,” said Kraft, “the less the program will rely on the individual skills that a tutor brings with them. And so it really opens up the potential labor supply pool to a much greater degree.”

The 74 spoke with Kraft over Zoom to find out what school leaders interested in tutoring should consider as they design their interventions and what pitfalls they should avoid.

This conversation has been lightly edited for clarity and length.

The 74: Can you tell me, in brief, why you’ve chosen to focus so much of your research on tutoring? What’s the potential to help students catch up on missed learning from the pandemic?

Kraft: I started studying tutoring as a doctoral student almost 15 years ago. Part of that was motivated because I have worked as both a volunteer and as a paid private tutor. It was a formative experience to be able to work one-on-one with students and see them make rapid gains in their understanding, whether it be algebra or reading. As a classroom teacher, you just don’t have nearly as many opportunities for sustained one-on-one interaction to develop those relationships. So from my own experience, I knew tutoring really had a lot of potential. 

As I began to learn more, it became clear that there’s a huge potential to move the spectrum of how we deliver instruction in public schools so that it’s not only group instruction but also individualized personalized instruction. So I started to study Match Charter School (where leaders had implemented a tutoring model), which led to the thought experiment of saying, ‘Why isn’t this something that we do more broadly? Why is it a service that those who can afford it pay for in the private market, but not something that we offer more widely in our public education system?’

Then the pandemic opened a larger public conversation. It created the opportunity to say, not only has there been huge potential for tutoring, but now we’re facing a moment where there’s an incredible need to support students at a more individualized level and accelerate their learning. So all of that came together to really motivate the work that I am doing right now.

You mentioned tutoring is what wealthy families turn to when their kids fall behind in school, and that resonated because it shows we intuitively know tutoring works. It’s just that not everyone has the resources.

Yes, there’s a lot of intuitive appeal to tutoring. In the antiquities, in Roman and Greek times, that’s how a lot of the privileged class were educated. And that remained the form of private schooling with one-on-one tutors in much of the Elizabethan era and even into the Colonial era. As we expanded access to education, the model moved away from that, to some degree by necessity, but there’s just the sense that one-on-one feedback is the natural way we learn.

Now, to add to that, we have a deep and growing body of evidence examining its efficacy through rigorous experimental research. And when you look at that body of evidence, it is very compelling. That said, it largely is built on evidence of small- to medium-scale programs implemented in person prior to the pandemic under favorable circumstances. And that’s not what we’re doing today, trying to scale tutoring to an entirely new level.

For school leaders trying to roll out new tutoring programs, what factors do you think they should consider when they’re picking out curricula?

The first question districts and schools need to answer is, ‘What is the intended outcome of a tutoring program? What are the goals?’ Because it may be that the goal is to support students belonging in school, their social-emotional development, as much as it is to accelerate their learning. It could be both. By first answering that question, I think that helps the program to backward map onto the type of curriculum that would be best suited to meet those goals. 

Often there’s a bit of a slippage that happens. There’s this notion that tutoring is a good thing writ large and so if we do something individually with tutors and students, that will produce positive outcomes. I think that’s wishful thinking. It requires a lot more purposeful alignment between choosing a curriculum that is both built on strong instructional materials, but that also complements the type of instruction students are receiving in their larger traditional classes. It doesn’t need to be the same curriculum, but it’s less productive to have tutoring use a curriculum that’s completely divorced from what students are doing in the classroom. They should be compliments.

Do you have practical tips for how school leaders can go about doing that? Who’s best suited to make the call on curricula? Maybe department heads?

The reality is there are a wealth of curricular materials designed for trained teachers [in a full classroom]. And then there is a potpourri of one-off materials for tutoring. And so there’s not as rich of a supply of curricular materials available that are designed to be able to be implemented by a tutor who has limited experience working with students. 

A key is to ask whether the tutoring materials could be used effectively by someone with some training and support, but not necessarily a formally trained teacher. When I’m searching for materials, do these materials feel accessible to a wide range of potential tutors?

That means they should be able to be broken down into very explicit instructional steps and should come with a scope and sequence that allow you to do very short formative assessments of students to figure out where to reinforce their knowledge and shore up their foundation. 

Materials also need a clear instructional sequence over the duration of a tutoring period. Some curriculum materials may be designed for a 60-minute class, but tutoring sessions may be for 30 minutes. And so, [leaders] need to map that on to the design of the tutoring program itself. And in every context, I think there is going to be some individualization required. 

So it sounds like I’m hearing that if schools get those structural components right — like finding curricula that align with classroom standards, for example, or designing lessons to match the amount of time for tutoring sessions — they can actually unlock a new pool of possible tutors. Which strikes me as important because teacher burnout is so high right now and staffing has been an issue for some programs.

I think that’s a key observation. The stronger the tutoring infrastructure — to support tutors with strong instructional materials, ongoing coaching and feedback, peer learning networks and a leadership team that will troubleshoot issues that come up like technical problems or attendance challenges — the more success tutors will have and the less the program will rely on the individual skills that a tutor brings with them. And so it really opens up the potential labor supply pool to a much greater degree.

So now in this current moment, we see a lot of districts actually moving to implement tutoring programs to help students catch up on missed learning from COVID. In these last couple years, what have we learned?

A huge advantage of the decentralized nature of public education in the United States is that there’s an amazing amount of innovation and experimentation that happens. A lot of districts are developing tutoring programs on their own and it looks different across a whole bunch of places. Those districts are individually learning a huge amount about what worked, what didn’t work. And if they continue to invest in those programs over time, there will hopefully be continuous improvement.

Where we fall short is in helping districts to share those best practices and [also what they learn about what] practices we should leave on the scrap pile of design improvement. There are efforts at the state and federal levels to build these networks and I think those have a huge role to play. 

At the same time, researchers like myself and a whole host of others are working in partnership with districts to study tutoring programs in dozens of different contexts. But it’s hard to do that while delivering rigorous research designs, which may not always be feasible in these Wild West contexts, at the same time as trying to roll it out as fast as we can. Research can be a slow process and is not always able to inform program design in real time.

We’re starting to see some new evidence coming out that districts are struggling to implement tutoring at scale and deliver the high-dosage model that we think is necessary. Some efforts to contract with 24/7 on-demand tutoring providers has led to less-than-expected uptake. And the uptake has skewed more toward students who may already be having some degree of success in school rather than [serving] students who are struggling most.

When you say the high-dosage model isn’t quite being hit, what exactly does that mean?

Districts are aiming to deliver tutoring on a regular basis to students multiple times per week. And maybe they’re shooting to do that for 1,000 students, but instead they’re only getting 100 students to come with regularity. Maybe that’s because there are transportation problems, communication problems, technology problems, tutor-supply challenges. All of those things are, at least in this initial rollout, expected implementation challenges. Taking an effective program and scaling it is, historically, something that our decentralized education system has always struggled to do well.

To school leaders who are thinking of iterating on these programs or rolling them out if they haven’t launched them yet, anything that, based on your research, they should really try to avoid?

There’s compelling evidence that efforts to scale tutoring by simply adding more kids to a tutoring session — while tempting because it means you can serve more students — is going to quickly lose its efficacy. It just ceases to be personalized and starts to look like group instruction. That’s one common pitfall. 

A second common pitfall is failing to eliminate the barriers to accessing tutoring on a regular basis. If tutoring is moved to after the school day, some kids have other commitments. [If it’s online], some kids may not have the support to troubleshoot technology or may not have the technology itself. The more that we can reduce those access barriers, the more successful schools will be at delivering tutoring at scale and delivering tutoring with the high-dosage frequency that we think is necessary for it to be effective.

I think schools have to be clear-eyed about the trade-offs of different design changes. [They should aim to] scaffold these programs so the program is driving success and tutors can parachute into it and be bolstered by the great infrastructure, with students showing up ready to roll, knowing the routine.

Get stories like these delivered straight to your inbox. Sign up for The 74 Newsletter

Republish This Article

We want our stories to be shared as widely as possible — for free.

Please view The 74's republishing terms.

On The 74 Today