Student struggles in math, which emerged as a major concern after the pandemic, have once again become an acute source of worry in the nation’s school districts.
Earlier this year, the results of the National Assessment of Educational Progress, commonly known as the “nation’s report card,” showed that 4th grade scores for American students in that subject rose only marginally — and remain below 2019 levels.
In 8th grade, test scores in math were flat compared with 2022.
The flagging performance has drawn a wave of responses from policymakers. Many school system leaders have suggested that new, creative strategies in math are needed to address student shortcomings.
About This Analyst
Katey Arrington is the director of systemic transformation at the Charles A. Dana Center at the University of Texas at Austin, overseeing all of the programmatic work of the Center. She is also the President of NCSM, the premier national mathematics education leadership organization. Previously, she served as the associate director of the UTeach Institute working to prepare STEM teachers across the country, served as a leader in state and national initiatives for improvement of mathematics education.
EdWeek Market Brief recently spoke with Katey Arrington, the president of NCSM: Leadership in Mathematics Education, an organization representing math leaders. Arrington is also the director of systemic transformation at the Charles A. Dana Center at the University of Texas at Austin.
Arrington spoke about the wave of concern about student math performance, and the kinds of resources that she believes organizations in the market need to be delivering.
What do you see as the most pressing areas of need in school districts when it comes to math resources?
There are a lot of tools available that are amazing. Curriculum is much better than it used to be. That said, there’s been little investment in working with teachers in effective ways to implement the tools that are out there.
For example, I was at a Council of Chief State School Officers meeting. We were talking about high-quality instructional materials. There’s a new report that shows that high-quality instructional materials — even though many districts have adopted what we consider [to meet that standard] — are not having an impact in math the way they’re having an impact in language arts or social studies.
The reason is that they’re taking these materials that are built to follow research on good instruction, and adjusting them, and building in scaffolds for students, when the whole point is students need to be thinking and asking questions to figure out what’s happening. We’re taking all the power out of the high-quality instructional materials.
New Directions in Math: Key Takeaways
Here are ways that NCSM President Katey Arrington believes math resources need to change to become more responsive to school needs:
1. Making resources branded as “high-quality instructional materials” more useful through ongoing, job-embedded professional development.
2. Putting a new emphasis on videos to model what sound classroom instruction looks like.
3. Creating math assessments that focus on testing a richer array of cognitive demands.
4. Focusing on a “stepping stone” approach to helping math educators implement resources effectively in classrooms.
How then do you believe the promise of “high-quality” instructional materials can be delivered?
I am a strong believer that high-quality instructional materials are necessary, but they’re not sufficient. We need to be thinking about how we provide teachers effective, ongoing job-embedded professional learning.
We need to be focused on effective math teaching practices. They were published in an NCTM book — the effective mathematics teaching practices were published 10 years ago in Principles to Actions — and were republished this year. The National Council of Teachers of Mathematics, NCSM, and Association of State Supervisors of Mathematics worked together on a new high school book that outlines effective teaching practices.
When you look at the universe of math resources out there, what’s missing that could help teaching and learning?
As far as actual tools go, videos of effective teaching, with people using the effective teaching practices, so that they can see it done [would help]. There was a video series — these are outdated now but the format is worth noting — that we had on our Inside Math site a few years ago that was not only the teacher teaching the students, but before the lesson, the teacher saying, “Here’s what I’m trying to do with this lesson.”
Then they see the actual lesson and an interview with the teacher afterward reflecting, “Did I get the results I wanted in the areas that I was focusing on?” That’s incredibly good professional development and a great resource for teachers attempting to change their practice.
What other advances with math resources are needed in supporting schools?
While curriculum has changed significantly, our assessments are still a problem. We know that the kinds of things that are important for students to do are about communicating their thinking and understanding other people’s thinking, connecting it to theirs.
Why are changes to assessments so important?
Those kinds of meaningful cognitive demand tasks are what we need our students to do to have them finish K-12 and go into higher ed with that problem-solving, productive, “I can take on a challenge and figure it out” attitude.
Our assessments are still focused on the things that are easier to measure, which is do you know this multiplication, or can you solve this equation? Work on assessments would be beneficial. It’s also expensive and hard, so I don’t expect most people will be for that suggestion.
Are there examples of new assessment models to follow, in your view?
There are people doing better jobs. Full disclosure: We work with Agile Mind. Dana Center is an author for the Agile Mind curriculum, and they are online and have lots of different interactive kinds of question types on their platform that I happen to be familiar with. Kids can answer in a range of ways because they’re given multiple options for how to answer.
There are some technology-based companies that are doing a better job of those questions that allow for multiple ways of thinking and answering, all those things that we talk about in instruction of, if you’re giving kids a question where there is one answer and really one way to get to that answer, it’s probably not even worth asking that question. That kind of thinking needs to be represented in the assessments, too.
How have you seen academic standards for math change over time, and where do they need to go?
The process standards — in Texas, they’re called process standards. In most places, they’re called the practices — are focused on students’ understanding, engaging in the math, communicating their thinking, showing that thinking using multiple representations, using both the quantitative measures but also the contextual to describe problems.
Now, versus where we were 10 years ago, everyone at least recognizes that as something that we need to talk about.
How close are we to adopting that contextual approach?
I’m pretty sure not everybody’s doing it. But there’s this acknowledgement of yes, if our goal is to have students graduate with all the things that come out in LinkedIn reports and from industry and business leaders of how they need people — they don’t care if they can factor an equation — to be able to figure out problems and talk to each other.
Then we’ve got to be teaching them to do those things.
Why aren’t those focuses on practical, real-world applications of math happen more often?
Admittedly, that’s not easy. A lot of our older teaching force was not focused on that as they were developing themselves as teachers. A lot of our younger teaching force, especially here in Texas, you have a wide variety of preparation for our teachers. It is a complicated thing to do to take a whole classroom of kids and get them involved and engaged.
We’re taking all the power out of the high-quality instructional materials.
How do schools make progress toward achieving that?
The emphasis right now needs to be on implementation. The materials are there, but what sort of stepping stone supports can we provide for them?
For example, we have created something called an innovation configuration … We describe what that might look like if we were living in a dream world and that was happening. And then we talk about, if that was happening, what would that look like?
We describe some steps in between and try to be as descriptive as possible so that teachers can find the description of where they are, see what it looks like for that next step, [and] know how to advance themselves in a reasonable way.
Do you believe technologies such as AI will end up influencing math instruction positively or negatively in the years ahead?
AI’s going to be huge. The cons that will come will [emerge] because we’re trying to avoid or pretend it doesn’t exist instead of embracing it and using the power of it for good.
We talk about things we want like culturally relevant instruction. Maybe you don’t know a lot about African culture. Or there are different data modeling things that we want students to do. AI can help with that. But we must be smart consumers and not just do the things that AI tells us to do.
Join Us for EdWeek Market Brief’s Virtual Forum
Join our virtual forum June 10 & 11, 2025, to hear directly from school district leaders and industry peers about important trends playing out in the sector—and the support school systems need from education companies.
How do we ensure that schools become smart AI consumers?
We’re working on putting out guidance for math leaders about how they can and should be working with AI themselves and teaching their educators to use AI. We want to be careful because if they have instructional materials they’re supposed to be using from their school, they should not be making things up. You lose a lot of coherence.
If you’re looking for another situation to model within a curriculum with a different context or something that’s locally connected, there’s a ton of technology, including AI, but not limited to it, that could be leveraged for that. We are working hard on how to provide good guidance for math leaders. And our curriculum vendors could be helpful in that realm.
The “science of reading,” or an evidence-based approach to curriculum and instruction, has taken hold in that subject. Some policymakers speak of a need to pursue a science of math.” What do you make of this idea?
My thoughts on math education — and [from] doing a lot of study of what the research says about math education — is we need a well-rounded approach to making sure that kids get what they need. My understanding of the science of math is that they are committed to a style of instruction, which is more direct teach and explicit instruction, where we say to students, this is how you do things, and then have them repeat and model.
I do think there’s a place for that, but I think very enthusiastically, that cannot be all that we’re doing for our students. We will be much better off if our goal is to produce problem-solvers and students who can find a problem in the real world and know some strategies to try and figure out what to do and how to do it, dig into that problem for themselves. And have some productive struggle and move themselves forward. We need to be giving them those kinds of opportunities.
What is the risk if policymakers do not encourage that approach in math?
If we limit ourselves to only direct and explicit instruction, we’re going to be doing students a disservice. What they’re advocating for is not necessarily bad. It just is limited to one type of instruction when the picture should be much broader. We will be happier with our outcomes if we keep our focus on providing different types of instruction.