PAUL PETERSON, HOST:
This is the Education Exchange. I am Paul Peterson, Director of the Program on Education Policy and Governance at Harvard University. Thank you for joining us. 2025 student performance in math and reading at ages 9 and 13 has just been released by the National Assessment of Educational Progress, the nation’s report card. It’s known as NAEP for short, and we’ll call it by that short name throughout this conversation. Younger children are doing better than three years ago, the report tells us, but older children are not. That’s it in a nutshell. But to discuss the findings and to probe deeper into them, I have with me on the Education Exchange, Marty West, the vice chair of the board responsible for overseeing the survey. He’s also the academic dean at the Harvard Graduate School of Education. Marty, thank you for joining our listeners on the Education Exchange.
MARTY WEST: It’s great to be with you, Paul.
PETERSON: So, Marty, my first question for you is, what’s the biggest surprise in this? What did you find the biggest surprise in the data that you looked at?
WEST: I think I was most surprised by the good news that has been hard to find in recent years, looking at trends in performance for American students, as measured both by the NAEP program and by other assessments. But here we did see clear evidence of progress among nine-year-olds, gains of four points in both math and reading. And I was particularly surprised by the improvements that we saw for nine-year-olds in reading. If you’ll recall, we had the main NAEP come out a year ago, which tested 4th graders in 2024. And when you compare their performance to 2022, you saw continued deterioration in their reading skills. So, I was surprised to see evidence of progress for this similarly defined cohort of students when they were tested in 2025. So, I was surprised by the good news. I think we need to give equal attention to the less positive news, which was that the scores for 13-year-olds were absolutely flat compared to the last time they were tested in 2023 and are still well behind where they were just prior to the pandemic and even further behind where they were in 2012. So, there’s certainly some hard news in these results as well, but I was surprised by the good.
PETERSON: Well, trying to figure out the complexity, unpack the complexity of these results, good news and bad news, strikingly different results for those at age 9 and age 13. When I looked at this, I said, well, those at age 9 really weren’t in school when Covid hit back in 2020. At least that’s the way I interpret it, that they may have been in kindergarten, but they were not, you know, in1st grade, they were not being instructed. Whereas those at age 13, they were in, you know, grade 4 or whatever at that time. And that was just when you needed to, you know, hone those skills and those skills become extremely important. So, isn’t this a Covid story here? Covid was bad, but if you manage to avoid going to school during Covid, you’re going to be fine.
WEST: I think that may be part of the explanation. You’re correct that a bit more than half of the nine-year-olds would have been in kindergarten in March 2024, but a bit less than half wouldn’t even have entered kindergarten at that time. And so presumably the disruption was less for them than for the 13-year-olds, most of whom would have been in grade 4 at the time of the school closures. And so, I think that certainly could be part of the story. But we have to keep in mind that when we’re thinking about recent trends in American education, that the story can’t be only about Covid. Really, it’s about what’s been happening for more than a decade. 2012 is when we saw results peak on the long-term trend assessment that we’re discussing the latest version of today. On the main NAEP, results peaked in 2013 or 2015, depending on the grade or the subject. So really, we need to be looking beyond the pandemic when we’re trying to understand trends and achievement.
PETERSON: Well, nonetheless, I think the improvement is compared to where they were back in 2019, just before Covid hit. They’re not back, those younger children, in my reading of it, correct me if I’m wrong, are not back to the levels back in 2015. All they have done is recover what was lost during Covid.
WEST: Yeah, and they haven’t even recovered all of that in math. So, if you compare nine-year-olds tested in 2025 to the last time they were tested in 2022, it is the case that in reading, the progress that they demonstrated in this assessment was enough to bring them back to where they were on the eve of the pandemic. For the average student, it also is comparable to where they were in 2012. But in math, they are still well behind where they were in 2020 and even further behind where they were in 2012.
PETERSON: So, I sort of see this as the die cast for those who were in school during the Covid years. A generation has been lost. Those students who were hit by Covid for a full year or something like that, it’s going to be very difficult for them to recover that loss.
WEST: Well, that’s certainly the case for the millions of students who experienced the disruption of the pandemic and have since moved out of the K-12 system. I’m not sure if we need to be quite so fatalistic when it comes to 13-year-olds who are still enrolled at this time; they would be roughly entering high school. We do have some evidence that strategies to help students recover lost ground have been productive. They haven’t been pursued at anywhere near the scale that they would need to be in order to offset the lost ground during the pandemic. But again, I think this sort of obsession with thinking about the pandemic learning loss obscures the fact that students lost just as much ground in the years leading up to the pandemic as they did during the pandemic. And so, we need to be thinking about strategies that will improve the quality of education delivered by our system, regardless of the disruption of the pandemic.
PETERSON: Right. Well, the younger kids probably are not so affected by the cell phone. So, if you’re talking about children who are…now being tested at age nine, they’re in 4th grade or something like that. We’ve seen from other data about cell phone usage that that’s having an impact on kids at school, but not so much in the elementary schools because cell phone usage is not as widespread at that age.
WEST: Yeah, I think that’s a good observation for those of us like me who have tried to call attention to the potential role of social media and cell phones and the increasingly digital childhood that students are experiencing. The fact that performance has been down for nine-year-olds, for 4th graders, I think does call that into question a little bit. I don’t think though it rules out the potential role of screens and their use in schools, which I think is quite widespread even in elementary school and could be part of the challenge.
PETERSON: Well, let’s break it out by categories. Do you see any regional differences? Is there a difference between the South and the rest of the country? Everybody’s talking about the Southern miracle or the Mississippi miracle. Do you see anything that suggests that we have any regional differences?
WEST: Well, one of the challenges with the long-term trend assessment is that the sample is much smaller since it’s designed to provide data for the nation only than is the case for the main NAEP that’s given every two years and provides information on the nation but also state by state. So, you only have about 8,000 students who are tested in each age and subject combination in the long-term trend. And that means that some of the subgroup comparisons that we often make with the NAEP are challenging to make with much precision. As I looked at the data region by region, I didn’t find much in the way of different patterns. But again, I’m not sure that this is the best source to look for that type of variation.
PETERSON: Well, how about the ethnic group variation? I’m sure you’re going to tell me the same thing there, but do you see anything? Did you see any disproportionate improvement among Black students or Hispanic students relative to white students, how are the Asian students doing?
WEST: Well, the data on race and ethnicity is actually a great illustration of the point that I was just trying to make about the limited use of the long-term trend assessment for subgroup comparisons. If you look at the data that NCES reports from the latest assessment you actually see for each of the major racial and ethnic groups, you see essentially numerically the same pattern. They’ve all made improvements for nine-year-olds since 2022. But none of those changes are statistically significant. So, we would actually say we didn’t see statistically significant evidence for white students, for Black students, for Hispanic students, for Asian students. Yet we know that when you take all of the groups together and talk about the nation as a whole, we do see, substantial and statistically significant progress. So, I think it’s a good illustration. I will say the subgroup pattern that I think is most useful and most interesting to look at with the long-term trend. I guess there are two, one is that we often like to look at the performance of higher achieving and lower achieving students. So, students, for example, at the 90th percentile, at the 50th percentile, and at the 10th percentile, the very bottom of the achievement distribution. And what was interesting about the gains for nine-year-olds that we saw in 2025 is that they were driven by larger gains for low-achieving students. And that is a reversal of one of the more troubling trends in recent NAEP assessments, which is that the declines in achievement we’ve seen over the past decade have been largest for lower-achieving students. So, to the extent that this reflects a reversal of that trend, I think it is potentially quite noteworthy.
PETERSON: I agree with you, Marty. I think that is noteworthy, but I am going to come back to the Covid thing because it was the disadvantaged students that were harmed the most by Covid. So, to see the greatest improvement in there makes me want to harp on the Covid story a little bit again.
WEST: It was also the low-achieving students who were… most affected by the declines in achievement prior to Covid. So, you know, whatever the cause of those declines was felt disproportionately by lower-achieving students as well. So that’s a long-term trend at this point that I think it’s quite noteworthy to see turn around. The other subgroup difference that I was going to call listeners’ attention to is actually the difference between girls and boys. And there’s been a lot of attention in recent years to the academic struggles of boys. But when you look at performance trends through and in the aftermath of the pandemic, you actually see girls experience larger declines and their scores have been slower to recover. And I don’t think we have a great explanation for that, but that is something that we’ve seen.
PETERSON: Well, except for the fact that women or young women, children, girls have suffered the most as a result of this dislocation, social and emotional dislocation. At least in my observation, this has had a bigger impact on females than males. I don’t know if you think that is the case or not.
WEST: I’m not sure if I know how to assess that for nine-year-olds, say. I think it’s certainly the case that for older students, there’s some evidence to suggest that girls have faced some pretty severe challenges, both from social media and from the isolation of the pandemic that may be part of the story. But we’re seeing this really across assessments by age in a way that I don’t fully understand.
PETERSON: Well, how about the gap between Catholic and public-school students? Did that widen in this? Are Catholic schools doing better relative to the public schools?
WEST: Those gains for nine-year-olds that I mentioned weren’t evident if you break out Catholic schools separately. And so, in a sense, you see the gains for public-school students, but not for Catholic-school students. Again, though, I don’t think we have adequate sample sizes to really call attention to that result. If you look at the data on Catholic schools over the course of the pandemic you just see much more stable performance. They lost less ground, and they’ve been sort of not recovering as rapidly, at least as judged by the long-term trend, but again, they’re not recovering as much because they didn’t lose much ground in the first place.
PETERSON: Well, to change the subject slightly, the National Center for Education Statistics says that enrollments, and this just came out today, so that’s something that I just took special notice of, it says that enrollments have dropped by 2.8% or 1.4 million students compared with the fall of 2019. So, this is the first official National Center for Education Statistics data that we have on what’s been happening to student enrollments in our public schools since 2019. What’s your assessment of that statement?
WEST: Well, I’d want to know a little more than what you’ve shared right now. And in particular, I’d be interested in how much of that reflects just some demographic changes we’ve seen in terms of the slowing of the birth rate. But it’s clearly the case that we’ve seen—
PETERSON: Well, that’s the dominant factor. Yeah, birth rate. The fertility rate has declined in the period. And these kids are now entering school. So that’s a 5-year delay between the decline in fertility. But it’s now impacting, yes.
WEST: Yeah, and I think another part of that story is that there’s been a shift outside of the traditional public sector into the homeschooling and the micro-schooling world. And that’s something that will affect our ability to measure progress with assessments like NAEP that don’t include home-schooled families. And that’s something that the program is thinking about and needs to be thinking about going forward.
PETERSON: Well, you’ve been mentioning the difference between the main NAEP and the long-term trend NAEP. And we haven’t really gone into that. But I think some of our listeners would like to know what in the heck this is all about. Are there two different types of NAEP surveys? And why in the world did this National Assessment of Education progress…and you’re the vice president of the governing boards. I’m going to hold you accountable for this. Why do you have two different NAEPs?
WEST: So, the long-term trend and what we refer to as the main NAEP differ in several key respects. The most fundamental is that the long-term trend, as much as possible, has tried to keep the assessment and the conditions under which students take the assessment, as closely matched as possible to exactly how things were done when the assessment was first launched in 1971 in reading and 1973 in math. Whereas the main NAEP is updated more regularly in order to try to better match the conditions students experience in schools. What does that look like? It means, for example, that the long-term trend is still a paper-based assessment where students are bubbling in the answers to items, whereas the main NAEP is now delivered entirely digitally. So, students are taking them on devices. It means that, to the maximum extent possible. The actual items on the long-term trend NAEP are exactly identical to how they were written back in the 1970s. There’s been some situations where that’s not exactly impossible because they become factually incorrect in how they refer to the world. But that’s the basic principle. Whereas the main NAEP, again, is regularly updated to make sure that it corresponds with how students are engaging with content in their schools. I mentioned that the long-term trend is age-based, tests 9, 13, and sometimes 17-year-olds, whereas the main NAEP is grade-based, looking at grades 4, 8, and 12. And then I also mentioned that the long-term trend is conducted for the nation only rather than also providing information on the performance of each state. And then finally, the frequency of the assessments differs. The main NAEP, Congress mandates that we administered every two years. Congress also mandates that we administer the long-term trend assessment program, but it allows the National Assessment Governing Board on which I serve to determine the frequency of the assessment. And that typically is less frequent than the every two-year cadence of the main NAEP.
PETERSON: Okay. Okay. So that’s all very precise. And that’s very valuable information. But in the big picture, aren’t they just the same thing? I mean, more detail from the one than the other, but aren’t they still measuring trends over time?
WEST: They are, and they’re using two really different philosophies about what’s the best way to measure trend and to make sure that scores are comparable over time. Is that best done by sort of locking the assessment in amber and carrying it forward, even though the world has changed around it? Or is that best done by, again, regularly updating the assessment to make sure that it matches what students have been exposed to? And arguments can be made for both of those approaches. The long-term trend because the main NAEP was only started in the 1990s also allows us to look at performance over a substantially longer period of time. For example, we can look with these latest results and see that despite the recent declines I’ve been talking about, that 13-year-olds, at least are performing substantially higher than they did back in the 1970s. And, you know, so that’s a useful affordance that the long-term trend provides that the main NAEP at this point does not.
PETERSON: So which do you believe? If you were going to get rid of one of these things, say there was a person in the White House by the name of Donald Trump, just for example, who said you got to get one of these because two of the same thing is too much. He isn’t going to be persuaded by your subtlety. So which one do you get rid of?
WEST: I want to say that first, that I have zero evidence to suggest that anyone in the White House or elsewhere in the current administration is thinking about that type of choice or forcing it on the program. I will say that for any short period of time, I think the best way to be able to preserve trend or maintain trend as we talk about it, which is so vital to the National Assessment of Educational Progress, it’s what allows us to speak about progress or the lack thereof. We want to be able to compare scores from these assessments given in different years. If you want to do that over any short period of time, say a decade; the best way is to follow the philosophy of the long-term trend, which is to keep everything exactly the same. I think once you start looking over much longer periods of time multiple decades you then begin to worry about whether that approach may not be the most reliable and may have some weaknesses that the approach that the main NAEP takes is designed to address. And so, I guess if you forced me to only pick one program rather than the other and maintain it over time, I would go with the main NAEP, but as a researcher as someone who enjoys looking at data, I certainly appreciate the opportunity to look at both of them.
PETERSON: Well, what’s going to happen going forward? Because there’s a lot of pressure on the Department of Education to make cuts in funding, and that’s probably not going to go away. What do you see as the near-term future of both of these surveys, the long-term trend and the main NAEP?
WEST: Well, the short answer to that, Paul, is that both of the assessments, as I mentioned, are mandated by Congress. And so the National Assessment Governing Board doesn’t have the luxury of thinking about them as an option. They are requirements and that’s the way we treat them. I think there have been longstanding conversations about the question that you raised. Do we really want and need to have these two different assessment programs running side by side and those conversations long predated the current presidential administration and they’ve continued during it, conversations among the governing board but if we were to make a recommendation in that area it would need to be a recommendation to Congress. It’s not something that we have the ability to decide on our own.
PETERSON: Well, Marty, I asked you a lot of questions, but I may have missed something. You want to have any final observations?
WEST: I think my final observation, I started with the good news for nine-year-olds, because I think that is what is most striking in these results against the backdrop of what we’ve seen recently. But I think we need to place as much if not more emphasis on the results for 13-year-olds who have flatlined over the past two years and are substantially behind where they were on the eve of the pandemic and even further behind where they were in 2012. And I think we need to act on those results by not considering them a lost cause, as you might have suggested, but realizing that we do have both time and the opportunity and proven strategies to help them get back on track.
PETERSON: Well, thank you, Marty. Thank you for joining us on the Education Exchange.
WEST: It’s been a pleasure. Thanks, Paul.
PETERSON: I’ve been speaking with Martin West. He’s the dean of the Harvard Graduate School of Education. He’s also the vice chair of the National Governing Board for the National Assessment of Educational Progress. And I think you can tell from his responses today just how valuable it is to have somebody as knowledgeable and as authoritative as he serving on such a key position at a time when the…testing regime in the United States is undergoing a lot of scrutiny. So, thank you all for joining me. And thank you, Marty, for joining me on the Education Exchange. I am Paul Peterson. Please join me every Monday when another Education Exchange podcast is released on the Education Next website at noon Eastern time.
