For Rachael Garnick, a former first grade teacher, the results were a reminder of how tough it’s been for schools to recover from historic declines in learning since the pandemic. 

鈥淭he literacy scores are still abysmal and we should be displeased,鈥� said Garnick, who heads the Pennsylvania Literacy Coalition. Made up of over 70 organizations, the group has pushed and state officials to fund and implement reading reform.

But despite the discouraging statewide results, she also sees districts, like in northeastern Pennsylvania and the Mohawk Area district, northwest of Pittsburgh, 鈥渢rending in the right direction,鈥� and demonstrating urgency over reading scores. Their attitude, she said, was 鈥渢he opposite of 鈥業f it ain’t broke, don’t fix it.鈥� Instead, 鈥業t鈥檚 broke; we鈥檝e got to fix it.鈥� 鈥� 

on pandemic learning loss from NWEA, an assessment company, captured that combination of frustration and hope over the state of academic recovery. About a third of schools have reached pre-COVID performance levels in reading or math, and just 14% have recovered in both subjects. But even some that were hit the hardest, like high-poverty schools, have made impressive gains.

The report was just the latest collection of results pointing to a long road ahead for most schools. Last year鈥檚 National Assessment of Educational Progress scores showed students in the majority of states losing more ground, but included a few standouts with strong progress, like Louisiana in reading and Alabama in math. And state test scores tell a similar story: few have topped pre-COVID performance.

It鈥檚 not like experts didn鈥檛 predict a slow recovery. 

鈥淚f student performance improvement follows historical prepandemic trends, it could take decades for students to fully catch up,鈥� researchers with McKinsey and Company, a consulting firm, .

Even the nation鈥檚 education chief isn鈥檛 expecting good news soon. 

鈥淚 would like to say that NAEP scores, when they come out again in January 2027, are going to show marked improvement,鈥� Education Secretary Linda McMahon said in a recent K-12 Dive . 鈥淚 don鈥檛 think they are.鈥�

But Dan Goldhaber, director of the Center for Analysis of Longitudinal Data in Education Research, said it鈥檚 important to put NWEA data, and all measures of kids鈥� learning, in context.

鈥淥ne of the reasons that we’re not seeing recovery and that the results aren’t better is because of what was happening in the decade ,鈥� he said. 鈥淭here was a slow degradation of academic achievement.鈥�

Resisters and rebounders

Schools that were able to resist further declines during the pandemic are those that are more likely to be back on track, according to NWEA鈥檚 data, which represents five million students who took the MAP Growth tests through fall 2024. Such schools make up nearly three quarters of the recovered schools.

The Los Angeles-area is one example. 

Before the pandemic, the high-poverty, majority Latino district was already seeing gains on state assessments. When testing resumed in 2022, reading scores held steady. Math scores caught up the following year, and the district has continued to post gains ever since. 

Superintendent Darin Brawley highlighted a mix of academic routines, like a math problem of the day, weekly quizzes and challenging writing assignments, that the district continued despite the disruption of school closures. Teachers were encouraged to dial back their use of smart boards in the classroom and require students to keep math and language arts journals to improve retention. 

鈥淓verything was being done on the smart board and kids weren’t notating anything,鈥� Brawley said. 鈥淐ertain things have to be worked out on paper.鈥�

NWEA data also pointed to what the researchers call 鈥渞ebounder鈥� schools, those that saw significant drops in achievement but have been able to climb their way back. High-poverty schools are among those with impressive gains, but even districts seeing higher-than-ever performance still struggle to close wide achievement gaps.

鈥淲e’ve never had scores this high in English language arts or math,鈥� said Buffy Roberts, associate superintendent of the Charleston County schools in South Carolina. 鈥淚t’s been quite phenomenal.鈥�

She was talking about , which, unlike NWEA and NAEP, aren鈥檛 comparable because states don鈥檛 all measure proficiency the same. But they can still reflect post-COVID trends if states haven鈥檛 changed their tests since 2019. 

South Carolina鈥檚 math test has remained constant. Results show that statewide, scores have nearly recovered. It鈥檚 a trend that NWEA noted as well, explaining that while schools 鈥渓ost significant ground,鈥� in math, many made 鈥渟ubstantial gains afterward.鈥�

In Charleston, 54% of students in grades three through eight met or exceeded expectations in math last year, up from 48% in 2019 and about 10 percentage points higher than the state average. The district also made the Harvard Center for Education Policy Research鈥檚 fully recovered districts in the nation last year.

Roberts pointed to a swift return to in-person instruction and high-dosage tutoring as some of the factors contributing to strong growth. But she said at the outset of the pandemic, leaders 鈥渒new there were some vulnerable groups鈥� that would need 鈥渟tructures and support to mitigate some of that learning loss.鈥�

The district鈥檚 , she explained, provided extra dollars to schools with high-poverty students even when the schools didn鈥檛 qualify for federal Title I funding. The schools used the funds for extra staff to reduce class sizes, incentives to increase attendance and mental health services.

But there鈥檚 still a lot of work to do. In fourth grade math, there鈥檚 a more than 50 percentage point gap between white and Black students, and students from wealthier families outscore students in poverty by 39 percentage points. 

鈥淲e agree that progress must be faster,鈥� the district on Facebook after a conservative community group to the disparities. 

In an analysis of scores, Education Data Center researchers, led by Brown University鈥檚 Emily Oster, were hopeful about continued math recovery in 2026. Of the 32 states that have kept the same math test since before COVID, seven met or exceeded 2019 proficiency rates: Colorado, Georgia, Iowa, Mississippi, Missouri, Rhode Island and Tennessee.

But even if they didn鈥檛, they all made some gains. Despite Pennsylvania鈥檚 decline in reading, for example, its performance in math is less than a percentage point from reaching the 2019 level. 

But the results in reading were less encouraging. Six out of 28 states have met or surpassed pre-pandemic performance. But several others, like Massachusetts, Minnesota and Oregon, remain well off that mark. 

Goldhaber, with CALDER, suggested that states haven鈥檛 seen improvement on tests because parents trust those scores less than the grades kids bring home on report cards and assignments. 

A recent reiterated that point. In a survey of over 2,000 parents, nearly three quarters said they believe grades more than tests when making decisions about their children鈥檚 learning. They鈥檙e also less likely to take action, like seeking out tutoring or other help for their child, when grades are good. 

The problem is that because of grade inflation, which was on the rise even before the pandemic, grades are a less accurate measure of how students are really doing. 

The results of that survey were no surprise to Bibb Hubbard, founder and CEO of Learning Heroes, a nonprofit that focuses on helping parents understand achievement data. She said she鈥檚 been 鈥渟creaming from the rooftops for 10 years鈥� that parents are about their kids鈥� performance. 

鈥淕ood grades do not equal grade level,鈥� she said. 鈥淧arents are deeply engaged, but we can鈥檛 afford to leave them on the sidelines relying on grades alone. The stakes are too high.鈥�

To make his case, Pondiscio relies on the skill level needed to achieve a proficient score or better on National Assessment of Educational Progress, a level that only 30% of tested students reached on 2024鈥檚 Grade 8 reading exam. Only 16% of Black students and 19% of Hispanics were proficient or more.

Yet naysayers argue that the NAEP standard is simply set too high and that NAEP鈥檚 sobering messages are inaccurate. There is no crisis, according to these naysayers.

So, who is right?

Well, of eighth graders from Kentucky indicates that it鈥檚 Pondiscio, not the naysayers, who has the right message about the NAEP proficiency score. And, Kentucky鈥檚 data show this holds true not just for NAEP reading, but for NAEP math, as well.

Kentucky offered a unique study opportunity. Starting in 2006, the Bluegrass State began testing all students in several grades with exams developed by the ACT, Inc. These tests include the ACT college entrance exam, which was administered to all 11^th grade public school students, and the EXPLORE test, which was given to all of Kentucky鈥檚 public school eighth graders.

Both the ACT and EXPLORE featured something unusual: 鈥淩eadiness Benchmark鈥� scores which ACT, Inc. developed by comparing its test scores to actual college freshman grades years later. Students reaching the benchmark scores for reading or math had at least a 75% chance to later earn a 鈥淐鈥� or better in related college freshman courses.

So, how did the comparisons between Kentucky鈥檚 benchmark score performance and the NAEP work out?

 Analysis found close agreement between the NAEP proficiency rates and the share of the same cohorts of students reaching EXPLORE鈥檚 readiness benchmarks. 鈥�

For example, in Grade 8 reading, EXPLORE benchmark performance and NAEP proficiency rates for the same cohorts of students never varied by more than four percentage points for testing in 2008-09, 2010-11, 2012-13 or 2014-15.

The same, close agreement was found in the comparison of NAEP grade 8 math proficiency rates to the EXPLORE math benchmark percentages. 

EXPLORE to NAEP results were also examined separately for white, Black and learning-disabled students. Regardless of the student group, the EXPLORE鈥檚 readiness benchmark percentages and NAEP鈥檚 proficient or above statistics agreed closely.

Doing an analysis with Kentucky鈥檚 ACT college entrance results test was a bit more challenging because NAEP doesn鈥檛 provide state test data for high school grades. However, it is possible to compare each student cohort鈥檚 Grade 8 NAEP performance to that cohort鈥檚 ACT benchmark score results posted four years later when they graduated from high school. Data for graduating classes in 2017, 2019 and 2021 uniformly show close agreement for overall average scores, as well as for separate student group scores.

It鈥檚 worth noting that all NAEP scores have statistical sampling errors. After those plus and minus errors are considered, the agreements between the NAEP and the EXPLORE and ACT test results look even better.

The bottom line is: Close agreement between NAEP proficiency rates and ACT benchmark score results for Kentucky suggests that NAEP proficiency levels are highly relevant indicators of critical educational performance. 鈥婽hose claiming NAEP鈥檚 proficiency standard is set too high are incorrect.

That leaves us with the realization that overall performance of public school students in Kentucky and nationwide is very concerning. Many students do not have the reading and math skills needed to navigate modern life. Instead of simply rejecting the troubling results of the latest round of NAEP, education leaders need to double down on building key skills among all students.

But that鈥檚 not what鈥檚 happening. Instead, across a range of tests, grade levels and subject areas, the scores of the lowest-performing students have fallen dramatically, while the scores of the highest-performing students have been flat or close to it.

The next assumption might be that these declines must be tied to long-running achievement gaps. Indeed, it is and remains true that white and Asian students tend to do better than their Black and Latino peers, and children with disabilities and those who are not native English speakers tend to do worse than average.

But over the last decade, declines have not been about specific student groups. Instead, the story is largely about a growing divergence between higher- and lower-performing kids.

To see this visually, consider the latest in 12^th grade math. As the chart above shows, scores were rising from 2005 to 2013. Not only that, but they were rising the fastest for the lowest-performing students, those in the bottom 10%.

But scores peaked in 2013 and then began falling, especially for the lowest-performing kids. While the performance of the top 12^th graders didn鈥檛 fall that much, even during COVID, scores for the lowest-performing students had fallen 6 points by 2019 and have plummeted another 5 points since then.

Looking at the graph, it’s clear there is a cascading effect down the performance spectrum. Students in the middle have lost far more ground than those at the very top, and kids at the very bottom have seen even steeper declines.

What鈥檚 more, these patterns in test after test. It鈥檚 worth unpacking the data a bit further to understand how these trends are playing out across groups. The first table below looks at the changes from 2013 to 2024 for the top and bottom 10% of students across racial and ethnic categories in 12^th grade math. 

Change in 12^th grade NAEP Math Scores By Student Race/Ethnicity and Performance Level (2013-2024)

The scores for Black and Hispanic students fell across the performance spectrum. But that鈥檚 not the main story here, because the bottom was falling so much faster than the top across all racial and ethnic groups. In fact, the lowest-performing white students showed the biggest declines.
Sorting the data by income, the same pattern holds:

Change in 12^th grade NAEP Math Scores By Student Income and Performance Level (2013-2024)

Again, low-income students scored worse than higher-income students. But the bottom 10% percent of students who do not qualify as low-income suffered the biggest slide.

Tim Daly has documented according to parental education levels for eighth grade math, and that holds for 12^th grade scores as well. Even among students whose parents graduated from college, the lowest performers suffered particularly large declines:

Change in 12^th grade NAEP Math Scores By Parental Education and Performance Level (2013-2024)

In 12^th grade math, at least, the scores of students with disabilities and English learners have actually held up pretty well. Performance rose 0.4 points for students with disabilities and 2.5 points for English learners. Neither group saw declines among the lowest performers.

But consider what happened to students without disabilities (scores for the bottom 10% fell 11.9 points) and native English speakers (the bottom plummeted by 10.1 points).

What鈥檚 behind these trends? The best evidence to a of in-school and out-of-school factors including screen time, instructional shifts that de-emphasize mastery of basic skills and the lack of school-level accountability for results. It may sound counterintuitive, but policymakers looking to raise the ceiling on student achievement should start by making sure they raise the floor.

Disclosures: The Future of High School Network and 社区黑料 both receive financial support from the Carnegie Corporation of New York, XQ and the Walton Family Foundation.

Leadership changed hands in 114 of those districts 鈥� 23% 鈥� within the past year, a jump from 20% the year before, according to data, shared exclusively with 社区黑料, by the  from ILO Group, a consulting firm. The project 鈥� the only current publicly available resource on leadership turnover in the 500 largest districts 鈥� listed about 15% of districts replacing their superintendents prior to the pandemic.

One surprise outcome of that turnover is an increase in female superintendents: Women now represent a third of district chiefs, up from 30% last year. Of the 114 new chiefs, 44 were women.

But even with those gains, it would take another 30 years for women to reach parity with men in district leadership, the authors said.

To Julia Rafal-Baer,  CEO of ILO Group, this year鈥檚 results offer a mixed picture, coming just days after the latest scores from the National Assessment of Educational Progress. The results showed declines in reading for 12th grade girls and in science for all 8th graders.

鈥淭here is a continued destabilizing of leaders at a time when we really need to have a coherent agenda that is driving instruction,鈥� said Rafal-Baer, also a member of the National Assessment Governing Board, which oversees the NAEP program.

Even so, she was pleased with the gains for women.  鈥淚鈥檓 encouraged by the fact that we鈥檙e starting to see some meaningful progress.鈥� 

Superintendent turnover happens for myriad reasons 鈥� from stagnant student performance to disagreements over salary. But it鈥檚 clear that COVID and the cultural debates that followed 鈥� embroiling districts in disputes over mask mandates, 鈥渁nti-racist鈥� curriculum and sexually explicit books 鈥�  transformed the nature of the position. 

鈥淚t’s always been political, but it’s never been so partisan,鈥� said Gustavo Balderas, superintendent of the Beaverton School District in Oregon. Since 2011, he has led five districts in the Pacific Northwest and will leave next year to become of the Puget Sound Educational Service District, a regional agency in Washington.

On top of local concerns, today鈥檚 superintendents have the added weight of responding to threats of funding cuts and policy shifts from Washington, Balderas said. 鈥淚 was just visiting a school 鈥� that had a family deported.鈥�

鈥榃orn out鈥�

Researchers who focus on the superintendency and school board politics echoed Balderas鈥� view. Rebecca Jacobsen, an education policy professor at Michigan State University, said district leaders are 鈥渨orn out.鈥�

鈥淚 think that the toll of the past few years continues to ripple and really push people out,鈥� she said. Several faced personal attacks, including , from angry members of their communities. 鈥淔or many who entered education 15-20 years ago, this is not the landscape that one envisioned.鈥�

The skills superintendents bring to the position sometimes don鈥檛 match the demands of the job, added Rachel White, an associate professor at the University of Texas at Austin and founder of the Superintendent Lab, a source of research and data on district leaders. Most were teachers and principals before moving to the central office and spent years overseeing instruction, finance or teacher development. 

Now they鈥檙e responding to social media, and the 鈥減roliferation of misinformation and disinformation campaigns often rooted in ideology,鈥� White said. 鈥淭his has shifted what superintendents are increasingly spending their time on 鈥� debunking stories being told about what is happening in their schools and classrooms that simply are not true.鈥�

As was the case in 2022, some of the turnover is due to school boards firing superintendents before their contracts expire. Since January, the in Georgia, the district in Tennessee and the district in Florida have fired their chiefs.

Most leaders, however, leave on their , sometimes because they鈥檙e seeking a new challenge.

Mary Elizabeth Davis spent nearly seven years as superintendent of Georgia鈥檚 Henry County Schools, overseeing the suburban-Atlanta district during a period of growth in both and . She eliminated a $12 million deficit and built teams to support instruction, facility planning and operations.

Last year, she started over in Cherokee County, another metro Atlanta district, where she aims to keep board meetings more focused on core academic issues rather than  just building projects and the budget. They still need to keep the public informed about finances, but 鈥渋t is no longer the only thing,鈥� she said. 

Inline photo

In Davis鈥� part of the country, women are still the least likely to be superintendents, the data shows. Twenty-two percent of chiefs in the Southeast are women, while the Northeast has the highest percentage of female district leaders 鈥� 46%, or 17 of the 37 districts on the list. 

This year鈥檚 report also delves into the routes leaders take to the top job. 

In 2018, when ILO began collecting the data, fewer than half of superintendents were internal hires. Last year, the majority, 58%, were hired from within, and about 40% had served as an interim superintendent in their district before the board officially gave them the job. Moving up within the same school district is slightly more common for women than men, 55% compared to 50%. 

Over a 20-year period, Cliff Jones worked his way up from teacher to of the Fulton County Schools in Atlanta. Once he entered the central office, he said he 鈥渢ook notes鈥� during a time of leadership turnover and learned the importance of communication in making relationships work with the board.

鈥淭he more successful superintendents that I saw were trying to be out in front, trying to create proactive communication,鈥� he said.  

Newly hired as the superintendent in Horry County, South Carolina 鈥� with an unusually large 12-member board 鈥� he has work to do. He said he doesn鈥檛 want to just be a 鈥�911 guy,鈥� contacting members when there鈥檚 an emergency. 

鈥楶riorities and values鈥�

Not all candidates spend that much time in a deputy or other cabinet position, which Balderas said is likely one reason why turnover remains high. He calls it 鈥渓eadership compression.鈥�

Among the 500 districts in ILO鈥檚 analysis, 10 male leaders skipped straight from principal to superintendent. They include , named interim superintendent of Texas鈥� Conroe Independent School District in May, and , who took over in February as acting chief of the South Bend Community School Corporation in Indiana.

鈥淧eople are just bypassing roles鈥� instead of serving four to six years in a mid-level role where they might tackle some of the same challenges as the superintendent, Balderas said. Maybe, they鈥檙e 鈥渓ess prepared to understand the political navigation that鈥檚 needed鈥� to stay in the position long enough to make lasting improvements. 

After leading multiple districts, Balderas said it鈥檚 possible to work with a politically divided board. He tried to build connections with members by taking on other responsibilities in the community outside of the education sector, from the local chamber of commerce to the Rotary club.

People active in those groups 鈥渟ee that you care about your community,鈥� he said. That word 鈥済ets back to your board in one way or another.鈥�

Despite division among board members, districts can stay focused on academic improvement, said Davis, who was hired in Cherokee on 4-3 vote. 

鈥淚 think that when you start from that position, you have a lot of work to do to understand the priorities and values of individuals,鈥� she said. She met with each board member, hearing concerns over teachers spending their own money on supplies and a desire for more presentations on student data. 

Having a divided board was familiar for Davis. The Henry County board hired her on a 3-2 vote. During her tenure, public meetings turned into over a mask mandate and a in 2023 that kept students locked out of the internet for nearly a month. 

But she had plenty to celebrate. The majority-Black district saw enough to come off the state鈥檚 list of failing school systems and a 9% increase in students scoring at the proficient level or above in .

鈥淚’ve never seen harmony as a requirement for effectiveness,鈥� Davis said. 

U.S. eighth graders are less prepared to be the scientists of tomorrow than they were before the pandemic. 

In the first nationwide test of students鈥� science knowledge since 2019, the percentage of students scoring at the proficient level fell to 29%, down from 33%, and the average score dropped back to levels last seen in 2009, when a new version of the test was introduced, according to the National Assessment of Educational Progress.

Students鈥� confidence in the subject area has also slipped, with 28% saying they 鈥渄efinitely can do various science-related activities,鈥� down from 34%. 

Performance fell across all three categories 鈥� physical, life, and earth and space sciences. Less than half of students can identify the major component of living cells, compared to 55% in 2019, and the percentage of students who can identify a characteristic of mammals declined from 72% to 68%. 

It鈥檚 not just the decline in skills that concerns science experts, it鈥檚 the dramatic decrease in their interest. The share of students saying they enjoy science activities plummeted from 52% to 42%. 

鈥淚f you’re not interested, it’s hard to learn,鈥� said Christine Cunningham, senior vice president of STEM learning at the Museum of Science in Boston and a member of the National Assessment Governing Board, which sets policy for NAEP. Students were also less likely than in 2019 to say they engage in tasks like designing research questions, debating scientific ideas and conducting experiments to explain why something happens. 鈥淎s someone who works a lot with students or with teachers who do that kind of inquiry, that’s why students get excited.鈥� 

COVID-era school closures derailed student learning in all areas, but science was hit especially hard as teachers tried to keep kids on track in reading and math. A from the Public Policy Institute of California showed that only about a quarter of districts emphasized science in their recovery efforts. Teachers were more likely to assign free online lessons and let students work at their own pace, compared with a typical school year. Widespread declines in reading performance have also hampered students鈥� ability to keep up in science at a time when technology is rapidly evolving. 

鈥淪cience is such a hands-on experience, and trying to find ways to bring that to different homes was challenging,鈥� said Autumn Rivera, a sixth grade science teacher in Glenwood Springs, Colorado, west of Denver. 鈥淓leven- and 12-year-olds really need a lot of activity.鈥�

She got families involved in 鈥渒itchen chemistry,鈥� asked students to and recorded videos of lessons to discuss with students on Zoom. One of their favorite experiments was studying the water cycle by hanging a plastic bag full of water in a sunny window. 

In the spring of 2021, , students had missed out on at least two months of science learning. By 2024, science achievement in third to fifth grade had returned to 2019 levels, but seventh and eighth graders, across all racial groups, saw the most significant declines and were still more than three months behind pre-pandemic performance.

One former education secretary warned against using COVID “as an excuse.” Margaret Spellings, who led the department during George W. Bush’s administration, noted that as with students’ achievement in other subject areas, performance in science did not improve between 2015 and 2019. Average scores for eighth and 12th graders were flat and declined for fourth graders.

A positive trend, Cunningham said, is that more elementary schools have added STEM as part of an elective rotation with art and music. Those classes can be highly engaging, but aren鈥檛 always focused on grade-level standards, she said. In addition, regular classroom teachers might scale back science lessons and focus more on reading and math. 

High and low performers

The declines in achievement were not confined to a few student groups. They affected students whether or not they live in the suburbs, come from wealthier homes or have parents who graduated from college. Students without disabilities and who speak English as a first language also scored lower than in 2019. 

But Matt Soldner, acting NCES commissioner, pointed out what he considers the one encouraging sign from the results  鈥� a 6-point increase in scores for English learners. 

鈥淣AEP describes the what, not the why,鈥� he said, 鈥渂ut that’s an interesting subgroup finding.鈥�

As with other NAEP assessments, the science results show a widening gap between students scoring at the highest and lowest levels. Scores for students in the 90th percentile dropped from 196 to 194, but fell further, from 106 to 101, for students at the 10th percentile.  In fact, for students at both the 10th and 25th percentiles, scores are at 鈥渉istoric lows,鈥� said Soldner. 鈥淭hese results should galvanize all of us to take concerted focused action to accelerate student learning.鈥�

Julia Rafal-Baer, co-founder of ILO Group, an education consulting firm, and also a member of the governing board, said access to books likely contributes to the disparities in scores. If science wasn鈥檛 a high priority in some schools, 鈥渉ow is it that high-performing kids are still absorbing enough to be able to be high-performing?鈥� she asked.

Many students, Rivera said, lack the reading skills to interpret science texts. 

鈥淚鈥檓 having to take a step back and really focus on basic reading 鈥� which is not something  that I am technically trained in as a sixth grade teacher,鈥� she said. Like many teachers, she also sees families place less emphasis on consistent attendance and good work habits. 鈥淲e鈥檙e seeing students missing work. We鈥檙e not really seeing 鈥� emphasis placed on school or on achievement.鈥�

Poor basic math skills are also hindering students鈥� progress in science, said Cunningham, who designs STEM curriculum materials for schools across the country.

鈥淭eachers are spending more time making sure that the kids are prepared to do some of the things that in the past they may have assumed kids would come equipped to do,鈥� she said. 鈥淐ould they make a table? Could they make a graph?鈥�

On NAEP, the percentage of students saying they frequently 鈥渦sed tables or graphs to identify relationships between variables鈥� fell from 43% to 39%. Less than a third 鈥渦sed math equations to explain or support scientific conclusions.鈥�

鈥楽tarving ourselves of knowledge鈥�

NAEP will assess students鈥� reading and math skills again in 2026, but the next won鈥檛 take place until 2028, again just for eighth grade. Students will take that includes a stronger emphasis on students applying their knowledge and will incorporate more technology and engineering topics. 

Because so many students 鈥� at least a third 鈥� score below basic, Cunningham added that the board felt it was important to expand the number of questions targeting students at that level.  

鈥淲e need to know more about what that population knows,鈥� she said. The questions, for example, might be simpler and require less reading.

Fourth graders were left out of the 2024 and 2028 science tests for budget reasons, Cunningham said. They鈥檙e scheduled to participate again in 2032. But one former governing board member said the absence of data from fourth graders is troubling.

鈥淚f there had to be a cut, I understand why we would, but it still raises the question of what we expect in science in early grades,鈥� said Andrew Ho, a testing expert and education professor at Harvard University. 鈥淲hy are we starving ourselves of knowledge about educational progress outside of [English language arts and math]?鈥�

Staff cuts to NCES of the results, which were expected earlier this summer. 

During a background call with reporters last week, a member of the governing board said the results were 鈥渁n opportunity for the field to see that these report cards are of the same quality that they have come to expect from the NAEP program.鈥� But an NCES official on the same call said that in light of Education Secretary Linda McMahon firing most of the center鈥檚 employees, the department will need 鈥渟ufficient staff and other resources in place鈥� to conduct the tests next year and plan for 2028.

McMahon reiterated her support for the NAEP program during a .

鈥淚f we have an objective measure across all states, like NAEP, then I think that’s the best way to go,鈥� she said. 鈥淲e will not get away from having NAEP scores and the research that we can all rely on to make sure that we’re doing the right things.鈥�  

鈥楢I-driven world鈥�

Beverly DeVore-Wedding, president of the National Science Teaching Association, still worries that the 鈥渃urrent political climate鈥� will diminish the program. 

鈥淚 am concerned about them changing the assessment picture and that NAEP could get reduced to only reading and mathematics,鈥� she said. 

The science results also have implications for other aspects of President Donald Trump鈥檚 agenda, such as incorporating artificial intelligence into learning. Last week, first lady Melania Trump hosted an event tied to the for students to use AI to address community challenges. 

鈥淚t’s not one of those things to be afraid of,鈥� McMahon said at the event. 鈥淟et’s embrace it. Let’s develop AI-based solutions to real-world problems.鈥�

Rafal-Baer said the rapid adoption of AI tools just reinforces the importance of science education.

鈥淎I is here and it鈥檚 already reshaping how we work, learn and solve problems,鈥� she said.聽 鈥淭he complexity is only going to accelerate, and we can鈥檛 afford to have a scientifically illiterate workforce trying to navigate an AI-driven world.鈥�

Correction: An earlier version of this story misstated whether the 8th grade NAEP science exam gathers supplemental data on students鈥� home environments or reading habits.

Just for fun in 2023, compared with 37% 30 years earlier. Op-eds and analyses mourn that ; meanwhile, children ages 8 to 18 spend 7.5 hours per day, on average, on screens. If there鈥檚 one institution that can make teenagers read, it鈥檚 school. And if there鈥檚 one time of year when teens have ample time on their hands, it鈥檚 summer. Indeed, to read a 300-page book over nine weeks of summer vacation is to read five pages a day. So it would make sense for schools to ask students to read more books when class is out.

I looked at 32 high schools in Maryland, the District of Columbia and Virginia to understand the state of summer reading assignments in 2025. The results are disheartening.

Only 54% assign any summer reading for incoming ninth-graders. While 90% of private schools studied have such a requirement, only 38% of public schools do 鈥� and at three of those, the 鈥渟ummer reading assignment鈥� doesn’t actually involve a book.

Maryland’s Bethesda-Chevy Chase High School asks incoming ninth-graders to read three New York Times articles and write a short paragraph about each. Catonsville High School in Baltimore County asks new high school students to read an article titled 鈥�.鈥� The at Albert Einstein High School in Kensington, Maryland, requires completion of one of four tasks: watching a documentary, listening to a music album, or reading an article or a six-page short story. In Montgomery County, Maryland; Fairfax County, Virginia; and neighboring districts, splashy graphics on school websites tell students they should strongly consider reading a book over the summer 鈥� which won’t convince anyone who isn’t interested in reading to start or entice a student who is reading to tackle something new or challenging. 

Five schools actually ask students to read more than one book 鈥� all of them private schools. Holton-Arms, an all-girls school in Bethesda, leads the pack with a four-book requirement. A six-minute drive away, at public Whitman High School, students are told to read any book that is at least 150 pages and to take notes in the margins or on sticky notes. Whitman is in the same district as Einstein and Bethesda-Chevy Chase high schools, so it鈥檚 clear the central office plays no role in requiring student summer reading. The same is true in the District of Columbia Public Schools, which before COVID required students to read a book and complete a writing assignment over the summer 鈥� and posted an 80% compliance rate 鈥� but now has nothing.

Even when districts do attempt to set a reading bar, the results are uneven. 

For example, the Newark Board of Education declared that this summer, incoming ninth-graders in the district’s 18 high schools would be required to read S.E. Hinton鈥檚 The Outsiders and complete three related activities. But on East Side High鈥檚 school website, the is to watch a TED talk and a movie. Central High prominently displays the assignment for The Outsiders, but most other Newark high schools don鈥檛 have anything listed on their websites at all about summer reading. 

Evidence from my survey suggests that public schools serving more affluent communities are more likely to mandate summer reading. Whitman is one of the most affluent neighborhood high schools in Montgomery County and the only one to require students to read a book. Similarly, Staples High School in Westport, Connecticut, where 2% of students are low-income, requires ninth-graders to read two books; Highland Park High School, outside Dallas (no low-income students), requires incoming ninth-graders to read Paulo Coehlo’s The Alchemist; and New Trier High School, just north of Chicago (3% low-income students), assigns specific books based on incoming students’ ninth-grade English placement. 

But affluence doesn鈥檛 guarantee a commitment to summer reading, either. Newton North High School, outside Boston, asks students to consider picking up a book, as does Harriton High School on Philly鈥檚 Main Line. Both districts have fewer than 18% low-income students. Georgetown Day School, the D.C. prep school where Supreme Court Justice Ketanji Brown Jackson sends her child, provides only a , as does New York City鈥檚 . 

The class that enters ninth grade this fall was in third grade when the pandemic hit. By the time these students began middle school, education was back to 鈥渘ormal,鈥� and they鈥檝e had years to be sold on the importance of academics, community and all that school has to offer. But a survey of middle schoolers in 2024 found that agree that 鈥渋n school, people don鈥檛 give up when the work gets hard.鈥� Less than half agree that 鈥渁t my school, we use our thinking skills, rather than just memorizing things.鈥� In their recent book 鈥淭he Disengaged Teen,鈥� authors Rebecca Winthrop and Jenny Anderson write that to address these sobering statistics, 鈥渨hat kids need now is to become better at learning.鈥� adolescent language ability and brain development, among other benefits, suggesting that literacy itself might be a tool for helping young adults further develop their ability to learn. Demanding a few hours of reading over the course of the summer is a small but meaningful step that can only help.

The annual , released last month from the , found that while there鈥檚 some bright spots nationally compared to 2019 鈥� including a growing number of children covered by health insurance and a decrease in teen pregnancies 鈥� many states are struggling to take care of children, whether it鈥檚 the number of children living in poverty, a growing number of teen deaths or older students who are not in school or working.

鈥淲hen we look at the overall numbers, we see a somewhat mixed picture,鈥� said Nicholas Munyan-Penney, assistant director of P-12 policy at , a national education policy group and grantee to the Annie E. Casey Foundation. 鈥淏ut, when we actually break it down by demographics, we see that there continues to be very large gaps between racial groups, in particular with Black and Latino students 鈥� [and their] educational outcomes.鈥�

Nationally, there was improvement in seven of 16 indicators, the report found. Of the remaining measures, six worsened since 2019 and three remained the same. In almost all 16 categories, however, American Indian, Alaska Native, Black and Latino children fared worse than the national average. 

The report found education topped the list for the weakest rebound in recent years with continued declines in reading and math proficiency for all demographic groups between 2019 and 2024; and a smaller percentage of children attending preschool across the country. 

Using federal NAEP test data from the National Center for Education Statistics, the report found 70% of American fourth graders in 2024 were not reading on grade level, worsening from 66% in 2019 鈥淸and] essentially undoing a decade of progress.鈥� About 73% of eighth graders are not proficient in math either.

Black, American Indian, Alaska Native and Latino students saw widening gaps compared to the national average and their white and Asian peers. In 2024, for example, about 84% of Black fourth graders and 90% of Black eighth graders were not performing on grade level in reading and math respectively compared to 61% of white fourth graders and 63% of white eighth graders.

Nationally, high school graduation rates have increased by one percentage point to 87% between 2018-19 and 2021-22, but similar to proficiency, most students of color lag behind the national average by between four to 13 percentage points.

鈥淭his really is indicative of the fact that we’ve had generations and generations of disproportionate resources going to students,鈥� Munyan-Penney said. 鈥淲e know that students of color and from low-income backgrounds have continually seen less investment in their schools and communities, and that is really borne out here in the data.鈥�

Children of color disproportionately lived in high-poverty areas in 2019-23, with around 20% of Black and American Indian or Alaska Native, followed by about 11% Latino children, who lived in areas of concentrated poverty compared to 3% for white, Asian and Pacific Islander children. 

Most states fund public schools through local property taxes, so there鈥檚 often a direct correlation between concentrated poverty and struggling student achievement, Munyan-Penney said.

Disparities also extended beyond education 鈥� particularly with the number of child and teen deaths per 100,000. 

From 2019 to 2023, the number of kids and youth who died between the ages of one and 19 per 100,000 children increased from 25 to 29, with cause of death mainly from accidents, homicides and suicides. That figure for Black youth is nearly double the national rate, with a 30% increase between 2019 to 2023, from 41 to 53 deaths per 100,000.

State-by-state child well-being has also been a moving target.

While New Hampshire, Vermont and Massachusetts topped rankings for overall child well-being, Mississippi, Louisiana and New Mexico scored the lowest. The report acknowledged that despite overall rankings, some states 鈥渟how vastly uneven scores across domains,鈥� including Maine, which scored overall at No. 17, but simultaneously ranked No. 41 in education or North Dakota which ranked first in economic well-being, but No. 42 in education. 

鈥淪trong performance at the state … level can mask the reality that millions of individual children are still struggling to access the resources,鈥� the report said.

Federal investments toward healthcare coverage and economic stability during the pandemic were credited in the report as sources of improvement in parental employment and children covered under health insurance. 

About 25% of children had a parent who lacked stable employment between 2019 and 2023, which improved by one percentage point. The report found financial aid, including pandemic relief funds in 2020-21 and an expanded child tax credit, helped “strengthen family financial security.鈥� The report also found an increase of children covered by health insurance from 5% in 2019 to 6% in 2023 was an 鈥渆ncouraging milestone.鈥�

But, these gains may too be in jeopardy in upcoming years as several pandemic-era supports expired and President Donald Trump鈥檚 administration has recently made cuts to SNAP and Medicaid.

鈥淭he pullback in federal investment鈥� is definitely a concern of mine,鈥� Munyan-Penney said. 鈥淚’m not optimistic that these numbers will continue to go up unless we sort of see a change in the way that the federal government is approaching this and or we have very robust state investment.鈥�

Take New York, for example. In its latest 鈥�, the Education Law Center adjusted school spending figures relative to their regional labor market costs. It gave New York鈥檚 school funding system an A for the total amount of money it sent to public schools, a B for the distribution of those funds among schools and an A for the amount of money it spent relative to the state鈥檚 overall gross domestic product per capita.

Overall, New York came out as one of the top-rated school funding systems, if not the highest-rated.

And yet, New York students perform slightly below national averages. While its schools $33,970 per student, $15,509 more per pupil than the rest of the country, in Fiscal Year 2022 its students overall student performance in fourth-grade math on the National Assessment of Educational Progress 鈥� the “Nation’s Report Card.”

But it鈥檚 not just New York. The relationship between school spending and student outcomes is weaker than you might imagine. To make the fairest comparison possible, I used the Education Law Center鈥檚 spending figures, which are adjusted for cost-of-living differences, and the from the Urban Institute.

The graph below shows the results for 2022, the latest year for which comparable spending figures are available. As you can see, New York is an extreme outlier: It spent more than any other state, and its results were in the bottom half. Other high-spending states like Vermont, New Jersey and Connecticut also got pretty poor results given their investments.

Which states do well on this metric? Texas, Florida and Mississippi all stand out for getting strong student outcomes despite not spending that much.

What about staffing levels? Education is mostly a people business, and the bulk of school spending goes toward salaries and benefits. But staffing levels are also not well correlated with student outcomes.

The graph below shows the number of staffers for every 500 students 鈥� think of a typical elementary school 鈥� versus the same demographically adjusted fourth-grade math scores as above. Here, Vermont and Maine stand out as having exceptionally high staffing levels without positive student outcomes to show for them. Meanwhile, Mississippi, Texas and Florida all stand out as states showing strong student test scores without high levels of staffing.

These are merely correlations, and readers should not take these arguments too far. For example, Matt Ladner, a senior adviser for The Heritage Foundation, made the case in a that the states that increased spending the most over the last two decades did not see equivalently large achievement gains. But it would harm students in, say, New York or Vermont, if state policymakers decided that schools needed to cut back on spending or staffing.

That鈥檚 because the on school finances suggests that a $1,000 increase in annual per-student spending improves test scores by 0.008 standard deviations and boosts college-going rates by 2.8 percentage points. Infusions of federal ESSER funds produced similar, albeit smaller, effects. Perhaps no one was or is with the magnitude of the returns on increased spending for public schools, and the gains are small enough that you can鈥檛 just eyeball them on a chart, but they are statistically significant and academically meaningful. Moreover, this research shows that school spending does cause test scores to rise. It would be irresponsible for policymakers to ignore these general trends.

At the same time, it is fair to note that the gains from higher spending are small, and policy is not made in a vacuum. Some places, especially the Mountain West, probably could see real gains from higher spending. Meanwhile, other places, especially in the Northeast, could benefit from more time thinking about cost effectiveness and how to drive improvements without additional funds.

The best modern example of the latter is Mississippi. Mississippi is the poorest state in the country, and it would not have been a positive outlier on these types of charts 10 or 20 years ago. But since 2013, the state has put in place a number of policy changes, including new curricular materials, a muscular school accountability system focused on the students who are the furthest behind and a third grade reading requirement that brought greater attention to children who struggle with the basics. Some of these initiatives even cost money, but they didn鈥檛 add up to that much relative to the state鈥檚 overall education budget, and they helped students in Mississippi their peers in higher-spending states.

It’s hard to have these types of nuanced conversations when some advocates continue for more money, even in well-funded states and communities, while others have 鈥� and using 鈥� the modest gains from spending increases as evidence in favor of school choice or other reforms. For policymakers, the only way to correctly understand the nuances of school spending is to recognize that it matters while also understanding its limitations. 

The need for a reliable compass is greater now more than ever before. Significant budget cuts and efforts to weaken the Department of Education call into question the federal government鈥檚 role in ensuring educational excellence. As policy and research leaders at the National Center for Learning Disabilities, an organization with nearly 50 years of experience advocating for the rights of individuals with learning disabilities, we fear the country may be entering an era in which we will be navigating this path without a reliable compass.

NAEP, often referred to as the 鈥淣ation鈥檚 Report Card,鈥� is the only standard assessment administered to students in all 50 states, allowing Americans to see trends in educational attainment over time. NAEP data is an invaluable indicator for understanding areas of growth and identifying unmet needs for students with learning disabilities. The center publishes NAEP Data Snapshots in and to compare the achievement of students with learning disabilities to that of all young people. Research and experience demonstrate that students with learning disabilities can achieve at the same level as their peers when given the appropriate support. Thus, these NAEP snapshots are a powerful tool for highlighting persistent discrepancies that warrant additional investment and innovation, and celebrating and learning from progress. 

Not long ago, substantial numbers of students with disabilities were barred from taking  NAEP. Only 15 years ago, over a third of students with disabilities were excluded from the test-taking sample. In 2009, representatives of the center testified before the National Assessment Governing Board to have students with disabilities included in the assessment. Since then, there has been progress. In 2024, about 89% of students who identified as having disabilities took the reading assessment. Their performance is part of the national conversation about how the nation navigates the journey toward educational excellence together.

NAEP also spurs innovation. Technology is changing rapidly, and because students take the exam on the computer, that data can be used in informative and transformative ways. Researchers have used not just the NAEP results, but also information about the testing process to help better understand how students interact with exams. For example, because of , it is clear that it isn’t just students with disabilities who use built-in accessibility features like text-to-speech.

In February, NCLD was alarmed to learn that the National Center for Education Statistics, the primary agency responsible for administering NAEP, had been reduced to a staff of three. Both administering the test and disaggregating and reporting the resulting data are critical work requiring expertise and staffing that are now at a skeletal level. Though Education Secretary Linda McMahon has stated that NAEP is safe and will be given as planned in 2026, we have many questions and concerns about its future and the ability of researchers in the field 鈥� disability advocates in particular 鈥� to use the data as we do now. The president鈥檚 budget request for Fiscal Year 2026 proposes a 29% cut to funding for NAEP, which only adds to our concerns about the test鈥檚 administration next year.

Policymakers at all levels need reliable data upon which to base their decisions. Researchers need valid and reliable achievement data to understand trends as well as differences in needs across states and student characteristics, including disability status. NAEP allows policymakers to compare their state with others nationwide, encouraging the adoption of best practices and reforms that have worked elsewhere, like Science of Reading-based reforms. At the national level, advocates like us use NAEP results to push for changes in legislation or investments that support students with disabilities.

This year marks the 50th anniversary of the passage of the Individuals with Disabilities Education Act. Yet, today, the federal government continues to underfund both special education and the broader effort to ensure all students, especially those from underserved communities, have the resources and opportunities they need to succeed. 

Educators, researchers, advocates and families committed to disability rights and inclusive education must speak up about the importance of NAEP data. This information has long helped school leaders and policymakers understand where students are excelling and where they are being left behind. Without this insight, funding, accommodations and student resources decrease, and NAEP data becomes a mirror, not a map forward.

Congratulations and welcome to a place we once knew well. You face any number of tough challenges on behalf of American students, parents, educators and taxpayers, as well as the administration you serve, but your 鈥淒epartment鈥檚 Final Mission鈥� shows that you鈥檙e well prepared to meet them. We particularly admire your commitment to making American education 鈥渢he greatest in the world.鈥�

But how will we 鈥� and you, and our fellow Americans 鈥� know how rapidly we鈥檙e getting there? By now, you鈥檙e probably aware that the single most important activity of the department you lead is the National Assessment of Educational Progress, known to some as NAEP and to many as the Nation鈥檚 Report Card.

That鈥檚 the primary gauge by which we know how American education is doing, both nationally and in the states to which you rightly seek to restore its control.

Almost four decades ago 鈥� during Ronald Reagan鈥檚 second term 鈥� it was our job to modernize that key barometer of student achievement. Five years after A Nation at Risk told Americans that their education system was far from the world鈥檚 greatest, state leaders 鈥� governors especially 鈥� craved better data on the performance of their students and schools. And they were right. At the time, they had no sure way of monitoring that performance.

That was one of our challenges, back in the day. Advised by a blue-ribbon study group led by outgoing Tennessee governor (and future U.S. senator) Lamar Alexander, and with congressional cooperation spearheaded by the late Ted Kennedy, in 1988 we proposed what became a bipartisan transformation of an occasional government-sponsored test into a regular and systematic appraisal of student achievement in core academic subjects, administered by the National Center for Education Statistics (part of your Institute for Education Sciences) and overseen by an independent group of state and local leaders, plus educators and the general public. (One of your responsibilities is appointing several terrific people each year to terms on the 26-member National Assessment Governing Board.) 

That 1988 overhaul made three big changes:

• Creation of that independent board to ensure the data鈥檚 integrity, accuracy and utility;
  
• Inauguration of state-level reporting of student achievement in grades 4, 8 and 12, i.e. at  the ends of elementary, middle and high school; and
  
• Authorization for the board to set standards 鈥� known as achievement levels 鈥� by which to know whether that achievement is satisfactory.

Much else was happening in U.S. education at the time: School choice was gaining traction. States were setting their own academic standards and administering their own assessments. Graduation requirements were rising as the economy modernized and its human capital needs increased. 

As these and other reforms gathered speed, NAEP became the country鈥檚 most trusted barometer of what was (and wasn鈥檛) working. You alluded to NAEP data during your confirmation hearing. President Donald Trump deploys it when referencing the shortcomings of U.S. schools. For example, his Jan. 29 executive order on school choice began this way: 鈥淎ccording to this year鈥檚 National Assessment of Educational Progress (NAEP), 70% of 8th graders were below proficient in reading, and 72% were below proficient in math.鈥�

Everybody relies on NAEP data, and its governing board鈥檚 standards have become the criteria by which states gauge whether their own standards are rigorous enough. Just the other day, Gov. Glenn Youngkin鈥檚 board of education used them to benchmark Virginia鈥檚 tougher expectations for students and .

Reading and math were, and remain, at the heart of NAEP, but today it also tests civics, U.S. history, science and other core subjects 鈥� exactly as listed in your speech.

But NAEP is not perfect. It needs another careful modernization. It should make far better use of technology, including artificial intelligence. It should be nimbler and more efficient. The procedures by which its contractors are engaged need overhauling. (The Education Department鈥檚 whole procurement process needs that, too 鈥� faster, more competitive, more efficient, less expensive!)

Yet NAEP also needs to do more. Today, for instance, it gives state leaders their results only in grades 4 and 8, not at the end of high school. It doesn鈥檛 test civics and history nearly often enough, and never in 12^th grade, even though most systematic study of those subjects occurs in high school. (It probably tests fourth- and eighth-grade reading and math too often 鈥� the result of a different federal law.) 

Doing more shouldn鈥檛 cost any more. Within NAEP鈥檚 current budget 鈥� approaching $200 million, a drop in the department鈥檚 murky fiscal ocean 鈥� much more data should be gettable by making new contracts tighter and technology smarter, squeezing more analysis from NAEP鈥檚 vast trove and having staffers put shoulders to the wheel. (Former IES director Mark Schneider has the .)  But making this happen will take strong executive leadership, an agile, hardworking governing board and your own oversight. You may decide it鈥檚 time for another blue-ribbon group to take a close look at NAEP and recommend how to modernize it again without losing its vital ability to monitor changes over time in student achievement.

Yes, this is all sort of wonky. NAEP results get used all the time, but it鈥檚 far down in the bureaucracy and doesn鈥檛 make much noise. Nobody in Congress (as far as we know) pays it much attention. Yet it remains 鈥� we believe 鈥� the single most important activity of your department. Which, frankly, is why it needs your watchful attention! 

We wish you well in your new role. Please let us know if we can help in any way.

Sincerely, 

William J. Bennett, U.S. Secretary of Education (1985-88)

Chester E. Finn Jr., Assistant Secretary for Research & Improvement and Counselor to the Secretary (1985-88) 

There are two ways for a gap to grow. The top can pull away, or the bottom can fall out. Here in the United States, the key problem is that the bottom is falling, and the changes are not small. At the national level, in fourth grade reading, the scores of the top 10% of students fell 0.5 points from 2013-24. Meanwhile, the scores of the bottom 10% fell 15 points. 

The state-level results are even more jarring. To put those in perspective, consider that the average student gains about 8 points per year on the National Assessment of Educational Progress reading tests. The chart below shows the change from 2013 to 2024 in the scores for the bottom 10% of students in each state. For this group, 40 states saw a decline of 10 points or more, 16 saw declines of 20 points or more and two states 鈥� Delaware and Maryland 鈥� had declines of more than 30 (!) points.

But one state is bucking this trend: Mississippi. Indeed, there鈥檚 been a fair amount of coverage of Mississippi鈥檚 reading progress in recent years, but its gains are so impressive that they merit another look.  

First, it鈥檚 worth remembering that Mississippi is the state in the country. Its per-capita income is below $50,000, and it spends less on its public schools than all but three states. But when the Urban Institute NAEP scores based on each state鈥檚 demographics, Mississippi鈥檚 fourth-grade reading scores came out on top. 

Even without those adjustments, though, Mississippi looks pretty great. Its Black students third nationally, and its low-income kids outperform those in every other state. Mississippi is also the only state to see gains across all performance levels over the last decade. Its average went up, but so did the scores of its highest and lowest performers. Mississippi raised the bar and the floor at the same time. 

So how did they do it? How did Mississippi go from 49^th in the country a decade ago to near the top today? And what can other states learn from it? 

According to a recent piece by Grace Brazeale, a policy associate with the advocacy group Mississippi First, the state implemented of changes starting with the 2013 . That law funded the state department of education to hire, train and literacy coaches to the 50 lowest-performing schools. It also required schools to administer universal screenings to identify students with reading deficiencies early and to communicate those results to parents, and it required schools to hold back students who were not reaching a certain threshold by third grade. 

These changes were not all that expensive, but they had big effects. EdWeek鈥檚 Elizabeth Huebeck in 2023 that the state spent $15 million per year to support its literacy work, and 60% of that went to coaching and intervention staff. A research paper last fall from Noah Spencer from the University of Toronto that the law helped drive the state鈥檚 gains.

Spencer estimated that the third-grade retention policy alone could be responsible for about one-quarter of the gains, and it was surely the most controversial element. Some have even tried to cast doubt on Mississippi鈥檚 NAEP gains by arguing they鈥檙e merely a function of testing older kids. But this has been : Mississippi does hold back more kids than other states, but it , and the average age of Mississippi鈥檚 NAEP test-takers has barely budged over time. 

on third-grade retention policies has that students who are retained tend to have better outcomes than those who are not, but that the process for identifying those children can be biased against Black, Hispanic, and low-income students. 

However, I think what matters most is not the students who are retained, but what the policy does to . Mississippi required schools to notify parents when their child was off track and to craft individual reading plans for those with reading deficiencies. In other words, the threat of retention may have shifted behavior in important ways. As evidence for that theory, consider that a study out of Florida also found positive effects on younger siblings of students who are retained. 

Other states have literacy policies too, but they are often weaker than what Mississippi implemented. For example, a 2023 paper from Michigan State University researchers John Westall and Amy Cummings that 41 states and the District of Columbia had early literacy policies, but only 12 could be considered . Many states had the superficial elements of a literacy plan, but they lacked requirements that all districts adopt high-quality instructional materials, screen all students for dyslexia and take dramatic action for kids who continue to struggle to read. Critically, Westall and Cummings found that only states with truly comprehensive policies saw student learning gains. 

This is perhaps the key lesson for state policymakers: The only literacy policies that are likely to lead to significant student learning gains are ones that meaningfully change schools and classrooms. With reading scores continuing to fall in many parts of the country, policymakers should look to replicate the lessons from Mississippi. 

One of three long-term studies that has measured student performance in math and reading since the 1970s, the assessment was set to begin in March and run through May.

But Westat, a research organization handling the assessment for the National Center for Education Statistics, notified state officials Wednesday that the department had canceled the test.

鈥淭he U.S. Department of Education has decided not to fund the NAEP 2024-2025 Long-Term Trend Age 17 assessment,鈥� Marcie Hickman, project director of the NAEP Support and Service Center, said in the email, which was shared with 社区黑料. 鈥淎ll field operations and activities will end today, February 19, 2025.鈥�

A long-term trend assessment of 13 year olds was conducted last fall. The age 9 administration is currently underway through March 14 and will continue. Age 17 data, however, hasn鈥檛 been collected since 2012, creating a significant gap in understanding older students鈥� academic performance.

The tests, which are mandated , were set to be administered during the 2019-20 school year, but were canceled due to the COVID outbtreak.聽

This year鈥檚 data would have set a new baseline for understanding how older students are recovering 鈥渇rom pandemic-era learning losses,鈥� said Andrew Ho, an assessment expert at Harvard University and former member of the National Assessment Governing Board, which sets policy for NAEP. 

The cancellation, he said, could undermine the nation鈥檚 trust in the assessment program. 

鈥淭his is just the first direct evidence that executive actions have weakened NAEP and its 鈥榞old standard鈥� infrastructure for monitoring educational progress.鈥� 

Every two years, NAEP tests fourth and eighth graders in math and reading. While results from 12th graders, collected last year, are expected this summer, they won鈥檛 be tested again until 2028.

鈥淣AEP’s biggest gaps already are at the end of high school, telling us what kids do [or] don鈥檛 know and can [or] can鈥檛 do as they prepare to enter the real world,鈥� said Chester Finn, president emeritus of the conservative Thomas B. Fordham Institute and former chair of the board. The long-term trend study has 鈥渉elped fill the gap.鈥� 

The move contradicts what department officials said last week when they canceled nearly $900 million in contracts for the Institute of Education Sciences, which includes NCES and NAEP. Madison Biedermann, a spokeswoman for the department, said at the time that work related to NAEP . 

But Elon Musk鈥檚 continues to cancel contracts it deems either wasteful or contrary to President Donald Trump鈥檚 executive order related to diversity, equity and inclusion.聽

The fact that the test was not administered in 2016 or 2020 鈥渄oes not make it a very effective longitudinal study,鈥� Biedermann told 社区黑料 Thursday. She did not indicate whether it would be rescheduled.

The department has also canceled a contract for conducting background checks on field staff who administer NAEP tests in schools. Biedermann did not provide details on the vendor or the amount of the contract.

鈥淲e’re going to try to re-scope and re-evaluate these contracts,鈥� she said. If they are 鈥渁bsolutely necessary,鈥� they will be re-bid, she said. 鈥淎 lot of these contracts, in our evaluation, are not cost effective and not meeting the standards.鈥� 

Biedermann maintained that the core NAEP program in fourth and eighth grade reading and math 鈥渋s not being touched.鈥� 

Results released last month showed that students continue to lose ground in reading. Eighth grade results in math were flat, and while fourth graders saw gains in math, those results were driven by the highest-performing students.

The long-term trend program is different from the primary NAEP assessment because it has tested students on essentially the same items for over 50 years. 

鈥淚t鈥檚 important that we have the long-term trend because of the consistency of the test,鈥� said Dan Goldhaber, director of the Center for Analysis of Longitudinal Data in Education Research at the American Institutes for Research. Measuring performance on the same test items allows officials to monitor changes across generations.

He added that at least the results from 12th graders, collected last year, will provide 鈥渋nformation about how kids are doing at the end of high school.鈥澛�

The 74’s Senior Writer Greg Toppo contributed to this report.

One of the most striking features of the troubling results from the most recent National Assessment of Educational Progress is the much lower percentage of students scoring proficient on NAEP than on many states鈥� own 2024 standardized exams.

By now, you鈥檝e likely seen the results: modest improvements in math, but not enough to get students back to pre-pandemic performance levels; fourth graders fell further behind in reading; a record 34% of eighth graders scored 鈥渂elow basic鈥� in reading. 

In addition to the national summaries, NAEP reported student achievement in each state, where proficiency rates ranged from a high of 51% in fourth-grade math in Massachusetts to a low of 14% in eighth-grade math in New Mexico.

States, of course, are required by federal law to administer their own annual standardized tests in math and reading. FutureEd students鈥� performance on NAEP and on their states鈥� tests and found that, for the most part, students met proficiency standards at significantly higher rates on their states鈥� exams, especially in reading.

The gaps were at least 15 percentage points in three-quarters of the states. In some, they were even greater. Seventy-two percent of Virginia鈥檚 eighth graders were proficient in reading, more than double the percentage on NAEP. Iowa reported more than three-fourths of its eighth graders proficient in reading in 2024, compared with less than a third of the state鈥檚 students on NAEP. We also found that the gaps increased in many states between 2022 and 2024, including in 26 states in fourth-grade reading and 22 states in eighth-grade reading.  

Why is there so much misalignment between NAEP and state results?

Perhaps more than any other factor, it鈥檚 lower state standards.

To achieve proficiency on the national assessment, students must show 鈥渟olid academic performance and competency over challenging subject matter.鈥� That鈥檚 where Rhode Island, Massachusetts and the District of Columbia set their proficiency bar. But most states’ of that benchmark, landing within the range of NAEP鈥檚 lower “basic” standard, which requires students to demonstrate only 鈥減artial mastery of fundamental knowledge and skills.鈥� In Virginia 鈥� which has introduced new academic standards 鈥� and Iowa, the bar for reading falls below even that.

What鈥檚 more, in  and other states, students can be performing 鈥渙n grade level鈥� without meeting the state鈥檚 鈥減roficient鈥� standard in the subject they鈥檙e studying. And some states have gone further, lowering the passing grades on some or all of their standardized tests in recent years.

The Oklahoma State Department of Education reported significant gains in 2024, including a 24-point jump in the percentage of students achieving proficiency in fourth-grade reading since 2022 and across-the-board improvements over pre-pandemic levels. But the gains coincided with a lowering of the state’s proficiency standards, which officials didn鈥檛 publicize when they released the improved test scores.  obtained by an Oklahoma news organization revealed that the 2024 scores would have been the same as or slightly lower than 2023 results if the standards had remained the same. On NAEP, Oklahoma鈥檚 proficiency rates declined in reading and improved slightly in math between 2022 and 2024, but they remained below pre-pandemic levels.

Similarly, New York reported across-the-board improvements in student achievement in 2024 after  in 2023. But these gains were not mirrored on all of the state鈥檚 2024 NAEP results. Wisconsin also registered higher proficiency rates on its 2024 assessments after , only to have most of its NAEP scores decline in 2024. This points to the value of an independent national measure of student achievement, like NAEP.

One of the more troubling findings from the 2024 state assessment cycle is the wide gap in proficiency rates between fourth and eighth grades, with eighth graders, on average, performing much worse than their younger counterparts. The gaps are far more pronounced in math than in reading. In New Jersey, for example, 45% of fourth graders were proficient, compared with only 19% of eighth graders. Similarly, in Washington, D.C., 29% of fourth graders and just 12% of eighth graders achieved proficiency.

With many schools struggling to return to performance levels that were declining even before COVID’s disruptions, having an accurate measure of achievement is critical. Aligning more state proficiency standards with NAEP’s would increase transparency and make it easier for everyone 鈥� students, parents, teachers, administrators and elected officials 鈥� to be clear on where every state needs to focus to improve educational outcomes for all students. 

Before fourth and eighth graders took the 2024 NAEP math assessment, they were asked how many days of school they had missed the previous month. Forty-nine percent of fourth-graders who would score at the 75^th percentile or higher on the test had missed no days the previous month, compared with 26% of those scoring below the 25^th percentile.

Equally striking, 45% of students in the bottom quartile reported missing three or more days of school in the previous month compared with just 20% of students in the top quartile. And at the extreme ends of the absenteeism spectrum, 7% of the lowest-performing eighth graders on the NAEP math test reported missing more than 10 days of school in the previous month, compared with just 1% of top scorers.

Correlation strongly suggests causation. It鈥檚 impossible to prove that students performed better because they were in school every day, but it鈥檚 the logical conclusion.

A detailed comparison of state test scores and student absenteeism by Rhode Island education officials suggests as much. They found that just 10% of students who had been chronically absent for three consecutive years scored proficient on Rhode Island鈥檚 own standardized math tests in 2024, and 13% were proficient in reading. In contrast, 40% of students who attended regularly were proficient in math and 38% were proficient in reading. As on the state’s dashboard, 鈥渓ong-term chronic absenteeism has a compounding negative impact on student performance.鈥�

Attendance influenced achievement significantly even among students facing the many challenges of poverty. While it’s hardly surprising that only 18% of Rhode Island鈥檚 low-income students who attended school regularly were proficient in reading and math last year, just 11% of those who were chronically absent were proficient in reading, and only 9% met that bar in math.

The upshot is there needs to be a relentless focus at the state and district levels 鈥� beyond the work of individual schools 鈥� on getting every student in school every day. Transparency is essential to progress. Rhode Island is the only state that publishes detailed, real-time attendance data for every one of its public schools, allowing officials to correlate state test scores and absenteeism.  More than a dozen states have yet to release attendance data from the 2023-24 school year, making it difficult for policymakers to even know which absenteeism problems they need to solve.

The quality of instructional materials, tutoring programs or new technology tools can’t make much of a difference if students aren鈥檛 in school.

I would argue that the causes are dual problems that are not only intertwined, but sit at the very core of this crisis of stalled learning recovery: Far too many boards of education have abandoned accountability and failed to embrace innovation. Yes, the challenges education stakeholders 鈥� parents, educators and policymakers alike 鈥� face are enormous, but the solutions are within reach if, collectively, we commit to a balanced approach. To recover lost ground, decisionmakers need to learn from bright spots in education, where accountability and innovation intersect to drive real results.

At the National Association of Charter School Authorizers we know high-performing charter schools can serve as models for driving meaningful results for students. Research shows how from authorizers contribute to charter school success and how high-performing charters have . NACSA鈥檚 own research has demonstrated the and approaches to learning that have found a home and are growing in the charter sector. Charter schools have consistently proven their ability to dramatically accelerate the learning of students furthest away from opportunity 鈥� a critical capacity, given the sobering NAEP outcomes for students academically behind. 

Here are three big lessons, based on the successes we have seen from authorizing high-quality charter schools, that decisionmakers can use to ensure all students receive an excellent education:

Understand accountability, encourage innovation and problem solving: For accountability to truly drive meaningful student results, there must be a shared understanding of what it is and is not. Accountability systems or frameworks ensure that all students meet high academic standards by defining expectations, creating conditions for innovation and accommodating the needs of families and communities to drive change when schools fall short. While accountability must impose consequences for failure 鈥� including making the tough decision to close schools that consistently fail to educate students well 鈥� it鈥檚 about so much more than that. It is about permitting excellent schools to expand so their successes can be replicated and, perhaps most importantly, ensuring that high and rigorous expectations for student learning are perpetuated.

When accountability is at its best, it fosters greater innovation. This is especially true in high-quality charter schools, where accountability and innovation are interwoven by design. These schools are given the flexibility to innovate and problem-solve in multiple ways 鈥� such as implementing evidence-based instructional models that draw from research-backed teaching strategies, finding creative ways of re-engaging students and using technology in unique ways 鈥� while being held to rigorous performance standards, ensuring they deliver for their students. It is this balance that needs to be more broadly replicated, allowing accountability and innovation to compliment one another in service of student learning and growth.

Align rigor with clear guidance: In order for teachers and school leaders 鈥� and, ultimately, students 鈥� to excel, they need policies and practices that promote achievement, which includes explicit guidance around expectations and clear measures of success.

Charter school authorizers play a critical role in ensuring expectations are both rigorous and attainable. By using tools like school performance frameworks and reports, progress monitoring systems, site visits, consistent parent communications and other evaluation and transparency systems, authorizers provide schools with the tools to understand how they鈥檙e performing and what鈥檚 expected of them. This alignment empowers educators to focus on what matters most: student learning.

Follow the excellence equation: High-performing charter schools have two important elements that drive their success: They are responsible for determining and executing their programmatic objectives, while their authorizers must set clear expectations and be transparent about performance. This clear delineation is the key to producing high-functioning schools where students are prepared for the next phase of learning and life.

At NACSA, we鈥檝e observed that this excellence equation 鈥� balancing accountability for outcomes with freedom to innovate 鈥� is a common feature of the highest-performing schools. All schools should be empowered to tailor their programs to meet their students’ needs while accountability structures ensure their efforts are aligned with clear, rigorous goals.

As states and districts consider changes to their accountability systems, lowering standards and misleading students and families about the reality of academic progress is not the answer. The new NAEP results make it clear that now is the time to raise expectations and work together to leverage every tool available to advance learning.

NAEP also offers insight into why achievement gaps are widening. Survey results show increasingly lower expectations and higher absenteeism rates among both students and their teachers. Nearly 60% of high-performing students said they were asked more than five times to write long answers to questions on tests or assignments that involved reading last school year, compared with only 32% of low-performing students.

No one should dismiss these results or try to explain them away. American students are not receiving the educational opportunities they deserve or the preparation they need to succeed in an increasingly complex society and changing economy. The current system is failing them, and the consequences will be severe and long-lasting.

Our analysis in September was clear: Unless something changes dramatically, this trend will continue for years. The current struggles are not just pandemic-induced: There has been little sustained progress in student achievement since the early 2000s; in fact, in some areas, progress has backslid. Some states and cities have managed to make progress, and they may offer useful lessons, but the overall picture remains bleak. There are no quick fixes, but there are certainly steps that can and must be taken.

The first is acknowledging the scale of the problem and responding with urgency. Policymakers and educators can implement evidence-based strategies to help students recover. We and others have consistently documented the positive impact of:

• High-dosage tutoring;
  
• Summer school programs;
  
• High-quality curriculum;
  
• Excellent teachers.

None of these are fundamentally partisan issues. Red and blue states alike are adopting these interventions, and they should be implemented everywhere. These are the basics. Are the decisionmakers who fail to put these proven interventions into place truly serious about solving the problems at hand?

Let鈥檚 be honest: The NAEP scores make their own case that what’s needed are fundamental reforms 鈥� not just tinkering around the edges. This is not just about playing catch-up from the pandemic; it is about redesigning an education system that has been failing too many students for too long. This means embracing bold, evidence-based reforms, even when they are politically difficult:

• Expand high-performing public charter schools. The best charters are delivering results for the students that NAEP shows are falling farthest behind. The nation needs more of them.
  
• Redesign high school. The current model is outdated and ineffective for too many students. The nation needs schools that are more relationship-based, relevant and engaging.
  
• Leverage emerging technologies. Tools driven by artificial intelligence that support personalized learning, tutoring, curriculum and assessment can help ensure all students get the support they need while empowering educators to be more effective.
  
• Provide families with honest data. Parents deserve to know how their child is really doing, not just receive meaningless report cards that obscure academic struggles.
  
• Hold adults accountable for student outcomes. Education leaders and policymakers must be responsible for results through thoughtful and fair accountability mechanisms.

Every city and state in this country has work to do, and that has been clear for a long time. The inertia, political resistance and implementation fatigue that have held back so many students must be confronted head-on. Now is the time for leadership.

If you have been paying attention, the NAEP results should not shock you. What should shock you is that education systems are not, on the whole, changing course. Isn’t the very definition of insanity doing the same things while expecting different results?

The data are clear. Young kids are not catching up. Gaps were widening even before the pandemic. The crisis is real, and it is not going away on its own. Believing the NAEP results means acting on them. Hope is not a strategy. Strong leadership, political courage and a commitment to evidence-based reforms are the only paths forward.

Governors, state chiefs, mayors and federal officials must commit to the long, politically challenging work of ensuring that all American students can realize their full potential. If they do not, this will be yet another opportunity squandered 鈥� and the cost will be measured in the futures of millions of children.

During a heavy snowstorm Tuesday that caused schools to close all over Wisconsin, State Superintendent of Public Instruction Jill Underly spoke by telephone with the Wisconsin Examiner about the health of the state鈥檚 public education system, student achievement, the growth of school vouchers, political attacks on diversity and her hopes for the coming year.

Parents bill of rights

As we spoke, Republican legislators were preparing to hold an executive session Thursday on , a 鈥淧arents Bill of Rights鈥� that encourages lawsuits by parents who feel that their rights have been violated because they were not informed about medical services offered at school or about the discussion of 鈥渃ontroversial subjects鈥�  in class, including gender identity and racism, or because they were not given the authority to determine the names and pronouns used to address their children.

Under the bill, a parent or guardian who successfully asserts a claim 鈥渕ay recover declaratory relief, injunctive relief, reasonable attorney鈥檚 fees and costs, and up to $10,000 for any other appropriate relief.鈥�

Get stories like this delivered straight to your inbox. Sign up for 社区黑料 Newsletter

鈥淭he reality is that meaningful parental engagement is happening every single day between our teachers and their students鈥� families and caregivers,鈥� Underly said. The Parents Bill of Rights 鈥渋s designed to shut down discussion and creates an environment of fear for our educators because it inserts them into a culture war that no one should be fighting in the first place.鈥�

She sees the bill as part of a larger pattern of attacks on public schools and democracy itself.

鈥淵ou think about the things that the Legislature picks up on,鈥� Underly said. 鈥淟et鈥檚 attack libraries. Let鈥檚 attack the curriculum. Let鈥檚 attack teachers, let鈥檚 attack school boards because they wanted to wear masks during the virus. 鈥� I think it鈥檚 really a way to make sure that we instill distrust in our public institutions.鈥�

There is 鈥渁 lot of misinformation out there,鈥� Underly added, propagated by people and groups insinuating that schools provide inappropriate materials to kids. 鈥淭hat鈥檚 by design. Misinformation is designed to stoke outrage.鈥�

Another Republican bill, , would require public schools to comply with written requests from residents in their districts to inspect a textbook, curriculum or instructional material within 14 days.

鈥淭hat鈥檚 really burdensome,鈥� said Underly. 鈥淟et me just say right now, if you have a question about curriculum, you can access that. You contact the school, the principal and the teacher will work to get you the information.鈥�

School voucher lawsuit

The message that public schools are 鈥渇ailing鈥� and do not adequately serve Wisconsin families has been promoted for decades by advocates for school privatization, including the Bradley Foundation, which also Milwaukee鈥檚 first-in-the-nation school voucher program. That program, which started out serving 350 kids, has mushroomed to include more than 52,000 students in the statewide, Racine and Milwaukee programs.

In December, the Wisconsin Supreme Court declined to hear a challenging Wisconsin鈥檚 private school voucher program. The suit, sponsored by Minocqua Brewing Co. owner Kirk Bangstad, named Underly, in her official capacity, as a defendant. It charged that taxpayer-financed private school vouchers are a huge financial drain, pushing local public school districts into a 鈥渄eath spiral鈥� and that they violate the state constitution鈥檚 promise to provide high-quality public schools for every child.

Asked to comment on the lawsuit, Underly said she couldn鈥檛 speak to the constitutionality of school vouchers. But, she added,  鈥淚 believe that we cannot afford two school systems.鈥�

鈥淲e need to robustly fund the system that serves all kids,鈥� she said, 鈥渁nd that鈥檚 our public schools.鈥�

(Late last year Underly another recent Supreme Court lawsuit, filed by teachers and other public employees challenging Act 10, the 2011 law that took away most collective bargaining rights from most public employees: 鈥淩eturning collective bargaining rights to public sector employees will strengthen our educator workforce, and strengthening our educator workforce will improve our children鈥檚 education and create a stronger future for our state,鈥� she said in a statement.)

Even though the voucher lawsuit was kicked back down to lower court, Underly said it could still help raise awareness  that, unlike public schools, which are open to every child, Wisconsin鈥檚 school choice programs 鈥渁re allowing these schools that accept vouchers to discriminate against students, students with disabilities, students who are LGBTQ+.鈥�

Worrying about LGBTQ kids

Underly said she worries 鈥渁ll the time鈥� about the well-being of LGBTQ kids in Wisconsin. She cited data showing that 鈥渢hese kids who struggle to feel included or to be seen, you know, their mental health struggles are higher.鈥�

鈥淎t the heart of all this I think what I would like people to realize, and I think many people do, [is that] at the center of all of this is a child.鈥�

鈥淎nd when we attack them,鈥� she added, 鈥渨hen we tell them, you know, their identity doesn鈥檛 matter or we have to take down symbols that show that they鈥檙e included, that鈥檚 hurting them. 鈥� It鈥檚 saying that you don鈥檛 belong here or you鈥檙e not wanted. 鈥� I just want to tell people, these are kids. These are human beings. And they deserve love and empathy.鈥�

Missing the Regents鈥� vote to cut back DEI

Along with recent efforts to ban books and remove LGBTQ Pride flags, Wisconsin schools have been at the center of a battle over diversity, equity and inclusion (DEI) programs. Underly, who serves on the University of Wisconsin Board of Regents, was absent for the vote in which the Regents reversed themselves and agreed to legislative Republicans鈥� demands that they eliminate DEI positions in exchange for promised funding for faculty raises and capital improvements.

Underly was out of the country, traveling with her elderly mother in Austria, on a vacation she said she鈥檇 had to reschedule several times, when the Regents voted 9-8 to reject the deal limiting diversity positions on Saturday, Dec. 9. She was still out of the country the following Wednesday, Dec 13, when the Regents reversed their decision in a second vote.

Between votes, Underly issued a statement asking that the second vote be postponed so she could attend. She had intermittent internet access, she explained, and wouldn鈥檛 be available at the meeting time. But the Regents went ahead without her.

鈥淧art of my frustration with that is that my position on diversity, equity and inclusion is very clear,鈥� Underly said. 鈥淚 think people knew how I was going to vote. Unfortunately, I couldn鈥檛 make it 鈥�  I wasn鈥檛 part of any of the discussions.鈥�

Like Gov. Tony Evers, Underly doesn鈥檛 believe there should have been any further negotiations between the Regents and the Legislature over funds that were already approved as part of the state budget.

Now, as Assembly Speaker Robin Vos pledges to eliminate every trace of DEI throughout the state, Underly said, 鈥淚t鈥檚 definitely that slippery slope argument. You give in on one thing, and they certainly will want to take more.鈥�

Still, she added, 鈥渢hese programs aren鈥檛 going to go away. 鈥� They exist to make sure that every citizen in the state of Wisconsin has access to higher education. That includes veterans. That includes kids from rural Wisconsin who want to study to become doctors. It includes women. It includes kids who come from economically disadvantaged backgrounds.鈥�

Will UW hold onto minority scholarship programs and other targets of Republicans in the Legislature, and somehow meet its agreement to eliminate the language of DEI without actually getting rid of programs that promote diversity?

鈥淚 don鈥檛 know,鈥� Underly said. 鈥淚 guess in my role as Regent what I do look forward to is having these conversations and in many ways protecting these positions [including] the scholarships and [other] components.鈥�

What about voucher schools that serve underserved kids?

On the flip side, what does Underly make of the argument made by school choice advocates like Madison鈥檚 One City independent charter school founder Kaleem Caire, that Wisconsin鈥檚 between Black and white students is unacceptable and the lack of diversity among teaching staff contributes to a lousy environment in the local public school district for Black kids?

鈥淚鈥檓 not going to say that his heart鈥檚 not in the right place,鈥� said Underly. 鈥淲e want all kids to be successful, and he is in a community and he interacts with children of color and their families all the time.鈥�

Still, 鈥淚 don鈥檛 think the answer is pulling kids out of public schools and funding private schools,鈥� Underly said. 鈥淚 would argue the opposite and say we need to put the resources in the public schools so that all kids can be successful.鈥�

Working on teacher training, curriculum, adjusting the length of the school day or the school year are all 鈥渨ays we could address the achievement gap, and the opportunity gaps that we see, especially among children of color,鈥� she said.

鈥淭his is really where we get at the root of what equity is,鈥� Underly added, 鈥済etting the schools what they need, so that their kids can be successful, and that鈥檚 not going to be the same thing in every school or in every community.鈥�

Poverty and student success

Among the biggest equity issues public schools must address, Underly said, is poverty.

Children facing housing insecurity and hunger are 鈥渘ot going to score as well on a standardized test,鈥� she said.

鈥淲hat public schools have done is they鈥檝e tried to level that playing field. They have provided food for kids, they provide stability, whether it鈥檚 for in-school or after-school programs, they provide the art and the music and these enrichment classes that kids in poverty perhaps can鈥檛 afford to get outside of school.鈥�

The whole purpose of public schools is to create a more equitable society by providing opportunity to kids whose families live in poverty. 鈥淭hat鈥檚 a fundamental value of democracy,鈥� said Underly. 鈥淭hat鈥檚 inclusion 鈥� making sure that not just the wealthy have access to these things.鈥�

Fundamentally, Underly agrees with the plaintiffs in the anti-voucher lawsuit that the private school voucher movement undermines democracy. 鈥淧ublic schools are among the most democratic institutions that you can think of because they accept everybody, regardless of their language, their socioeconomic status, their gender, who their parents are, their immigrant status. Because that鈥檚 what inclusion is. And when you have these outside groups attack public schools, they鈥檙e really attacking that democratic institution.鈥�

School report cards

The latest round of released by DPI showed students test scores continuing to improve after the disruption of the COVID-19 pandemic.

None of Wisconsin鈥檚 school districts is rated as 鈥渇ailing鈥� in the latest assessments and 94% of districts meet or exceed  expectations. But critics say DPI is setting the bar too low. Will Flanders of Wisconsin Institute for Law & Liberty told : 鈥淲hile DPI may tout there has been an increase across the board, we still have districts like Milwaukee where proficiency rates are less than 20% and somehow that seems to be meeting expectations.鈥�

Public school student proficiency rates for 2022-23 were better than in 2020-21 and 2021-22. But they still seem low:  38.9% were proficient in English language arts and 37.4% were proficient in math. Students participating in the state鈥檚 Private School Choice Programs, however, had even lower proficiency rates of 22.1% in English language arts and 17.9% in math in 2022-23.

Student assessment scores are only one factor in determining district report card scores, a spokesperson for DPI explains. For districts with high percentages of low-income students, growth is weighted more significantly than achievement 鈥� a .

鈥淥ur public education system should be about getting every kid what they need 鈥� in the way they need it 鈥� in order to achieve success,鈥� Underly said.

In announcing the latest assessment data, DPI pointed to a that found Wisconsin鈥檚 performance standards in reading and math were among the highest in the nation, corresponding to higher levels of proficiency as measured by the National Assessment of Educational Progress (NAEP).

Big financial challenges for public schools

Still, schools face big challenges, particularly those with large numbers of low-income and special education students and English language learners. The biggest challenge, Underly said, is revenue.

After more than a decade of school funding that , and a less than 30% state reimbursement for special education 鈥� a mandatory cost that is eating up school districts鈥� budgets, driving deep cuts in other programs, public school advocates with the latest state budget.

Gov. Evers had adopted DPI鈥檚 proposals in his own budget, including a big increase in the state reimbursement for special education from less than 30% to 60%, lifting local revenue limits and providing a total funding increase of $2.6 billion. The Legislature stripped that down to $1 billion, and left 40% of school districts with less funding this year than they had under the previous, zero-increase budget.

Remaining hopeful part of the job

Despite the existential challenges facing Wisconsin public schools, including the elimination, next year, of the cap on enrollment for voucher schools, Underly said she has a lot of hope for 2024.

鈥淲hen we talk to kids, especially the ones that remember COVID 鈥� middle school, high school kids 鈥� they have a lot of hope for the future.鈥�

She is already working on her next budget proposal, which will include teacher recruitment, increasing funding for mental health and, once again, an increase in the state鈥檚 special education reimbursement, as well as programs including free meals that address poverty.

鈥淲e need to get kids what they need, so that they can be successful and making sure that they鈥檙e not hungry is really critical for them to be able to focus and concentrate,鈥� she said.

鈥淚 think it鈥檚 important that we continue this hopeful outlook because that鈥檚 what our schools need,鈥� Underly added. 鈥淥ur schools don鈥檛 need to be attacked. Our students don鈥檛 need to be attacked. So just supporting our schools, supporting our students and supporting that hope is part of supporting their education.鈥�

is part of States Newsroom, a network of news bureaus supported by grants and a coalition of donors as a 501c(3) public charity. Wisconsin Examiner maintains editorial independence. Contact Editor Ruth Conniff for questions: info@wisconsinexaminer.com. Follow Wisconsin Examiner on and .

The survey, conducted by the research and public relations firm Global Strategy Group, was commissioned by the Bill & Melinda Gates Foundation. Many of its themes 鈥� including the necessity of math skills in both promising career paths and the demands of everyday life 鈥� by the Foundation over the past six months, after its leaders announced a major refocusing of resources toward post-pandemic math recovery last fall. Standardized tests like the National Assessment of Educational Progress have shown steep declines in math achievement during the COVID era.

Angela Kuefler, a partner at Global Strategy Group, said on a call with reporters that existing methods of teaching math were regarded as 鈥渦nengaging, outdated, not connected to the real-world experiences that kids are having or will have.鈥� 

鈥淧arents especially鈥on’t feel like their kids are getting fully adequate preparation for what they will need to succeed later in life,” Kuefler said.

Asked to choose among a list of academic subjects, 60 percent of the survey sample (over 730 teachers and 800 parents) named math as either 鈥渆xtremely important鈥� or 鈥渧ery important,鈥� more than for any other discipline. Ninety-three percent of parents (and 94 percent of teachers) also agreed that 鈥渨hen students succeed in math, they are more likely to succeed later in life.鈥�

But math was also the most frequently named subject that needed 鈥渦pdating and improvement鈥� in how it is taught, according to respondents; 56 percent listed math, far more than career and technical education (41 percent), social studies and history (36 percent), English (35 percent), science (31 percent) and the arts (14 percent).

In the main, complaints about math education related to its applicability to the world that students live in. While over 60 percent of the survey鈥檚 participants said that math should be relevant to the real world, just 21 percent said that it currently is. The subject was similarly described as less useful and engaging than it ideally would be, while sizable minorities of parents and teachers described our current approach to math as complicated, boring, useless and irrelevant.

Some of those sentiments were echoed in accompanying focus groups that Global Strategy conducted with parents of varying racial and socioeconomic backgrounds. Participants cited by the Foundation described their children鈥檚 math coursework as 鈥渁 chore, instead of a challenge,鈥� noting that compelling instruction 鈥渄oes not reach all students.鈥� 

The influential philanthropy has already embraced the task of overhauling math education, awarding large grants to organizations like the and the that aim to make math more culturally relevant to disadvantaged and minority students. Some of those efforts have also come under criticism from some conservatives, that the project could weaken math standards and harm student outcomes.

But Bob Hughes, the Foundation鈥檚 director of K鈥�12 education, cited the 鈥渁bysmal鈥� data on student learning loss as evidence that more dramatic interventions were necessary.

鈥淧arents, teachers, and the general public see a disconnect between the math education they believe our young people need to thrive and the one that students are actually experiencing in too many classrooms,鈥� Hughes said. 鈥淧arents and educators point to a solution: Making math education more relevant for students and more connected to the real world.鈥�

But amid the carnage, one feel-good story emerged. Los Angeles, the nation鈥檚 second-largest school district, appeared to have accomplished a reading miracle, with eighth-grade scores jumping an incredible nine points.

Peggy Carr, the nation鈥檚 top testing official, called the district鈥檚 performance one of the few 鈥渂right spots鈥� on the National Assessment of Educational Progress, otherwise known as 鈥淭he Nation鈥檚 Report Card.鈥�

The media took note. and trumpeted the development, one that Alberto Carvalho, the district鈥檚 superintendent, said 鈥渂odes very well for our L.A., and is really a testament to our strategy.鈥�

Almost as soon as the news sank in, however, a host of naysayers emerged to pop L.A.鈥檚 balloon. Among wonks on edu-Twitter, the skeptic鈥檚 case went something like this:

• Nine-point jumps on NAEP are extremely rare.
• It happened in a huge urban district and during an educationally punishing pandemic, when most school systems鈥� scores tanked or, at best, stayed flat.
• NAEP is typically harder than most state assessments. Yet L.A. scores on the national test eclipsed those on California鈥檚 Smarter Balanced assessment, where the district鈥檚 eighth-grade readers inched forward a barely perceptible half a percent. 

Speaking for many researchers, Robin Lake, director of the Center on Reinventing Public Education, took to Twitter to say the results

At first, they didn鈥檛 sit right with the test鈥檚 administrators either. That鈥檚 why prior to NAEP鈥檚 release, the National Center for Education Statistics conducted extensive internal reviews and called on outside experts to crunch the numbers to understand what they were seeing.

鈥淭he L.A. results 鈥� certainly caught our attention, because they stood out as different from the nation as a whole and the other鈥� school systems that take NAEP, said Daniel McGrath, branch chief of the assessment division at NCES. Twenty-six districts participated in the Trial Urban District Assessment, which allows for comparisons across the nation鈥檚 large urban districts. 

The researchers 鈥渢urned over many additional rocks鈥� to back up a result they knew would raise eyebrows, he said. The American Institutes for Research conducted a separate analysis of the 2022 scores. 

The results offer the kind of frustrating nuance researchers are known for, confirming neither the unmitigated triumph of the district鈥檚 cheerleaders nor the worst-case scenarios of the skeptics.

After performing a series of statistical checks and balances, McGrath said L.A.鈥檚 schools saw a 鈥済enuine increase in student performance.鈥� Because of all the interest, NCES plans to add some technical notes to its explaining factors that affected this year鈥檚 results in Los Angeles.

A 鈥榩iece of the puzzle鈥� 

L.A. district staff participated in NCES鈥檚 Oct. 11-13 pre-release workshop to discuss the results. And on Oct. 19, the NCES team held a separate meeting on Zoom with Carvalho to answer questions about the extra steps it took to verify the data. 

The researchers explained that one factor in the district鈥檚 performance was the sample of schools in which students were tested. , the last time the nation鈥檚 students took NAEP, L.A.鈥檚 high-performing, affiliated charter schools weren鈥檛 included. This year, they were.

Including those scores 鈥� from five of the 60 schools in the sample 鈥� accounted for roughly two points in the nine-point climb in eighth-grade reading, McGrath said. That鈥檚 鈥渙ne piece of the puzzle.鈥�

Affiliated charters are district schools, but operate with some flexibility. They perform significantly higher than district schools overall, and this year鈥檚 state test scores show just how much. Forty-two percent of the district鈥檚 eighth graders met or exceeded state standards in English, compared to 66% in the district鈥檚 , according to 社区黑料鈥檚 review of .

Since the scores鈥� release, Carvalho the district鈥檚 efforts to provide summer school, tutoring and internet connections during remote learning. 鈥淗ow about those NAEP scores!鈥� he said during public comments Friday. In addition to the nine-point increase in reading for eighth graders, he touted the district鈥檚 fourth-grade reading scores, which went up two points compared to an average three-point decline in other districts. And he offered a pointed rebuke to those he described as 鈥渋nteresting fellows鈥� on social media who 鈥渨ork at places like Harvard.鈥�

鈥淭hey do nothing but study what we do and write about it,鈥� he said. 鈥淭hat data is accurate, and to assert anything other than that is to diminish the impact that our teachers, our principals, our support staff and 鈥� our students produced.鈥�   

Carvalho doesn鈥檛 need to be reminded of NAEP鈥檚 reputation as the 鈥済old standard鈥� in testing. He鈥檚 a of the National Assessment Governing Board, which sets policy for the test.

While a nine-point increase is rare, leaps in that range are not unheard-of, according to examples provided by NCES. Boston Public Schools, for instance, saw a 10-point increase in fourth-grade math between 2003 to 2005. And eighth-grade reading scores in the District of Columbia schools jumped eight points between 2011 and 2013. 

Demographic shifts

Some experts noted that Los Angeles has also experienced major demographic shifts in recent years that would affect student performance. Immigration in Los Angeles County, and Carvalho has often noted that officials can鈥檛 account for thousands of . 

Between the 2019-20 and 2021-22 school years, district enrollment fell by a staggering 10%. This factor deserves more attention, said Thomas Dee, a Stanford University economist and education professor. 

鈥淭he composition of that enrollment changed by a variety of student traits and the impact of such changes on LAUSD’s recent test scores is far from clear,鈥� said Dee, who has been tracking pandemic enrollment trends. 

NCES officials thought of this, too.

It turns out that demographic changes didn鈥檛 affect the overall increases. In fact, students in each racial and ethnic group made gains between 2019 and 2022. The district鈥檚 Hispanic student enrollment fell during that time period while enrollment of white students increased. But the addition of the affiliated charters effectively canceled out the impact of that population shift, McGrath said.

NCES also considered one group of students who didn鈥檛 participate in this year鈥檚 NAEP 鈥� those who were still learning remotely in spring 2022. Last year, the district鈥檚 virtual school, City of Angels, included 76 eighth graders. If they were all lower-performing students and if they had been included in the sample, it鈥檚 possible the district鈥檚 scores would have been lower. But to drag it down even one point, those students would have had to score 20 points below the district鈥檚 average, McGrath said.

The NAEP results also didn鈥檛 jibe with typical comparisons to state tests. Usually, districts on state assessments than on NAEP. 

In this case, the overall percentage of Los Angeles students meeting or exceeding standards on the Smarter Balanced test declined since 2019. In eighth-grade English language arts, however, there was a miniscule increase 鈥� from 41.14% to 41.74%.

The discrepancy could be chalked up to the fact that the state test includes a writing section and NAEP does not, said Andrew Ho, an assessment expert at Harvard and part of the research team behind a new that uses NAEP and state test data to calculate learning loss. 

In of why these discrepancies occur, the team notes that it鈥檚 also possible for the students who took NAEP to include more high performers than the larger state sample. 

鈥楽till a full two years behind鈥�

The district鈥檚 results confused some education advocates in Los Angeles as well. They expressed doubt that the strategies Carvalho points to were robust enough to spark such an increase.

鈥淣one of it was happening at the scale any of us were hoping it would,鈥� said Hannah Gravette, regional vice president for Los Angeles with Innovate Public Schools, an advocacy organization that works with families in low-income communities. 

The district is offering and homework help, for example, but she said the type of in-person, consistent tutoring known to is still in the early stages.

But the superintendent who was in charge when the pandemic hit said he is not surprised by L.A.鈥檚 testing windfall. 

鈥淵ou can鈥檛 point to one thing. We did everything possible, and I don鈥檛 know of any other school district in the country that did,鈥� said Austin Beutner, listing factors such as the district鈥檚 extensive COVID testing program, the 140 million it provided to families and delivery of . 

鈥淲e sent trucks around the country to Apple stores to take inventory out of the back because they were closing,鈥� he said. And hiring aides and reading specialists to run breakout rooms over Zoom allowed students to get more individualized attention than some would have received even in a traditional classroom, he added.

Ray Hart, executive director of the Council of Great City Schools 鈥� the network of 78 urban school systems that helped launch the district assessment program 鈥� said looking further back than 2019 helps put the district鈥檚 performance in context.

The affiliated charters were part of that 2017 NAEP sample, but not in 2019, when the NAEP score in eighth-grade reading dropped from 254 to 248. McGrath said that just like charters didn鈥檛 account for the full increase in 2022, their absence wasn鈥檛 the only reason for the decline in 2019. Eighth-grade reading scores fell nationally that year, not just in Los Angeles.

Comparing apples to apples 鈥� 2017 to 2022 鈥� the difference in eighth grade reading is much smaller, a three-point increase from 254 to 257 on NAEP. This relatively flat performance is more in line with that of many districts this year, which both Hart and McGrath view as positive, considering the pandemic鈥檚 crushing impact nationally. 

鈥淭he interventions during the pandemic and after are starting to show signs that they are paying off,鈥� Hart said.

Regardless of the reason for the district鈥檚 bump, some advocates remain unimpressed.

鈥淚t is really not an improvement, but a return to pre-pandemic performance, which means students are still a full two years behind,鈥� said Alicia Montgomery, CEO of the Center for Powerful Public Schools, a nonprofit supporting school improvement efforts in the district. 

鈥淭hree points is nothing over a five-year period,鈥� she said. 鈥淚f there was no pandemic, students would have gained far more.鈥�

The prospect makes the federal official in charge of measuring student progress a bit anxious. 

鈥淭he likelihood that the scores would be anything but down is pretty small,鈥� said Peggy Carr, commissioner of the National Center for Education Statistics. 

While performance among the lowest-scoring fourth- and eighth-graders on the National Assessment of Educational Progress was falling well before the pandemic, Carr predicted, 鈥淚t鈥檚 more than likely we’re going to see the bottom drop even more.鈥�

Known as the Nation鈥檚 Report Card and a chief 鈥渂arometer鈥� of educational achievement in the U.S., the congressionally-mandated NAEP is the only assessment with results broken down by gender, race and socioeconomic status that can be compared across all 50 states. As such, it is a major gauge of achievement gaps that are likely to have grown larger since COVID鈥檚 arrival.

Carr鈥檚 predecessor, James 鈥淟ynn鈥� Woodworth, described this year鈥檚 administration as “the most important NAEP assessment that鈥檚 ever been done in the history of this country.鈥� In a wide-ranging interview with 社区黑料, Carr fleshed out why, explaining that the pandemic had added layers of 鈥渘oise鈥� that could make results harder to interpret.

鈥楾rue change鈥�

Even the most superficial alteration in a student鈥檚 testing experience can throw off their performance. In the , researchers determined that a change in the color of the ink on the test booklets contributed to an otherwise unexplained drop in reading performance for 9- and 17-year-olds. In 2002, the mistake was accidentally repeated with a random sample of students, and again, scores dropped. 

But the pandemic has exploded the universe of possible variables: The sample of test-takers includes masked and unmasked students, as well as smaller groups. Social distancing and other changes in the environment could also affect student performance.

鈥淚t makes me a little nervous about what we鈥檙e going to see, and how I鈥檓 going to be able to separate out what is noise and what is true change in students鈥� performance,鈥� she said. 

At the same time, collecting those results has been far from easy. From staff quitting due to illness to schools rescheduling because of students in quarantine, this round of testing is unlike anything the center has faced in the past. 

鈥淚’m getting notices every day that people are quitting or people have 鈥� caught COVID in the schools,鈥� Carr said. 

In December, there were 3,560 NAEP staff members in the field. More than 850 have quit, with over half of those leaving in December and January as Omicron started to spread, according to NCES. 

Because of COVID鈥檚 lingering interference, Carr said she was pleasantly surprised schools haven鈥檛 pulled out of the assessment. Only one district, Fresno Unified in California, opted not to participate in the Trial Urban District Assessment, which provides results for more than two dozen districts nationwide. 

Nonetheless, state and district chiefs have already expressed concern about whether Carr can guarantee the validity of the results.

鈥淭hey said, 鈥楶eggy, I鈥檝e got 900 vacancies. I have people who normally teach art teaching some academic subject,鈥欌�� she said. 

To the doubters, she emphasized that this year鈥檚 tests include the same items used in 2019, which will further give the public a 鈥渟olid trend line鈥� through the pandemic years, she said. 

NAEP, she stressed, is 鈥渟till the standard by which other large-scale assessments judge themselves, and even in the context of COVID that has not changed.鈥�

But because of the impact of the pandemic, it might seem as if this year鈥檚 results are setting a new 鈥渂aseline,鈥� Carr said. A baseline, which technically refers to official changes in the test, is the starting point researchers and policymakers use to track student performance over time.

鈥淚t鈥檚 a new day in many ways,鈥� Carr said. 鈥淗ow tests are being administered, how students are being taught and how they learn in schools today is a little different than it was before COVID.鈥� 

Despite those challenges, NCES鈥檚 responsibility is to maintain the public鈥檚 trust in NAEP as an accurate measure, said Andrew Ho, an education professor at Harvard University and a former member of the National Assessment Governing Board, which sets policy for NAEP.

Carr must help parents and educators understand how the pandemic has affected fourth- and eighth-graders鈥� math and reading achievement, he said. 鈥淚t鈥檚 not a new baseline if we do our job right,鈥� he said. 鈥淚t is a decline.鈥�

He and Carr said the urban district results will be especially valuable when viewed against the backdrop of school closures at those sites. 

鈥淭here are always policy differences; they just haven鈥檛 been so confounded with historic health issues,鈥� Ho said, adding that it鈥檚 inevitable the results will become fodder for political arguments over how leaders responded to the pandemic. 鈥淓veryone likes to attach a policy story to NAEP results.鈥�

In addition to the reading and math tests, which were delayed a year because of school closures, NCES is testing eighth graders in civics and history and 9-year-olds as part of its long-term trend study. Nine-year-olds were also tested in 2019, which will allow NCES to provide pre- and post-pandemic results. Data from three years ago showed stagnant performance in both reading and math, except for girls, whose math scores dropped five points. Now researchers will be able to see how students in that age group, who were in first or second grade when schools shifted to remote learning, are performing. Next year, 13-year-olds will be assessed.

A in the Senate, introduced this month, proposes that NCES add a new component to measure the long-term impact of COVID on a representative sample of students.

It could be much harder, however, to see how U.S. students fared during the pandemic compared to their peers in other countries. While states and districts generally participate in all non-mandated NAEP tests, such as those in history, economics and technology, Carr struggles to get an adequate sample 鈥� 350 schools 鈥� for the Program for International Student Assessment and other global comparisons. 

School leaders are bombarded with requests to participate in surveys and an optional assessment can feel like one more burden. When Betsy DeVos was education secretary, Carr asked her to recruit schools for the international assessment. Michael Casserly, who led the Council of the Great City Schools and pushed for the urban assessment results, also helped.

鈥淲hen Betsy DeVos was here, we had her calling schools, and we got Mike Casserly, who’s a good friend of mine, calling schools and we barely made it,鈥� Carr said. 鈥淚t’s a hard sell. I’ve got to figure out another way to develop a relationship with the stakeholders on the ground and make it worth their while to participate.鈥�

Cloud-based tests and AI scoring 

As Carr prepares to analyze this year鈥檚 NAEP data, she鈥檚 also overseeing a modernization of the program, which has been 鈥渇ast-tracked鈥� by the pandemic, she wrote in a recent , co-authored with Lesley Muldoon, executive director of the assessment鈥檚 governing board.

Future tests will be cloud-based and downloaded to districts鈥� own devices. And beginning in 2024, NCES will no longer hire 3,500 to 4,000 administrators to deliver devices with the assessment to schools. That model, which prevented the center from conducting mandated tests in 2021, typically costs about $62 million. 

While some field staff will still be on site, using local administrators could save $22 million, according to released last week.

Also in 2024, NCES will begin using to score students鈥� essays. In January, the center announced four winners of a competition who showed AI can score an essay with 88% accuracy compared to trained individuals, Carr said.

鈥淚t may be good enough with a little bit of tweaking,鈥� she said, adding that the center will still have human scorers in 2024 to remain 鈥渟cientifically defensible.鈥�

The center will also continue running its monthly 鈥� the result of an early Biden administration to produce data on the impact of the pandemic. The survey tracks the percentages of students in in-person, hybrid and remote learning and has expanded to add questions on , quarantines and mental health. 

The project has pushed the center toward a quicker turnaround 鈥� something the governing board and state and local leaders would like to see with NAEP as well. 

鈥淚f I don’t have to put together the full-blown report card with all the bells and whistles, maybe I can get it out faster,鈥� Carr said. 鈥淏ut I’m not going to cut short the statistical analysis that I need to make sure we can stand behind the data. I鈥檒l put asterisks on it. I’ll caveat it, and then 鈥� whatever it says, I’m going to report it.鈥�

The High School Transcript Study, from the National Assessment of Educational Progress, or NAEP, shows that high school graduates鈥� overall scores in math declined between 2009 and 2019, while science scores remained flat. 

Only students taking the most rigorous math courses 鈥� including precalculus or higher 鈥� scored at the proficient level. But even their average score declined from 189 to 184 over the 10-year period. During that same decade, the typical high school senior鈥檚 graduating GPA rose .11 points to 3.11, or roughly a B average.

Peggy Carr, commissioner of the National Center for Education Statistics, which administers the NAEP program, suggested that the content of high school courses is not always 鈥渁s advertised.鈥�

鈥淎lgebra I is not Algebra I just because it鈥檚 labeled Algebra I,鈥� she said during a Tuesday call with reporters, but noted that the reasons behind the mismatch are complex. During a separate interview with 社区黑料, she added, 鈥淲e do have evidence from prior investigations that these courses may say they’re doing one thing, but they鈥檙e suspect about the rigor.鈥�

The good news, she said, is that more students of all backgrounds are taking higher-level courses 鈥� often because schools require them for graduation. She also noted that the average scores of Black students taking calculus have increased, from 161 to 177.

The lack of alignment between NAEP results and student performance in high school is not a new phenomenon. Officials reported the same trend in 2009, Carr said.

GPAs have increased for all racial groups over time, but since 1990, the gaps between Black and white students and Hispanic and white students have increased. The study also shows that students are earning their highest GPAs in career and technical courses or in those described as 鈥渙ther鈥� 鈥� not in the core academic courses measured by NAEP. 

The findings, based on a sample of 14,300 graduates from 1,400 public and private schools, follow a series of NAEP results that point to sagging academic performance for the nation鈥檚 students. Data released last fall showed disturbing declines among 13-year-olds in both reading and math between 2012 and 2020. And the gaps between the highest- and lowest-scoring students have grown over time. More students, however, are taking tougher courses in high school. The percentage taking 鈥渕id-level鈥� or rigorous courses, including Advanced Placement and International Baccalaureate courses, has climbed since 2000 from 46 percent to 63 percent.

Some say the mismatch between students鈥� grades and NAEP performance reflects that students earn credit for tasks that don鈥檛 necessarily reflect learning. Others argue the traditional A-F grading system is and that grade inflation is rampant: Higher grades might boost students鈥� confidence and increase graduation rates, some experts say, but leave them less prepared for college.

Seth Gershenson, an associate professor of public policy at American University, who studied in 2018, said the pandemic has probably exacerbated grade inflation, but added he hasn鈥檛 seen any recent data on the issue. 

鈥淟earning decreased, but a variety of pressures likely kept grades from dipping too hard,鈥� he said. The current NAEP study only reflects scores and GPAs prior to the pandemic.

Carr said her team is already considering what adjustments they鈥檒l have to make when they conduct the next transcript study in 2024, since it will focus on students whose education was disrupted by the pandemic, when many students completed courses online. Most districts across the country shifted to policies that kept grades from slipping below what they were when schools shut down in March, 2020. Some students also had opportunities during the 2020-21 school year to raise their lowest grades. 

In Ohio鈥檚 Oberlin City Schools, near Cleveland, high school history teacher Kurt Russell said there was some pushback from teachers at his school when administrators decided to give all students an A or a C on the work they submitted once schools shut down. 

But even before the pandemic, he said he noticed a shift toward allowing students to make up assignments that were significantly past due. 

鈥淚n the past, it was a 0 in the gradebook. Now I see a lot of teachers giving full credit for assignments that are very tardy,鈥� said Russell, one of four current finalists for National Teacher of the Year. His policy is to knock off a letter grade for each day an assignment is late. 

鈥淚 think we still need to hold our students accountable,鈥� he said.

Gershenson also studied teachers鈥� in 2020 and found that students learn more when teachers are tougher. 

鈥淭eachers who set a higher bar for a good grade had students who went on to learn more, even after they left that teachers’ class,鈥� said Michael Petrilli, president of the conservative Thomas B. Fordham Institute, which published the study. 鈥淪tudents learned to exert more effort. They inferred that their teacher held high expectations for them. Teachers face a ton of pressure to give easy A’s. Those that don’t are real heroes.鈥�

Non-achievement factors 

Some experts who study grading policy note that teachers traditionally consider a lot more than a student鈥檚 academic work when assigning final grades.

Matthew Townsley, an assistant professor of educational leadership at the University of Northern Iowa, said there鈥檚 a 鈥渨idening disconnect鈥� between GPAs and standardized assessments because course grades reflect a lot of non-achievement factors, such as attendance and assignment completion 鈥� regardless of the quality of student performance on  assignments.

The NAEP study, he said, could help strengthen the case for a movement called 鈥� giving students credit for what they actually learn and often more than one chance to learn it.

Some educators consider the practice more equitable because submitting assignments early or racking up extra credit points might be easier for students with high-speed Wi-Fi at home and access to private tutors, but can be an ongoing struggle for those in lower-income families. 

While were moving toward such a grading system before the pandemic, interest has spread as educators look for methods that don鈥檛 unfairly disadvantage students in poverty. 

鈥淚 believe schools seeking to separate achievement from non-achievement factors in their grading were not only well positioned to assess and communicate learning during the pandemic,鈥� Townsley said, 鈥渂ut also to communicate learning in the future that better aligns with NAEP and other external measures.鈥�

Those are among the updates to a 鈥渇ramework鈥� for the reading assessment that the governing board in charge of the National Assessment of Educational Progress approved Thursday. The unanimous vote represented a significant shift since May, when some members of the panel bitterly opposed a version they thought overly emphasized issues of equity and fairness for students taking the test.

Over the summer, former Mississippi Gov. Haley Barbour, chairman of the National Assessment Governing Board, and Alice Peisch, vice chair, brought together members to hash out their differences and finalize what one observer called a 鈥渇orward-looking鈥� approach to measuring students鈥� reading skills.

鈥淵ou worked extraordinarily hard to build consensus across an array of perspectives, ensuring an update for NAEP Reading that everyone here can stand behind,鈥� Lesley Muldoon, executive director of the governing board, said during the hybrid meeting, with some members assembled in Tysons Corner, Virginia, and others joining remotely. 鈥淚t would have been really easy for you all just to sort of hit pause after the May meeting and say, 鈥楲et鈥檚 just keep the old framework.鈥�

The board passed a that resulted from those summer meetings, stating its commitment to equity, but dropping some of the lengthy passages on the topic in an earlier draft. Patrick Kelly, a board member and history teacher in South Carolina, who led a committee working on the document, said he felt confident that none of the members feel 鈥渢he final product is 100 percent reflective of their personal views. That鈥檚 also what consensus is about.鈥�

The board has had a long history of finding common ground despite political differences. Andrew Ho, an assessment expert at Harvard University and former member of the board, said agreement on this issue is 鈥渋n everyone’s best interest.鈥� Approval of the document means the National Center for Education Statistics, which administers the congressionally mandated testing program, can move ahead with developing and piloting assessment questions. With last year鈥檚 school closures putting students roughly half a year behind in reading, according to end-of-year results, future tests will also gauge the extent of the pandemic鈥檚 impact on learning.

鈥淚 think the gap is going to be wider between the racial and ethninc groups,鈥� Peggy Carr, acting commissioner of NCES, told the board, noting the data showing racial differences in remote and in-person learning last school year. 鈥淚 think the implications are not good for achievement.鈥�

In May, the majority of board members were ready to approve a draft with detailed explanations for how elements in the digital test 鈥� such as student writing samples and pop-up hints for some words or concepts 鈥� could improve understanding for students who might be unfamiliar with those terms because of their language and cultural backgrounds.

But a few members, especially Russ Whitehurst, argued that highlighting those features of the test at a time when the nation has been arguing over issues of race and discrimination would prove divisive. Whitehurst, a senior fellow at the Urban Institute, even offered his own draft, eliminating most references to 鈥渟ociocultural鈥� factors that influence how students comprehend what they read.

The board 鈥渋tself almost came unglued鈥� over the issue, Chester Finn Jr., president emeritus of the conservative Thomas B. Fordham Institute, wrote in his latest on the saga. But he called the new version 鈥渁 masterful, praiseworthy piece of work鈥� and 鈥渁 coherent, forward-looking and intelligible approach to the modern assessment of students鈥� reading prowess and comprehension.鈥�

Other changes in the new framework include reporting students鈥� reading skills within content areas 鈥� literature, social studies and science. The existing version only provides the categories of literature and 鈥渋nformational texts.鈥� Kelly said unlike in 2004, the last time the document was updated, literacy skills are now 鈥渋nfused鈥� into academic standards.

In addition, results will break out performance levels for former, current and non-English learners. This change will 鈥渟hed light on any progress 鈥� or lack thereof 鈥� that might be detectable in the group of former English learners,鈥� the document says.

But even with his glowing review, Finn, a former chair of the board, has some lingering complaints about the issue that led to the board members鈥� dispute 鈥� the purpose of those digital hints and nudges that are intended to motivate students to do their best on an exam that never affects their grade. In the real world, readers don鈥檛 get that kind of help, Finn wrote.

鈥淜eep in mind that reading, that most fundamental of subjects, is not in good shape in America today,鈥� he wrote. 鈥淲e need to bend every effort to teach reading better and ensure that kids learn it well. We don鈥檛 need to conceal their deficiencies.鈥�

When students read, do their personal and cultural backgrounds determine how they understand the text, or are the skills and knowledge they pick up in the classroom more important?

That鈥檚 a question currently dividing the government body that oversees what is known as The Nation鈥檚 Report Card.

The dispute centers around the role of equity in a lengthy document that could determine how the federal government designs future reading tests. The authors filled it with references to 鈥減op-ups鈥� or short videos defining words or terms some students might not recognize, such as a talent show.

But others on the board view the emphasis on removing bias and increasing fairness as divisive in the current political climate.

The board won鈥檛 vote until August, and the updated 鈥渇ramework鈥� wouldn鈥檛 affect tests until 2026. But the disagreement over the testing program, often referred to as the 鈥済old standard鈥� of student assessments, comes amid a nationwide push to improve education for children of color. And it reflects a longstanding tension between those wanting to make the program relevant to contemporary issues and others primarily concerned with accurately measuring how students鈥� reading skills have changed over time.

Such conflicts aren鈥檛 new for the National Assessment Governing Board, which Congress established in 1988 as an independent and nonpartisan overseer of the National Assessment of Educational Progress. But remote meetings necessitated by the pandemic have made those differences harder to resolve than they would have been in person, said Andrew Ho, an assessment expert at Harvard University and former board member.

He 鈥渇ewer unanimous votes ahead鈥� when he left last year.

鈥淣AEP has always been above the fray, relative to everything else,鈥� he said in an interview, adding that if issues before the board are becoming more controversial, that鈥檚 because 鈥渆verything else is, too.鈥�

A panel of 18 curriculum experts completed a draft of the framework on April 21. Board members in favor of that version have directed criticism at colleague Russ Whitehurst, who this month offered his own , cutting most references to 鈥渟ociocultural鈥� factors and words such as equity, bias and fairness. He emphasized other areas that influence reading comprehension, including curriculum and instruction, teacher quality and students鈥� background knowledge.

During a May 7 virtual meeting, board member Christine Cunningham, a professor at Pennsylvania State University, said Whitehurst鈥檚 version is 鈥渄ecidedly not a document that I can agree with.鈥� The terms he removed, she added, 鈥渁re the words that experts use. I don鈥檛 think they should be omitted or treated as objectionable.鈥�

鈥楳easuring what鈥檚 changing鈥�

The governing board decided in 2019 to , which hasn鈥檛 changed since 2004, to reflect current research on how students make sense of what they鈥檙e reading and that students now read a variety of both digital and traditional materials.

The challenge is 鈥渢rying to measure what is relevant while also measuring what’s changing,鈥� Ho said. 鈥淭he way we read now is not the way we read 20 years ago.鈥�

The panel writing the document drew on the findings of a 2018 , 鈥淗ow People Learn II,鈥� which concluded that learning depends not only on the knowledge students gain in the classroom but on the cultural experiences they bring to the process.

The experts鈥� draft focused heavily on how the pop-ups and other 鈥渦niversal design elements鈥� in a computer-based test would make the reading passages more accessible for all students. The elements are essentially hints that might explain a word or phrase not all students would recognize.

Such features are intended to pique students鈥� interest or prepare them for what they are about to read. They could even motivate students to try harder on a test that doesn鈥檛 affect their grade, said Mark Miller, a board member and a math teacher in Colorado Springs, Colorado.

The elements are not meant to affect the material students are actually being tested on or give one group of students an advantage over another, according to an explanation from the board鈥檚 鈥渢echnical advisory committee.鈥�

That assumption is open to debate. During a public comment period in 2020, the board held eight webinars and collected over 2,600 comments. Some groups supported the emphasis on social and cultural factors, but others were concerned adding too many elements 鈥渕ight yield a test that all students would be able to pass,鈥� according to of the changes.

In response to the public and other board members, the panel toned down references to the role of culture and background in the document, and opted not to add more of the elements in the test.

Board member Julia Rafal-Baer, chief operating officer of Chiefs for Change, said she鈥檚 concerned that if critics of earlier drafts don鈥檛 understand how the panel addressed their concerns, 鈥淲e are going to completely lose both sides of the aisle.鈥�

鈥楢n undesirable position鈥�

But those changes didn鈥檛 satisfy Whitehurst, past research director at the U.S. Department of Education during the second Bush administration and a senior fellow at the Urban Institute. For example, he deleted a sentence that reads, 鈥淩esearch has shown that a student鈥檚 background, language, and experience is important in how they interpret assessments.鈥�

He wrote that he doesn鈥檛 want the board members or the testing program to be forced to accept 鈥渁 particular point of view of what is most important in learning to read.鈥�

At another meeting last Thursday, he said he didn鈥檛 think the edits he wants 鈥渟hould be onerous or upsetting to people unless they have an agenda that is not on the surface.鈥� He added that while the panel鈥檚 draft included several references to fairness, it鈥檚 鈥渕issing explanation on what we find unfair that we hope to correct.鈥�

Board member Eric Hanushek, an economist at Stanford University, shares Whitehurst鈥檚 concerns. He said the board has gotten itself in 鈥渁n undesirable position鈥� and has conflated the assessment of reading with the goal of improving reading.

鈥淭he latter is not in our charge,鈥� he said.

Work-related communication

One additional layer to the governing board鈥檚 recent discussions is an internal letter from former Mississippi Gov. Haley Barbour, board chair, to other members. The letter, according to two conservative observers who have read it, said members can no longer communicate with each other or outsiders regarding official matters unless they share those messages with board staff. The chat function in Zoom has been disabled to further prevent side conversations.

, president emeritus of the Thomas B. Fordham Institute, and , director of education policy studies at the American Enterprise Institute, suggested the letter was meant to stifle criticism of the reading framework document.

But Stephaan Harris, a spokesman for the board, said the letter complies with guidelines for work-related communications that now apply to all special government agencies and employees.

鈥淚t had nothing to do with the framework or any specific issue, action or policy,鈥� Harris said, adding that Finn鈥檚 and Hess鈥檚 鈥減ieces incorrectly conflated the two, so that caused some unfortunate confusion.鈥�

Finn, a former chair of the governing board, also linked the controversy to the larger and wrote that an updated view of reading would extend beyond the assessments and 鈥済radually percolate through the entirety of K鈥�12 education in America.鈥�

鈥淚n the 鈥榬eal world鈥� of reading outside school, there are no supports,鈥� he said in an email. 鈥淓ither you possess the [vocabulary] and knowledge that you need to understand what you’re reading or you don’t. And if you don’t, you likely won’t understand stuff.鈥�

But Patricia Anders, a University of Arizona reading professor, said it鈥檚 important to focus on children鈥檚 strengths. She was among those who wrote letters last week to the governing board rejecting Whitehurst鈥檚 edits.

鈥淓veryone knows that culture and language influences learning to read,鈥� she said. 鈥淭eachers don’t 鈥榝ix鈥� language and culture. They work with what the learner brings to the learning-to-read context.鈥�

Ho, at Harvard, said he thinks the board members will eventually work out their differences because NAEP is such an important 鈥渂arometer鈥� of student progress.

鈥淎t the end of the day, everyone needs NAEP,鈥� he said. 鈥淣o matter what happens with this framework, I’m confident that people on both sides of this issue will come to support what comes out of it.鈥�