Share

This is the first of a four-part series on recent math and reading achievement in the IES REL Midwest states of Illinois, Indiana, Iowa, Michigan, Minnesota, Ohio, and Wisconsin. It’s also part 26 of a larger series comparing results of states on their own tests with their results on NAEP (see other posts from the series at my Substack webpage).

First, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 4th-grade reading results for the Midwest states:

The first point I take from this table is that the Midwest states do not vary as much in the reading skills as you might think if you looked only at their state tests. This has been generally true for neighboring states across the nation. If you looked only at the state tests, you’d think kids read much better in Iowa (72 percent proficient), for instance, and Ohio (64 percent proficient) than in Indiana (42 percent proficient) and Illinois (38 percent proficient). However, according to the National Assessment (NAEP), all four states are quite similar, none have particularly high percentages of students NAEP would consider proficient, and Indiana actually may be the top Midwest state in terms of 4th-grade reading proficiency. In 2024, the Midwest states ranged from 34 percent NAEP Proficient in Indiana to 25 percent in Michigan (most of the Midwest states hovered around 30 percent NAEP Proficient).

The second main point I’d take away is that none of the states except Indiana made real reading progress between 2022 and 2024 in 4th grade. The Midwest states are not alone in this, of course. But, even if Illinois found a 5-percentage-point gain in percent proficient on its state test and Ohio a 2-point gain, these did not show up in the National Assessment.

Let’s look at the charts and see what they say. Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

We’ll start with Indiana and Minnesota, which have the best-looking charts.

Indiana

Minnesota

Both Indiana and Minnesota show exposed magenta at the tops of their charts. This suggests some narrowing of discrepancies in skills between the highest and lowest skilled students. The exposed magenta at the top shows an increase in the percentage of students with average skills. In the case of Indiana, this is accompanied with quite a bit of exposed purple down the left slope and a long stretch of magenta down the right slope. This indicates a decrease in the percentages of students with lower skills and an increase in the percentages of students with higher skills.

In the case of Minnesota, it looks like mixes of purple and magenta on each slope and a bit of magenta at the very bottom left. This suggests maybe a slight decrease in the percentages of students with higher skills and slight increase in percentages of students with lower skills, including at the very lowest skill levels.

The rest of the states are on “purple-to-the-right, magenta-to-the-left” patterns, which means larger percentages of higher-skilled students in 2022 and lower-skilled students in 2024. See the charts for Illinois, Iowa, Michigan, Ohio, and Wisconsin below.

Illinois

Iowa

Michigan

Ohio

Wisconsin

None of this is unexpected given the disruptions to schooling during these students’ kindergarten and first grade years. Perhaps the most concerning charts are those for Iowa and Michigan, which show distinct, large increases in the percentages of students with lower skills.

In Sum

Indiana’s progress in 4th-grade reading between 2022 and 2024 is remarkable given the pandemic experiences of these 2024 4th-graders during key years in their literacy development. The other Midwest states struggled more or less as we’ve seen inmost other states across the nation. The next post will look at reading achievement in 8th grade for the Midwest states.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Illinois, Indiana, Iowa, Michigan, Minnesota, Ohio, and Wisconsin). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point for Iowa and Wisconsin and with 3 points for the other states. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Wisconsin’s is based on 2024 data only, because they had changed assessments between 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 25 of a series comparing results of states on their own tests with their results on NAEP (see other posts from the series at my Substack webpage). It’s also the last of four posts featuring the IES Appalachia region states: Kentucky, Tennessee, Virginia, and West Virginia (see spelled out acronyms in the appendix).

Below, first, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 4th-grade mathematics results for the Appalachia states:

As we’ve seen in math at grade 8 and reading at both grades 4 and 8, Virginia’s proficiency standards are out of whack with the other Appalachia states and with NAEP. You can see this in the “Threshold for proficient” column for state tests. Virginia’s threshold is much lower than the other states and NAEP. As a result, Virginia reported 71 percent of its 4th-graders proficient in math—much higher than states like Kentucky and Tennessee—while NAEP reported Virginia with 40 percent NAEP Proficient—which is about the same as Kentucky and Tennessee.

A second big takeaway I want to point out from this table is how large the gains were for these states and how robust the consistency between the state tests and NAEP (with somewhat of an exception in Virginia). NAEP estimated gains of at least 6-percentage-points in proficiency rates and 4 points in average score for each of the states except Virginia, and Virginia’s estimated gains were still sizable.

Let’s turn to the charts. Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Kentucky and Tennessee

I’ve put the Kentucky and Tennessee charts together because they’re similar

Kentucky

Tennessee

These are just the kind of charts most states would want to see. You see purple on the left, indicating larger percentages of lower-skilled students in 2022 and magenta on the right, indicating larger percentages of higher-skilled students in 2024. Kentucky’s chart is particularly remarkable because there’s a sizable patch of exposed purple at the very bottom left, meaning that Kentucky reduced the percentage of students with the very lowest skills and their lowest-skilled students in 2024 were more highly skilled than their lowest-skilled students in 2022.

Virginia

Virginia’s chart shows gains: the exposed magenta at the top indicates an increase in the percentage of students in Virginia’s mid-skill range and the long slope of magenta down the right indicates increases in the percentages of highly skilled students, including at Virginia’s highest skill levels. The magenta at the bottom left, however, shows a small increase in the percentage of Virginia 4th-graders at the very lowest skill levels.

West Virginia

West Virginia’s chart is something like what we saw in 8th-grade math in Tennessee and in the District of Columbia in math at both grade 4 and grade 8. That is, you see an increase in disparities between higher and lower-skilled students but it’s not because of an accumulation of students with very low skills (which is often the case). Instead, it’s because of the tremendous growth in the proportion of students scoring at the highest levels. To see the increased disparities, look at the widening and shortening of the figures between 2022 and 2024. But you can also see that the widening is all about growth in higher skilled students in that thick magenta slope down the right side of the chart. Top-scoring students in West Virginia in 2024 scored higher than the top students in 2022.

In Sum

This wraps up the tour of recent National Assessment and state assessment results in the Appalachia states. In math at grade 4, the Appalachia states experienced strong recoveries from the pandemic—particularly so in West Virginia, Kentucky, and Tennessee. Up next, the Midwest.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Kentucky, Tennessee (including some 2026 results), Virginia, and West Virginia). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point for Virginia, about 2.5 points for Tennessee and West Virginia, and just under 3 points for Kentucky. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 24 of a series comparing results of states on their own tests with their results on the National Assessment (NAEP). See other posts from the series at my Substack webpage. This is also part three of four posts featuring the IES Appalachia region: Kentucky, Tennessee, Virginia, and West Virginia (see spelled out acronyms in the appendix).

In these posts, first, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient on NAEP. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 8th-grade mathematics results for the Appalachia states:

As with all these posts on the Appalachia states, my first takeaway from this table is that Virginia’s proficiency standards are out of whack with the other Appalachia states and with NAEP. Virginia reported half their 8th-graders were proficient in math in 2024 at the same time Tennessee said just one in three of their 8th-graders was proficient. NAEP reported both states with about 30 percent NAEP Proficient.

My second takeaway is that, according to NAEP, all the Appalachia states, with the exception of Virginia, had relatively strong recoveries in the 2022 to 2024 post-pandemic period. Nationally (not shown in the table), the percent NAEP Proficient in 8th-grade math rose 2 percentage points between 2022 and 2024 and the average math score didn’t grow at all. But in the Appalachia region, NAEP estimated gains in percent NAEP Proficient and average scores for each of the states, except Virginia.

What do the NAEP charts say? Let’s see. Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Kentucky

Kentucky had somewhat more growth in NAEP Proficient than on their state test, and you can see in the chart why NAEP found a bigger gain. See the exposed purple wrapping right around the top of the figure and down the right slope to the charcoal tower and then magenta the rest of the way down that right hand slope. Purple represents larger percentages in 2022 than 2024 and magenta represents larger percentages in 2024 than 2022. That charcoal tower represents the NAEP Proficiency cutpoint. So, what the figure shows is higher percentages of students at the skills above the NAEP Proficient threshold in 2024. The green tower represents the Kentucky proficiency cutpoint and you can see purple on either side that cutpoint: there just wasn’t as much gain across the Kentucky threshold. I also want to note the exposed magenta near the bottom left. This indicates a worrisome increase in the percentages of Kentucky 8th-graders with very low math skills.

Tennessee

Tennessee has the “purple-to-the-left, magenta-to-the-right” pattern that indicates unmitigated growth. However, it’s not the classic pattern, which would have the magenta mountain as tall as or taller than the purple mountain, indicating growth in skills without a growth in the disparities of skills. The Tennessee pattern shows an increased disparity in skills but in the most positive way: through a considerable growth in the the percentages of students with the very highest skills. We saw this pattern also in DC in math at both grade 4 and grade 8 and something like it in 8th-grade reading in Tennessee.

Virginia

Virginia’s state test showed a 4-percentage-point gain in proficiency. It’s hard to see how that could be in this chart of results from the National Assessment. The exposed purple up the left slope and down the right slope, combined with a large patch of magenta on the lower left slope, indicates a state in decline in math skills.

West Virginia

In West Virginia, NAEP estimated a 3-percentage-point gain in percent NAEP Proficient but just a 1-point gain in average score. You can see in the chart how that might be. There is a good deal of blue to the left of the charcoal tower that represents the NAEP Proficiency threshold and magenta to the right, indicating an increase in percent NAEP Proficient. However, you can see, also, some magenta at the lower left. This suggests West Virginia experienced an increase in both higher and lower-skilled students and a decrease in students with skills near the West Virginia average (see the purple at the top of the chart).

In Sum

As with 8th-grade reading, Tennessee stood out among the Appalachia states in recent progress in 8th-grade math. Also, as with 8th-grade reading, Virginia appeared to struggle over the 2022 to 2024 period in 8th-grade mathematics. Next is the final post in the Appalachia set: mathematics at grade 4.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Kentucky, Tennessee (including some 2026 results), Virginia, and West Virginia). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. I did not try to estimate thresholds for Tennessee and Virginia because enough 8th-graders took an end-of-year assessment in algebra rather than general mathematics that NCES could not line up the general test and NAEP populations sufficiently well to estimate a proficiency threshold on the state test—and so I didn’t either. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 23 of a series comparing results of states on their own tests with their results on the National Assessment (NAEP). I’m in the midst of four posts on the four states in the IES Appalachia region: Kentucky, Tennessee, Virginia, and West Virginia (see spelled out acronyms in the appendix and see other posts from the series at my Substack webpage).

In each post, first, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient on NAEP. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 8th-grade ELA/reading results for these states:

As with 4th-grade reading, Virginia stands out in this table because of the disconnect between what Virginia’s test says about its proficiency rate and what NAEP finds. Virginia reported 72 percent of its 8th-graders proficient in English/Language Arts in 2024, a much higher rate than the other Appalachia states reported on their state tests and much, much higher than what NAEP found. NAEP found 29 percent of Virginia 8th-graders NAEP Proficient in 2024. That’s about the same rate as in Kentucky and Tennessee. The disparity between what Virginia’s state test and NAEP say is probably because Virginia’s threshold for proficiency is so low (see the “Threshold for proficient” columns).

Tennessee, on the other hand, stands out because it’s doing so well in 8th-grade reading. Both, Tennessee’s state test and NAEP found 3-percentage-point gains in percent proficient between 2022 and 2024. Nationally, average NAEP scores in reading at 8th grade declined between 2022 and 2024, and the other Appalachia states besides Tennessee all experienced stagnation or declines as well (although West Virginia’s state test found an increase).

The charts below tell us more, at least as far as NAEP goes. Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Kentucky

Kentucky reported a 3-percentage-point decline in proficiency in 8th-grade reading between 2022 and 2024, while NAEP reported no change. You can see in the chart how both of these things could be true. Look at the green tower running right through the middle of that purple Kentucky mountain. That’s my estimate of Kentucky’s proficiency threshold placed on the NAEP scale. The exposed purple at the top means there was a smaller percentage of students scoring in that range in 2024 than there had been in 2022; the exposed magenta down the left slope means larger percentages of lower-skilled students in 2024 than 2022. That’s how you could find a 3-point decline on the Kentucky test. But all this activity (except a bit of magenta at the lower right) happens to the left of the charcoal tower. That means it happened below NAEP’s threshold for NAEP Proficient. And so, NAEP showed no change. So, although these chart is not good news for Kentucky, it’s not nearly as bad as you would think if all you saw was the Kentucky test results.

Tennessee

As expected given the table above, Tennessee has the best-looking 8th-grade reading chart among the Appalachia region states. There is a concerning bit of magenta at the lower left—meaning increased percentages of Tennessee students with very, very low skills. However, most of the chart is “purple-to-the left, magenta-to-the-right,” meaning increased percentages of students with higher skills.

Virginia

Virginia did not have a good 2022 to 2024 period in 8th-grade reading. This was true for 4th-grade, as well. Here, you see quite a bit of magenta at the very bottom left of the figure. This means a large increase in the percentages of students at the very lowest Virginia skill levels. You can see also some purple at the upper right, meaning a decrease in the percentages of students in the middle and above middle skill levels.

West Virginia

Despite West Virginia reported a 2-percentage-point gain in 8th-grade ELA Proficiency between 2022 and 2024, NAEP found a bit of decline in reading performance in West Virginia. The exposed purple at the top and upper right and the magenta down the left hand slope suggest a slide—a decrease in the percentage of students with middle-to-upper-middle skills and an increase in the percentage of students with lower level reading skills.

In Sum

Tennessee bucked the national trend between 2022 and 2024 and showed gains in reading skills among its 8th-graders, and Kentucky, if you take NAEP into account, may not have done as badly as you’d think from Kentucky’s own state results. Virginia and West Virginia looked more like the nation as a whole, showing slippage in the reading skills of 8th-graders between 2022 and 2024.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Kentucky, Tennessee (including some 2026 results), Virginia, and West Virginia). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 2 NAEP scale points for all the states. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 22 of a series comparing results of states on their own tests with their results on NAEP (see other posts from the series at my Substack webpage). It’s also the first of four posts focused on the four states in the IES Appalachia region: Kentucky, Tennessee, Virginia, and West Virginia (see spelled out acronyms in the appendix).

In each post I, first, show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 4th-grade reading results for the Appalachia states:

The first thing that leaps out to me from this table is how out of whack Virginia is with the rest of the Appalachia states. They reported 73 percent of their 4th-graders proficient in English/Language arts in 2024. The other three states all reported about half their kids proficient. Are Virginia 4th-graders really reading that much better than in Kentucky, Tennessee, West Virginia? No. According to the National Assessment, Virginia is very much like Kentucky and Tennessee in 4th-grade reading—about 30 percent NAEP Proficient (West Virginia lags the other Appalachia states).

The second big thing I take away from this table is that the Appalachia states mostly bucked the national trend between 2022 and 2024, which saw declines in 4th-grade reading proficiency. These 4th-graders in 2024 had been kindergartners when the pandemic hit. Many were learning crucial literacy skills at a time when schooling was disrupted and very little in-person instruction was occurring. The nation as a whole saw a 2-percentage-point decline in percent NAEP Proficient between 2022 and 2024. The Appalachia states appeared to have held steady (according to NAEP) or actually made strides (according to their state tests).

Let’s look at the charts to try to learn more about what was going on. Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Kentucky

Kentucky has a good-looking chart for a year when most of the country struggled in 4th-grade reading. See the magenta at the lower right? That indicates increased percentages of students with higher skills in 2024, compared with 2022. In fact, some kids in Kentucky in 2024 scored higher than any Kentucky kids in 2022. This was quite unusual in the U.S. (or probably anywhere in the world) in 2024.

Tennessee

Tennessee’s chart tells a more complicated story but, in some ways, a positive story. The magenta at the top of the chart suggests higher percentages of students in 2024 scoring at the Tennessee average level. The large exposed purple patch on the left slope indicates lower percentages of low-to-mid-skilled students in 2024 than 2022. This translated in part into larger percentages of very low-skilled students (the magenta at the lower left) but also possibly to that increase in the percentage of students at the average level. Also, the magenta on the right indicates larger percentages of readers at the highest skill levels in 2024.

Virginia

Virginia’s chart is not so great. On the positive side, you can see some exposed purple down the left slope, near the bottom. This suggests smaller percentages of very low-skilled students in 2024 than in 2022. But the rest of the chart is mainly purple-on-the-right, magenta-on-the-left, indicating larger percentages of higher-skilled students in 2022 than 2024 and larger percentages of lower-skilled students in 2024 than in 2022.

West Virginia

West Virginia’s chart tells a mixed story, one we’ve seen in other states and in other subjects since the mid-2010s: increasing disparities between higher and lower-skilled students. Here, we see purple at the top and magenta down each slope, indicating smaller percentages of mid-skilled students in 2024 than 2022 and larger percentages of, both, higher and lower-skilled students. However, it looks like more magenta on the right than the left, meaning a bigger gain in the percentage of higher-skilled students than lower-skilled students.

In Sum

By and large, the Appalachia states seemed to have fared better than the nation as a whole in 4th-grade reading over that two-year period between 2022 and 2024, when the kindergartners of 2020 reached 4th-grade. This was especially so in Kentucky, which saw unmitigated gains. The next post turns to reading in 8th-grade.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Kentucky, Tennessee (including some 2026 results), Virginia, and West Virginia). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 2 NAEP scale points for all the states. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 21 of a series comparing results of states on their own tests with their results on NAEP (see other posts from the series at my Substack webpage). It’s also the last of four posts featuring the IES REL Southeast states of Alabama, Florida, Georgia, Mississippi, North Carolina, and South Carolina (acronyms are spelled out in the appendix). First, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 4th-grade mathematics results for these states:

In the last post, when we looked at 8th-grade math, I said my point 1a from the table was that the states reported higher proficiency rates than NAEP did. The same point 1a holds for 4th-grade math. In 4th-grade math, the states ranged from 58 and 57 percent proficient (in Florida and North Carolina) to 38 percent in Alabama based on their own tests and thresholds for proficiency. NAEP estimated the Southeast states ranged from the mid-40s to high 30s in percent NAEP Proficient. My point 1b at 8th grade was that North Carolina really stood out as the highest achieving Southeast state, at least on NAEP. Here, in 4th-grade, the states are grouped more closely and Florida leads the group with 45 percent NAEP Proficient in 2024.

My point 2 for 8th-grade math was that NAEP was less sunny than the states’ own tests about progress in the Southeast states between 2022 and 2024. That’s not really so at 4th-grade. NAEP found about the same progress as most of the Southeast states, with the possible exceptions of Alabama (NAEP found more progress) and Mississippi (NAEP found a bit less progress). Gains in percent NAEP Proficient between 2022 and 2024 ranged from 4 points in Florida and Georgia to 10 points in Alabama. In a good year for 4th-grade math across the nation, the Southeast states were among the leaders.

Let’s look at the charts and see what more they say. Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

The charts are pretty similar to each other, so I will show all the charts and then discuss afterwards.

Alabama

Florida

Georgia

Mississippi

North Carolina

South Carolina

1. Across the board, these charts show more purple to the left and more magenta to the right, which is just what you want to see. That pattern means larger percentages of students with lower skills in 2022 (purple) and larger percentages with higher skills in 2024 (magenta).

2. Mississippi’s chart is what I imagine nearly every state would like to see. The figure is taller in 2024 than 2022. The magenta peak actually escapes the top of the chart. What this suggests is a narrowing in the discrepancies between the higher and lower-skilled students—students are piling up at the middle Mississippi skill level. Also, there’s a pretty thick layer of exposed magenta, all down the rightward slope of the chart. That means, in addition to an accumulation of mid-skilled students, Mississippi was piling on the higher-skilled students, as well, including to the very highest skill levels. Even if NAEP did not show quite the progress Mississippi found on its own test, NAEP found impressive progress.

3. North Carolina maintained the height of its distribution between 2022 and 2024 and moved it to the right. That is, it kept the same proportion of students at the middle North Carolina skill level even as that middle skill rose. There’s a slight bit of magenta at the lower left in North Carolina, indicating a small increase in percentages of students at the lowest skill levels but the vast majority of the movement of students was toward the higher skill levels.

4. Georgia seems not to have made the same sort of unmitigated progress as in the other Southeast states. Georgia’s chart displays a pattern we’ve seen regularly across the country since the mid-2010s. This pattern shows increases in percentages of students at both the low and high ends. The purple at the top and magenta down both slopes indicates greater disparities between higher and lower-skilled students—a decrease in the percentages of students near the Georgia average and increases in the percentages of higher-skilled students (good) and the lowest-skilled students (not good at all).

In Sum

Across the Southeast region, states made progress consistent with or greater than the rest of the nation in 4th-grade mathematics between 2022 and 2024. The exception was Georgia, where they did not see a real post-pandemic bounce among their lowest-skilled students. This concludes a 4-part series on recent progress in math and reading in grades 4 and 8 in the Southeast. Up next, Appalachia.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Alabama, Florida (including its 2026 results), Georgia, Mississippi, North Carolina, and South Carolina). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point, except for Florida (1.9 points) and Mississippi (1.5 points). Also North Carolina was off by 12 points but that seems to be because NCES used NC’s level 4 and above for its threshold and I used level 3 and above, which is what the Education Data Center used. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. For Florida and Georgia I used 2024 data only, because they had changed their assessments between 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 20 of a series comparing results of states on their own tests with their results on NAEP (see other posts from the series at my Substack webpage). It’s also part three of four posts featuring the IES REL Southeast states of Alabama, Florida, Georgia, Mississippi, North Carolina, and South Carolina (acronyms are spelled out in the appendix).

In these posts, first, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient on NAEP. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 8th-grade mathematics results for these states:

Point 1a that I would take from this table is that the range across the Southeast states in proficiency rates on state tests is more about their definitions of proficiency than the actual math skills of their 8th-graders. According to the state tests, the states ranged in 2024 from 55 percent proficient in Mississippi and 54 percent in Florida to 22 percent proficient in Alabama. But NAEP shows all the states closer to Alabama’s finding than Mississippi’s and Florida’s. All the states are around 30 percent NAEP Proficient, or below. You can see why in the “Threshold for proficient” column for the state tests: Alabama’s threshold (297 points on the NAEP scale) is about the same as the NAEP Proficient threshold (299 points), while Mississippi’s is at 264 points on the NAEP scale. Florida’s threshold is probably pretty low too. I couldn’t estimate a threshold for Florida (and North Carolina) because of the way they administer and report their general math and algebra end-of-year exams in 8th grade; but Florida’s threshold is probably pretty low because NAEP finds just 21 percent of Florida’s 8th-graders NAEP Proficient, which is pretty close to Alabama’s 18 percent.

Point 1b is North Carolina is the standout Southeast state in 8th-grade math. NAEP found 31 percent of North Carolina’s 8th-graders NAEP Proficient, which is higher than the neighboring states, and North Carolina’s average NAEP score stands above the others, as well.

My point 2 is that NAEP found less progress for a few of the Southeast states than the states found with their own tests. Sometimes you can see why in the charts below and sometimes you can’t. This discrepancy was also apparent in reading at grades 4 and 8, and we’re going to see it in the next post on math at 4th-grade, too. Most striking is Mississippi, where the state test showed an 11-percentage-point gain in proficiency rate between 2022 and 2024 and NAEP found a much smaller gain. As I said, discrepancies show up for Mississippi in the other grades and subjects too. The differences are so big that I wondered if Mississippi had changed their test or their proficiency cutpoints. They changed their cutpoints in 2025 but I’ve not seen anything that said they changed them back in 2024. However, when we get to the charts I think you’ll see that, at least for 8th-grade math, it looks like Mississippi may have made gains that just didn’t get picked up on NAEP because Mississippi and NAEP have different thresholds for proficiency. On the other hand, we’ll see that different proficiency thresholds probably can’t explain the Alabama-NAEP differences.

Anyway, let’s look at the charts. Quickly on how to read them: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficiency would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Alabama

I’ve put the Alabama, Florida, and Georgia charts together because they’re similar.

Florida

Georgia

These charts do not look good at all. They are “magenta-to-the-left, purple-to-the-right” charts, which is not what you want to see. The magenta-to-the-left shows increased percentages of 8th-graders with low math skills in 2024, including increases at the very lowest scores. The purple at the top means decreased percentages of students in 2024 near the state average. And, the purple down the right means decreased percentages of higher-skilled students too. This is just not good.

We can’t compare NAEP results with state results for Florida and Georgia, because they both changed their tests in some way between 2022 and 2024, which means you can’t compare their state test results between 2022 and 2024. However, Alabama found a 4-percentage-point gain in proficiency on its state test. There’s no clue (that I can see) in the Alabama chart suggesting why the Alabama state test and NAEP results are so different. For instance, Alabama’s threshold for proficiency is very close to the NAEP threshold and so it’s unlikely that kids crossed the Alabama threshold and not the NAEP threshold.

Mississippi

Mississippi’s is a really good looking chart and you can see how Mississippi could report a much larger gain in percent proficient than NAEP found. First, this is “purple-to-the-left, magenta-to-the-right,” which is what you want to see. It means there were increasing percentages of 8th-graders in Mississippi at the higher skills levels in 2024. This includes the very highest levels, where kids in 2024 scored higher than anyone in 2022. Look also at where Mississippi’s proficiency threshold is, relative to the purple and magenta patches and relative to the NAEP threshold. The Mississippi threshold (the green tower) is right through the top of that purple mountain, right at the average skill level for Mississippi in 2022 and to the left of the top of the magenta mountain, where the average skill level was in 2024. This indicates a much larger percentage of students above the Mississippi proficiency threshold in 2024 than 2022, which is what the Mississippi state test showed. The NAEP threshold (the charcoal tower) is well to the right of the Mississippi threshold. There’s plenty of magenta to the right of the NAEP tower but, still, the percentage crossing this threshold would be smaller than the percentage crossing the Mississippi threshold.

I’ve put the Carolinas together, too, because their charts look similar to each other (but, unfortunately, not similar to Mississippi’s).

North Carolina

South Carolina

The North and South Carolina charts suggest growing disparities in skills between the lowest and highest skilled students. We’ve seen this across the nation since around 2013: states and the nation continue to produce the same or higher percentages of highly skilled students but, at the same time, more and more students with very low skills. You can see this in the post-pandemic period between 2022 and 2024 in the Carolinas. The exposed purple at the tops of these charts indicate larger percentages of students with middle level skills in 2022 than 2024. The exposed magenta down the left and right slopes indicate larger percentages of students in 2024 with lower skills (left slope) and higher skills (right slope). This is especially pronounced in North Carolina. Remember, point 1b above was that North Carolina is the standout Southeast state in terms of average score and percent NAEP Proficient. This chart suggests North Carolina does not standout for all students; there’s a sizable and growing divide between the haves and havenots.

In Sum

What we saw in these charts is why it’s good to have them. We could see that Mississippi could have made a large gain in its proficiency rate in 8th-grade math even though the similar metric on the National Assessment (NAEP) didn’t show it. We could see also, in the other Southeast states, broader and less favorable developments that don’t show up in the state test results. Next week is the final post in the Southeast set: mathematics at grade 4.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Alabama, Florida (including its 2026 results), Georgia, Mississippi, North Carolina, and South Carolina). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 2 NAEP scale points, except for Alabama (2.7 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024; for Georgia, I used only their 2024 results because they changed cutpoints between 2022 and 2024. I did not estimate state test thresholds for Florida and North Carolina because the way that they report their general math and algebra data makes it difficult to align the state test and NAEP populations to estimate the location of the general math test threshold on the NAEP scale. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 19 of a series comparing results of states on their own tests with their results on NAEP (see other posts from the series at my Substack webpage). I’m in the midst of four posts on the six states in the IES Southeast region: Alabama, Florida, Georgia, Mississippi, North Carolina, and South Carolina (see spelled out acronyms in the appendix).

In each post, I first show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient on NAEP. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in measures of percent proficient or average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent results in 8th-grade English/language arts (ELA) on state tests and reading on NAEP for these states:

The first thing I take away from this table is that you’d come to a different conclusion about which states in the region have the top 8th-grade readers if you use states’ own tests than if you use NAEP. According to the state tests and their thresholds for proficiency, the top states seem to be Alabama, Florida, North Carolina, South Carolina, all with about 50 percent of their 8th-graders proficient in reading, and the states lagging are Georgia and Mississippi, with 45 percent and 42 percent, respectively. However, NAEP actually has Georgia at the top in the region, though with just 31 percent NAEP Proficient. The other states range all the way down to 21 percent (1 in 5!) NAEP Proficient in Alabama. This is because the rigor of the proficiency thresholds vary across the states and it is lower on all the state tests than on NAEP (though it’s also possible part of the difference between the states and NAEP is that the state tests are measuring something different from NAEP).

My second big takeaway is that, just as in 4th-grade reading in the Southeast, NAEP is less bullish than the state tests are on progress since 2022. This is true for all the states except Alabama and North Carolina (and maybe Florida, which had changed tests between 2022 and 2024). The differences are less stark than at 4th-grade but, pretty much across the board, the states found sizable gains in proficiency rates, while NAEP found essentially no gains. You can see this in the “Average score on NAEP” and the “Two-year gain” columns. The exceptions to this pattern are Alabama and North Carolina, whose state tests found similar lack of progress as NAEP found, and maybe Florida, which changed its assessment and therefore doesn’t have a state-test comparison between 2022 and 2024.

But let’s look at the charts and see what they say. How to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Alabama

Alabama (along with North Carolina) had state test results most similar to NAEP. They reported almost no change between 2022 and 2024. You can see in the chart the lack of change on NAEP. The 2022 and 2024 figures overlap almost perfectly, except for some exposed purple at the top and some magenta mostly down the left slope but also a tiny bit at the bottom of the right slope. This indicates a small increase in disparities in skills between the highest and lowest skilled students, mostly because of a decrease in the percentage of students with average skills for Alabama and an increase in the percentage of 8th-graders with lower skills.

Florida

As in 4th-grade reading, 2022-24 was not a good period for 8th-grade readers in Florida on the NAEP. This chart is just what you don’t want to see. Lots of exposed magenta on the left and lots of exposed purple on the right. There were considerably higher percentages of 8th-graders with very low reading skills in 2024 than 2022 and they were scoring lower than ever.

Georgia

The Georgia chart is interesting because you can see why Georgia may have found a 4-point gain in percent proficient and NAEP found almost no change in percent NAEP Proficient or in average score. The figures for the two years closely overlap—hence NAEP finding little change—but the differences in the figures occur right where the Georgia proficiency threshold stands and the movement looks in the direction of increased percent-Georgia-proficient. That is, the green tower, representing the Georgia proficiency threshold, runs through the purple patch at the top of the chart and you can see some magenta running down the slope to the right of that green tower. This would be consistent with an increase in the percentage of students reaching Georgia-proficient.

Mississippi

As with 4th-grade reading, Mississippi looks mostly unchanged in NAEP reading at 8th-grade and I don’t see in the chart why Mississippi would find a large gain in its proficiency rate. In 8th-grade reading, the figures for the two years overlap very closely and the differences that are apparent suggest an increase in the percentage of 8th-graders with very low reading skills.

North Carolina

North Carolina shows up as flat on their state tests in terms of percent proficient and on NAEP both in terms of percent NAEP Proficient and average score. But this chart shows that it happened in a way that you don’t want. The chart shows change that doesn’t register in percent-proficient and average-score metrics, and it’s not good change. This is an increase in the percentages of 8th-graders with very low skills (magenta at the bottom left) and a decrease in percentages of students with mid-level skills (purple up the left slope and the top). This can register in percent proficient and average score as no change but it’s really an increase in disparities between higher and lower-skilled students and one that is caused mainly by an increase in the percentage of very low-skilled students.

South Carolina

In South Carolina, like Georgia, you can see in the chart why the state test and NAEP found different trends between 2022 and 2024. There is exposed purple on the left slope and magenta at the top, including to the right of the green tower marking South Carolina’s proficiency threshold. This would be consistent with the gain in percent proficient found by South Carolina’s state test. There’s not much magenta to the right of the charcoal tower marking the NAEP Proficient threshold—and that’s consistent with NAEP finding no gain in percent NAEP Proficient.

In Sum

As in 4th-grade reading, when you look at National Assessment (NAEP) results alongside state tests in the Southeast states, progress in reading at 8th-grade between 2022 and 2024 does not look great. It looks a good deal worse than if you looked at the state test results alone. The next two posts turn to math in these states.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Alabama, Florida (including its 2026 results), Georgia, Mississippi, North Carolina, and South Carolina). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point, except for Florida (2.3 points) and Mississippi (1.9 points). Also North Carolina was off by 20 points but that seems to be because NCES used NC’s level 4 and above for its threshold and I used level 3 and above, which is what the Education Data Center used. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024; for Georgia, I used only their 2024 results because they changed cutpoints between 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 18 of a series comparing results of states on their own tests with their results on NAEP (see other posts from the series at my Substack webpage). It’s also the first of four posts focused on the six states in the IES Southeast region: Alabama, Florida, Georgia, Mississippi, North Carolina, and South Carolina (see spelled out acronyms in the appendix).

In each post I, first, show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests, as well as the percentage NAEP Proficient. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 4th-grade reading results for these states:

First, look at the big differences between states’ reports on percent proficient vs NAEP’s reports of percent NAEP Proficient, for all the southeastern states except Georgia. All the other states reported at least 50 percent of their 4th-graders were proficient readers in 2024. Georgia reported 37 percent proficient in 2024, and this was much more consistent with what NAEP found, not just for Georgia, but for all the Southeast states. NAEP had all the Southeast states right around 30-percent NAEP Proficient in 4th-grade reading.

Second, NAEP also reported smaller gains than the states did between 2022 and 2024 (again, with the exception of Georgia and maybe North Carolina). Remember, these 4th-graders were kindergartners when the pandemic hit in the spring of 2024 and so it would not be surprising to see declines between 2022 and 2024 in percent proficient. Instead, Southeast states were showing gains of 5 percentage points (Alabama), 7 points (South Carolina), and 13 (Mississippi). We’ll get to the charts next to try to understand this but I will tell you now the charts are not going to explain how Mississippi could have a 13-percentage-point gain in percent proficient while NAEP is estimating a 1-point gain.

Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Alabama

Alabama’s NAEP chart doesn’t suggest the 5-point gain in percent proficient that Alabama found on its own test. The chart shows not much change between 2022 and 2024 except a bit of widening in the distribution of skills. The purple at the top and magenta down each slope indicates more kids with skills near the Alabama average in 2022 and, then, more kids with both lower skills and higher skills in 2024. In the case of Alabama, that large magenta patch at the lower right means considerable growth in the percentage of Alabama kids with very high skills relative to the Alabama average. They did not have so much increase in the percentage of students at the very low end. It looks like the growth in percentages of lower-skilled students was not at the very lowest skill levels—which is better than we’ve seen in some states.

Florida

Florida introduced a new state assessment between 2022 and 2024 and so you can’t measure growth on the state assessment between these two years. However, NAEP did not change and it shows a really tough two years for Florida. This chart is almost all purple-to-the-right and magenta-to-the-left, which is not what you want to see. In this case, because the purple is at the top of the chart, it indicates a decrease in the percentage of middle-skilled students and an increase in the percentages of lower-skilled readers, including very, very low-skilled readers (but see also a little magenta at the lower right, indicating a slight increase in the percentages of the highest-skilled students). So, disparities in skills grew in Florida and they grew mainly because of much larger percentages of students with very low reading skills.

Georgia

Georgia was the state with state-test results most consistent with NAEP results. It’s also the state with the least apparent change between 2022 and 2024. The figures for 2022 and 2024 almost perfectly overlap with, most noticeably, some growth in percentages of students with relatively low reading skills.

Mississippi

As I said above, I can’t see in the NAEP chart how Mississippi made such a large gain in its state assessment. NAEP shows little change. There is some purple on the left and magenta on the right, indicating lower percentages of lower-skilled students and higher percentages of higher-skilled students in 2024. However, the figures overlap pretty closely, except at the top, where the purple peak indicates a reduction in the percentages of students near the Mississippi average. You’ll see in these charts that sometimes students move across the state threshold for proficiency (the green tower) and not the NAEP threshold (the charcoal towers), which could suggest growth picked up by the state test and not NAEP. But I don’t see that here. The growth in higher-skilled students (the magenta on the right) is to the right of both the Mississippi and the NAEP thresholds; the big reduction (the purple at the top) is below the NAEP threshold and above the Mississippi threshold and so, I’d expect NAEP to pick up growth not caught on the Mississippi test, rather than the other way around.

Still, this is not a bad chart for Mississippi. It looks more like growth than decline. Also, note how much more narrow the Mississippi figures are than the figures for the other Southeast states. This indicates smaller disparities between the highest and lowest skilled 4th-graders (though the disparities grew a bit between 2022 and 2024).

North Carolina

The North Carolina chart is a less extreme version of the Florida chart. Nearly all the exposed magenta is to the left and purple to the right, indicating growth in the percentages of students with very low skills.

South Carolina

South Carolina’s chart also shows declines in the percentages of middle-skilled students (the purple in the middle) and increases in the percentages of students with very low skills (as well as possibly small increases in the percentages of highly-skilled students). I don’t see in this chart how South Carolina found a 7-percentage-point gain in proficiency on its state test. As always, particularly with English/Language Arts and reading, it’s possible that South Carolina’s English/Language Arts test is picking up something different from NAEP’s reading-only test.

In Sum

The National Assessment (NAEP) data cloud the picture of recovery in 4th-grade reading we’d seen on some of the Southeast region’s state tests. In addition to reporting lower proficiency rates, generally, than the states reported, NAEP didn’t find nearly as large gains as were found on some of the state tests. It could be that the state tests were picking up something NAEP did not. After all, NAEP’s reading assessment doesn’t include writing, while the state ELA tests do. However, it’s also possible the state tests are overestimating real growth and, in some cases, missing a decline in students’ reading skills at 4th-grade.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Alabama, Florida (including its 2026 results), Georgia, Mississippi, North Carolina, and South Carolina). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point, except for Alabama and Florida (1.2 points, each). Also North Carolina was off by 14 points but that seems to be because NCES used NC’s level 4 and above for its threshold and I used level 3 and above, which is what the Education Data Center used. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Florida’s is based on 2024 data only, because they had changed assessments between 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is the last of four posts reviewing recent state test and National Assessment (NAEP) results in reading and math at grades 4 and 8 for the IES Mid-Atlantic region states of Delaware, DC, Maryland, New Jersey, and Pennsylvania (see spelled out acronyms in the appendix). This is also part 17 of a broader series region-by-region across the country (see other posts from the series at my Substack webpage). First, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient. I use the years 2022 and 2024, because they are the most recent with NAEP results in math and reading at grades 4 and 8. Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of recent 4th-grade reading results for these states:

First, note that the states are reporting higher proficiency rates on their own tests, with their own standards, than NAEP reports for its NAEP Proficient standard. On state tests, the Mid-Atlantic states ranged from 51 percent proficient in Pennsylvania to 35 percent in the District of Columbia. On NAEP Proficient, the states ranged from 38 percent NAEP Proficient in New Jersey to 26 percent in Delaware. This may, in part, be because the state ELA tests measure somewhat different skills from NAEP reading. For instance, the ELA tests measure writing skills. However, the bigger difference is probably because the states vary in their thresholds for proficiency and because they have lower thresholds than NAEP. Look at the “Threshold for proficient” columns. I’ve estimated that all the Mid-Atlantic states have lower thresholds for proficiency than NAEP in 4th-grade reading.

Second, the state tests and NAEP are saying pretty much the same thing about progress in the two post-pandemic years between 2022 and 2024. The exception is maybe Delaware, where NAEP seems a bit more positive about progress than the Delaware state test is. In fact, if you look at the “Average score” column, it looks like Delaware may have progressed as well as DC.

But let’s look at the charts and see what they say. Quickly on how to read the charts: Magenta represents 2024 and purple represents 2022. The green towers mark where I’ve estimated the states’ thresholds for proficient would be on the NAEP scale and the charcoal towers mark the NAEP Proficiency threshold. See more on how to read the charts in the appendix.

Delaware

There was not much change in the distribution of 4th-grade reading skills in Delaware between 2022 and 2024 but it was change in the right direction. Look at the way almost all the exposed magenta is to the right and all the exposed purple is to the left. That means higher percentages of higher-skilled students in 2024 than 2022. I’m not sure why this wasn’t picked up on the Delaware state test, which actually found a 1-point decline in percent proficient. It could be that most of the movement between 2022 and 2024 occurred at higher skill levels than the Delaware threshold, at levels closer to the NAEP threshold. It could also be differences in the content of the tests between the Delaware test and NAEP.

District of Columbia

Look at how squat the DC picture is compared with Delaware. We’ll see this again in Maryland. The DC chart peaked just over 8 percent in 2022 and actually dropped under 8 percent in 2022. That peak is the percentage of students at the middle skill level in a state. When you have shorter, wider distributions like you see in DC, it means you have large differences in skills between your highest and lowest skilled students. In DC, the distributions are short and wide and they’re getting shorter and wider over time. However, and this is a big however, in DC the widening is occurring because of an increase in the percentage of students with higher skills. We’ve seen this in DC in the other subjects and grades, too. It’s not ideal of course—it means DC is not reducing its percentage of lower-skilled students—but it still means they’re growing their percentages of higher-skilled students. In most cases where we see a widening of the distribution it’s largely driven by a growing population of lower-skilled students. Not so in DC.

Maryland

Maryland has a relatively squat distribution too (peak just above 8 percent) but it’s not getting squatter. And it’s moving to the right. Maryland is the standout state in the Mid-Atlantic region for 4th-grade reading between 2022 and 2024. This chart is just what you want to see: all magenta to the right all purple to the left. Maryland experienced both an increase in the percentages of highly skilled students and a decrease in the percentages of lower-skilled 4th-grade readers.

New Jersey

The New Jersey chart is what you often see when states’ distributions of skills widen: purple at the top and magenta down both slopes. What this means is that New Jersey saw growth among its highest skilled students but also an accumulation of students with very low reading skills. New Jersey had higher percentages of students scoring at the state’s highest levels in 2024; they even extended the range of NAEP scores higher than in 2022. However, they also had higher percentages of students at the state’s lowest skill levels and extended the range to ever lower skills in 2024.

Pennsylvania

Pennsylvania probably had the worst overall experience in 4th-grade reading among the Mid-Atlantic states. What stands out is the large patch of magenta down the lower left slope, representing increasing percentages of lower-skilled students. But there’s also exposed purple down the right slope all the way to the far right. What that means is that Pennsylvania had smaller percentages of students scoring at the highest levels in 2024.

In Sum

Three Mid-Atlantic states bucked the national trend toward lower reading skills in 2024 among 4th-graders. Delaware, the District of Columbia, and, especially, Maryland all experienced growth in reading skills where many states experienced declines in the wake of the pandemic. New Jersey appeared to have a more typical experience—growth in the proportion of students with high reading skills in 2024 but also growth in the proportion of students with very, very low skills. Pennsylvania appeared to struggle—growing proportions of lower-skilled students and a slight decline in the proportion of highly skilled students, as well.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Delaware, the District of Columbia, Maryland, New Jersey, and Pennsylvania). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point, except for Maryland (2.7 points), New Jersey (2 points), and Delaware (1.6 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the charts: The charts show the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 16 in a series that reviews recent state test results at grades 4 and 8 in math and reading in the context of National Assessment (NAEP) results. Region-by-region, I compare states’ results on their own tests with their results on NAEP. First, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests—because percent proficient tends to be the focus of state tests results (“proficient” or some other similar term)—as well as the percentage NAEP Proficient and states’ average NAEP scores. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8.

Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

In this post, we’re looking at grade 4 mathematics for the five states in the Mid-Atlantic region of the IES REL program: Delaware, the District of Columbia, Maryland, New Jersey, and Pennsylvania (see other posts from the series at my Substack webpage).

Here’s a table of recent 4th-grade math results for the IES REL Mid-Atlantic states:

The first thing I note in this table is that the range in percentage of 4th-graders proficient in math is tighter when you use NAEP as a common yardstick than when you use the states’ own tests and standards. According to the state tests, there’s a range of 21 percentage points between the highest and lowest proficiency rates. Pennsylvania reported 49 percent proficient in 2024 and DC reported 28 percent. NAEP shows the states performing much more similarly to each other. The range in percent NAEP Proficient is 11 points, between New Jersey at 44 percent and DC at 33 percent. This variation is probably because the states range a bit in their standards. DC and Maryland have relatively high thresholds for proficiency—as high or higher than NAEP’s—while Pennsylvania’s is a bit lower.

Second, both the state tests and NAEP found large gains in percent proficient between 2022 and 2024. Nationally, NAEP reported a 3-point gain in percent NAEP Proficient between 2022 and 2024. Delaware and DC posted 9-point gains in NAEP Proficient and big gains on their state tests. The other states all showed large gains in percent proficient on their state tests and all but Pennsylvania showed large gains in NAEP’s percent NAEP Proficient.

Third, New Jersey had a pretty sizable gain in percent NAEP Proficient but, if you look to the far right of the table, NAEP didn’t find much growth for New Jersey in terms of average scores. How does that happen? Well, it helps to look at the whole distribution of skills and not just the percent reaching proficient. So, let’s look at those charts.

I’m going to show all the charts first, because some are pretty similar, and then discuss below. The charts show the percentage of students at each skill level on NAEP from the lowest to the highest. The shapes tend to look like mountains or normal curves because there tend to be more students with average skills and fewer with either very high or very low levels of skills. Purple represents 2022 and magenta represents 2024. The green towers indicate my estimate of where the state threshold for proficient would land on the NAEP scale and the charcoal towers mark the NAEP Proficient threshold (for information on how to read the charts, please see the appendix).

Here are the charts:

Delaware

District of Columbia

Maryland

New Jersey

Pennsylvania

Comments:

1. These are charts you like to see: What you want to see in these charts is purple to the left and magenta to the right, because that indicates larger percentages of lower-skilled students in 2022 and larger percentages of higher-skilled students in 2024. And, that’s largely what you see in the Mid-Atlantic states—though it’s true that in nearly every state there is a bit of magenta on the left, too, indicating some increase in percentages of students with lower skills. However, with the exception of New Jersey and maybe Delaware, these are very small increases in percentages of lower-skilled students.

2. Delaware, Maryland, and, particularly, DC standout as making strides in moving the student population toward higher skills. Delaware and Maryland have kept more or less the same shapes, the same distributions of lower and higher-skilled, and moved them to the right, toward higher skills. DC has changed the shape of its distribution. In 2022, DC had a relatively narrow distribution, meaning relatively small disparities between lower and higher-skilled students, but the distribution was centered on relatively low skill levels. By 2024, the distribution had broadened. DC still had relatively high percentages of students at the very low end of skills but it had much larger percentages of students at the high skill end than they’d had before: As in 8th grade, skill disparities had grown but by extending the skills at the high end.

3. You can see why New Jersey and Pennsylvania made more progress with percent proficient than with average scores: In the case of New Jersey, there’s a good deal of magenta down that right slope and purple near the top left. That indicates increases in the percentages of students at or above the proficient range. However, see the magenta at the bottom left, as well: New Jersey’s gains at the higher end were mitigated by increases in the percentages of students at the very low end of skills. In the case of Pennsylvania, NAEP shows very little change between 2022 and 2024. However, Pennsylvania’s threshold for proficiency runs right through the highest part of the figures—the part with the largest proportion of students—and right through a purple patch at the top of the chart. In that area, relatively small changes in skills can produce large changes in percent proficient. It could be that Pennsylvania saw more gains at its threshold than at the NAEP threshold.

In Sum

As with most of the nation, the Mid-Atlantic states experienced gains in the math skills of 4th-graders between 2022 and 2024. Even in states like New Jersey and Pennsylvania that didn’t see large gains in percent NAEP Proficient (Pennsylvania) or average NAEP scores (both states), they appear to have made gains in the percentages of students reaching what the states defined as proficient. The other states—Delaware, Maryland, and, particularly, DC—experienced more widespread gains. We’ll wrap up the review of the Mid-Atlantic states with 4th-grade reading in the next post.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Delaware, the District of Columbia, Maryland, New Jersey, and Pennsylvania). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except Maryland (1.9 points) and Delaware (1.1 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 15 of a series reviewing recent state test results in the context of National Assessment (NAEP) results for math and reading at grades 4 and 8. It’s the second part discussing results for the IES REL Mid-Atlantic region states of Delaware, the District of Columbia, Maryland, New Jersey, and Pennsylvania (see the appendix for spelled out acronyms for IES, REL, and others; to see other posts from the series, go to my Substack webpage).

In each post, I first show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests based on each state’s definition of proficient (or whatever term the state uses). The table also includes the percentage NAEP Proficient and states’ average NAEP scores. I use the years 2022 and 2024 because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Next, I run through charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in measures of percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of recent 8th-grade math results for the Mid-Atlantic states:

There’s that 7 percent proficient for Maryland and 11 percent for DC in the second column. If you went only by states’ own tests and standards, you’d find that the Mid-Atlantic region ranged from 29 percent proficient in Pennsylvania to 7 percent in Maryland in 2024. However, NAEP found a range between 37 percent NAEP Proficient in New Jersey and 19 percent in Delaware. NAEP estimated 25 percent NAEP Proficient in Maryland and 20 percent in DC, more than three times what Maryland found on its own test and double DC’s state Test results.

I think the difference in what NAEP and the states are finding in terms of percent proficient is about the high standards for proficient in these states, at least relative to NAEP. Using methods similar to (but not as rigorous as) those NCES uses to estimate where states’ thresholds for proficiency fit on the NAEP scale, I estimated Pennsylvania’s threshold at 304 points (5 higher than the NAEP threshold) and Delaware’s at 292 points (7 points lower than NAEP). I couldn’t estimate them for DC, Maryland, and New Jersey because they all administer their general math and algebra assessments in a way that makes it too difficult to compare their state tests and NAEP and get a good estimate of state test thresholds on the NAEP scale. But their thresholds must be quite high because look at how much lower their reports of percent proficient are than NAEP’s.

NAEP and the state tests are similar, though, in the stories they tell about changes in percent proficient between 2022 and 2024. If you look at the two-year gain columns, you see similar findings for NAEP and the state tests, with the possible exception of DC where NAEP seems to be more positive about DC than DC is.

Let’s look at the charts across the skills distributions, so we can see what NAEP says about developments between 2022 and 2024 across all students, rather than just those crossing the proficiency threshold (for information on how to read the charts, please see the appendix).

Delaware

This is not a great chart for Delaware. What it shows is a widening in disparities between the skills of the highest and lowest-skilled students. Purple in this chart represents 2022 and magenta 2024 (the reddish purple is the overlap between 2022 and 2024). Purple at the top and magenta down the two slopes indicates larger percentages of middle-skilled students in 2022 and larger percentages of both lower-skilled and higher-skilled students in 2024. And there’s more magenta on the left than the right, meaning more growth in the percentage of lower-skilled students than higher-skilled.

District of Columbia

DC looks better. Like Delaware, DC shows increased disparities in skills. However, in DC, this is mostly because of tremendous growth in the percentages of higher, in fact the highest, skilled students. This is the shape I said is very unusual. You’ll often see states with figures that skew to the left, meaning there are lots of lower-skilled students and not so many higher-skilled. In DC, you see almost the opposite: a skew to the right, meaning higher percentages of students at DC’s highest skill levels than at their lowest. And, while both ends grew between 2022 and 2024, the upper end grew more.

Maryland

I said DC’s shape, skewed to the right, was unusual. And now here’s Maryland with a similar though less pronounced skew to the right, indicating relatively higher percentages of very high-skilled students, compared with very low-skilled. Here, in the period between 2022 and 2024, there appears to be similar growth at both the very low and very high skill levels. So, in the end, this is not a chart that’s showing much progress between 2022 and 2024.

New Jersey and Pennsylvania

I’m discussing New Jersey and Pennsylvania together because their progress between 2022 and 2024 is pretty similar, at least as shown in NAEP results.

Pennsylvania

In both cases, NAEP and the state tests showed substantial increases in percent proficient but NAEP found almost no change in average scores. Here, you can see why. Like every other Mid-Atlantic state, there’s purple at the top and magenta on the slopes, indicating increasing disparities in skill levels. In the cases of New Jersey and Pennsylvania, there are really large swathes of magenta down the right—large increases in the percentages of students at the higher skill levels, increases large enough that they registered as increases in the percentages of students reaching those high bars of state proficient and NAEP Proficient. However, on the other hand, there are sizable magenta patches at the lower left, indicating increased percentages of students at the very lowest skill levels. These kinds of patterns can produce increases in percent proficient and, at the same time, no increase in a state’s average score: the increases at the higher and lower skill levels balance out.

In Sum

The post-pandemic years between 2022 and 2024 saw increasing disparities in the math skills of 8th-graders across the Mid-Atlantic states. This is similar to the nation as a whole. On the one hand, in the Mid-Atlantic states, we saw states that made tremendous progress because they increased the percentages of kids reaching the state’s (and NAEP’s) proficient threshold. On the other hand, these states—with DC as a notable exception—experienced sizable increases in the percentages of students with the very lowest skills, as well.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Delaware, the District of Columbia, Maryland, New Jersey, and Pennsylvania). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. Also, I couldn’t do it for DC, Maryland, and New Jersey because they administer algebra tests instead of general math tests as end-of-year assessments for relatively large numbers of students, and so there NAEP and general math test populations don’t line up. For Delaware and Pennsylvania, I checked my results for 2022 against NCES’s published results and my results were 1 NAEP scale point or less different from the published result for each. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

See this short, readable discussion of issues associated with tracking change using percent proficient.

In this post, we’re looking at grade 8 reading for the five states in the Mid-Atlantic region of the IES REL program (see acronyms spelled out below in the appendix). Those states are Delaware, the District of Columbia, Maryland, New Jersey, and Pennsylvania. You can see all the other states reviewed so far at my Substack page.

Share

Here’s the table of recent 8th-grade ELA/reading results for these states:

If you look at the first four columns, you’ll see that the states report higher percentages of their 8th-graders as proficient than NAEP does. The states ranged from 53 percent proficient in New Jersey to 36 percent in the District of Columbia in 2024, using their own tests and standards for proficient. However, NAEP reported 38 percent NAEP Proficient in New Jersey, and then a range for the other states down to 23 percent in Delaware. The difference between NAEP and the state tests is probably largely a difference in standards, though it’s possible part of the difference is in the tests, because NAEP and state tests are somewhat different. For instance, the states all test English/language arts, including writing, as well as reading; NAEP reading assessments solely test reading.

I noticed in this table also that three of the Mid-Atlantic states have essentially the same standards for proficient: I’ve estimated DC, Maryland, and New Jersey all to have set their threshold for proficiency at 269 points on the NAEP scale; Delaware is at 263 points and Pennsylvania at 259 points. NAEP Proficient is set at 281 points on the NAEP scale.

Finally, if you look at those middle two columns that show trends in percent proficient, you’ll see they’re not so consistent between the state tests and NAEP for three of the states: Maryland reported a 4-percentage-point gain in percent proficient between 2022 and 2024, where NAEP reported no change in percent NAEP Proficient. New Jersey reported a 1-point gain and NAEP estimated a 3-point decline. And, on the other hand, Pennsylvania reported a 3-point decline in percent proficient and NAEP reported no change.

Sometimes it helps to see the charts that show movement across the whole distribution of achievement on NAEP, not just the percent crossing the proficient threshold. In this case it doesn't help, but let’s go to the charts, anyway.

Delaware and the District of Columbia

Let’s look first at the two states (I’m calling DC a state) with consistent state test and NAEP results: Delaware and the District of Columbia (I’m calling DC a state). To read these charts, the purple figures represent 2022 and magenta 2024; the reddish purple is the overlap between 2022 and 2024. The green tower on each chart represents where the state threshold for proficient is placed on the NAEP scale and the black tower shows the NAEP Proficient threshold.

Delaware

The lack of change in percent proficient in Delaware, as shown in the table, seems to mask a real decline in the reading skills of 8th-graders in Delaware. The reason this decline doesn’t show up in the percent proficient metric is probably because almost all the movement happened at skills lower than either the Delaware or NAEP proficiency thresholds. You see a good deal of exposed magenta at the lower left of the chart. Magenta represents skill levels in 2024 and so magenta to the left means larger percentages of lower-skilled students in 2024 than 2022. There’s a large patch of purple at the top of the figure, mostly to the left of the green tower representing the Delaware threshold for proficient and entirely to the left of the black tower representing the NAEP Proficient threshold. The middle of the chart is the middle skill level for Delaware, and so, the chart shows a shift from larger percentages of middle level skills (though still lower-skilled than proficient) in 2022 to larger percentages of very low-skilled students in 2024.

District of Columbia

DC, as is often the case, is a tale of two cities. You see purple at the top and magenta on both the left and right slopes. That means that between 2022 and 2024, DC had fewer students in those middle skill levels and more at the lowest skill levels and at the higher skill levels. This shows up as an increase in percent proficient and, essentially, no change in average scores (see the table, which shows an estimated 1-point gain for DC on NAEP).

Maryland and New Jersey

These two states appeared to be doing better on their state tests than on NAEP between 2022 and 2024.

Maryland

There’s nothing in this chart that says to me why Maryland and NAEP would differ. You can see why NAEP would report no change: The 2022 and 2024 figures almost overlap. Also, the differences in the figures happen almost entirely to the left of the black tower representing the NAEP Proficient threshold. There’s a purple patch at the top and a magenta patch at the lower left, a thin layer of purple down the right slopes and a bit of magenta at the upper right. All this suggests an increase in the percentage of the lowest-skilled students, a decrease in mid-skilled students, a decrease in highly skilled students, and an increase in the percentage of students just above mid-level. I read this as a bit of a slide across the distribution of skills but not necessarily in the percent reaching NAEP Proficient. But, I don’t see in the chart why NAEP and the state test would differ in results. As I look at that green tower, there’s nothing to suggest that, say, students were crossing the Maryland threshold and not the NAEP threshold, which would have explained an increase on the state test and not on NAEP. The difference between Maryland and NAEP results could be that Maryland was picking something else up on its test that is not in NAEP (for instance, remember that the state 8th-grade tests measure writing, as well as reading).

New Jersey

New Jersey looks like a substantial decline in skills between 2022 and 2024. Purple on the right and magenta on the left is just what you don’t want to see in these charts, because they indicate smaller percentages of higher-skilled students in 2022 and larger percentages of lower-skilled students in 2024.

You can see, though, why New Jersey’s state test didn’t pick up as large a decline in percent proficient. Although the thresholds for proficient for New Jersey and NAEP (the green and black towers) are close to each other, the NAEP tower is in the middle of a big patch of purple. There were more students in that range in 2022 than in 2024. It could be that there were decreases in the percentage of students reaching that NAEP threshold but not the New Jersey threshold. Again, however, it could also be that the New Jersey ELA test is picking up different skills from NAEP reading.

Pennsylvania

Pennsylvania reported worse results on its state test than NAEP found. Here’s the chart of NAEP results:

The Pennsylvania chart shows mostly a narrowing of the extremes of skills. The first thing that you may notice is the magenta peak. This indicates an increase in the percentage of students at Pennsylvania’s middle skill levels. There’s a little bit of exposed purple on the right slope and a bit more on the left slope, meaning a bit fewer students at the higher skill levels and at the lower skills levels (though see the magenta at the lower left indicating higher percentages of students at the lowest skill levels). Most of this suggests fewer disparities in skill levels but is also consistent with no overall change in percent NAEP Proficient or the average NAEP score. It’s surprising to me, though, that Pennsylvania found a decline in percent proficient when it looks like, at least on NAEP, Pennsylvania had higher percentages of students reaching that green tower in 2024. Again, I’ll hide behind the state-may-measure-something-different-on-its-state-test caveat. It’ll be interesting to see in 2026, though, if Pennsylvania doesn’t show up with some gains in 8th-grade reading.

In Sum

These NAEP charts did not do much to help explain differences between state test results and NAEP results where we found them. However, I think they provided some color on developments in 8th-grade reading between 2022 and 2024 in the Mid-Atlantic states, at least on NAEP. Across the nation, 2024 was a tough year for 8th-grade reading—and 2025 was tough too (see the recent NAEP Long-term Trend report on 13-year-olds). And, that was largely true for the Mid-Atlantic states, though there were some bright spots in DC and maybe Maryland and Pennsylvania.

Over the next three posts, look for similar posts with hopefully some better news for the Mid-Atlanti states in 8th-grade math, and 4th-grade math and reading. Again, you can see all the other states reviewed so far at my Substack page.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Delaware, the District of Columbia, Maryland, New Jersey, and Pennsylvania). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except Maryland (1.2 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

I’m focused on (1) states’ reports of “percent proficient,” which is the percentage of students reaching the threshold of “proficient” (or whatever the state calls proficient) on their state test; (2) where the state’s threshold for proficient lands on the NAEP scale and how the state’s threshold compares with the NAEP Proficient threshold; and (3) where the state threshold and the NAEP Proficient threshold fit in the wider distribution of achievement in the state.

You can look to the appendix for more on methods but, otherwise, let’s get to the results for 4th-grade math. See the table below.

In fourth grade math, there’s a good deal of consistency between what you find in the states’ results and the NAEP results. That was true for the Northeast states in math at grade 8, too, but not for reading at either grade. As usual, Massachusetts and Rhode Island have set high standards—as high or higher than NAEP (see the “Threshold for proficient” columns). At grade 4, Vermont’s new threshold for proficiency may actually be highest of all (but this is my unofficial estimate, so I’d take it as uncertain). Thresholds for New Hampshire and Connecticut are also pretty close to the NAEP Proficient threshold. It’s Maine and New York that are different, with lower thresholds according to my unofficial estimates. But that’s not to say they’ve misplaced them; they may be well placed for efforts to improve skills for large numbers of kids (see discussion below and in the other posts on the Northeast).

In terms of actual student performance, regardless of what the states report for percent proficient, Massachusetts stands out. In 2024, NAEP reported half the Massachusetts fourth graders as reaching NAEP Proficient (see the Percent NAEP Proficient columns).

What do things look like across the whole distribution of student achievement? See the charts below. The 2022 distributions of skills are in purple and the 2024 distributions in magenta. The green towers show about where the states’ thresholds for proficient land on NAEP. The black towers mark the location of the NAEP Proficient threshold.

I won’t say much about the charts but please take a moment to look at them. I’ll just say three things:

1. First, and, in italics: NAEP found gains in grade 4 math almost across the board in the Northeast states between 2022 and 2024—and often without a widening in their distributions. That is, at most skill levels, things looked better in 2024 than 2022: fewer students with lower skills and more students with higher skills.

2. Take a look at Connecticut, Massachusetts, New York, and Rhode Island, where you see big improvements in achievement without much spread in achievement between high and low. In the case of New York, I don’t see any spread in achievement. It’s like they moved a mountain: Higher percentages of the highest-skilled students and lower percentages of the lowest-skilled students. Rhode Island has a bit of an increase in the percentage of very low-skilled students but the highest-skilled Rhode Island 4th-graders in 2024 were more highly skilled than the highest-skilled in 2022. That’s remarkable. There were Rhode Island students scoring higher in 2024 than any Rhode Island kids had scored in 2022.

3. As in math at grade 8 and reading at both grades, Maine has placed its threshold for proficient in a high traffic area, where there are lots of students. If they could help students in that range improve, they could help a lot of students.

Here’s Connecticut:

Maine:

Massachusetts:

New Hampshire:

New York:

Rhode Island:

Vermont:

See more posts on state tests and NAEP

In Sum

This concludes a tour through the recent achievement results in the Northeast states in math and reading at grades 4 and 8, combining state test and NAEP results to try to get a sense of progress since 2022. Across the Northeast, as in the nation more generally, 2024 saw progress in 4th-grade math in the wake of the pandemic. This was especially true in Connecticut, Massachusetts, and, most especially, New York and Rhode Island.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Connecticut, Maine, Massachusetts, New Hampshire, New York, Rhode Island, and Vermont). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except for Connecticut (1.9 points) and Vermont (1.2 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Maine and Vermont had significant changes in their state tests between 2022 and 2024, which means you can’t really assess their growth between 2022 and 2024 using their state tests. So, I’m not using their state test scores for looking at change between 2022 and 2024 and I’m only using their 2024 data to estimate their state threshold on the NAEP scale. Please note that my estimates are my estimates only and not official results from NCES, where they do a much more rigorous job.

The performance distribution figures were also produced from the distribution percentages statistic in the NAEP Data Explorer.

For a more comprehensive but short and readable discussion of issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences and REL is the Regional Educational Laboratory Program. Thanks for reading! Subscribe for free to receive new posts.

In this series of posts, we’re looking at (1) states’ reports of “percent proficient,” which is the percentage of students reaching the threshold of “proficient” (or whatever the state calls it) on their state test; (2) where the state’s threshold for proficient would land on the NAEP scale (given that all the states set their own threshold for proficient) and how the state’s threshold compares with NAEP’s threshold for NAEP Proficient; and (3) where the state threshold and the NAEP Proficient threshold fit in the wider distribution of achievement in the state.

Share

Adding NAEP to the state test results provides a way to compare across states and to get a look at the whole spectrum of student achievement and not just the percentages hitting achievement thresholds. I should note, though, that the contents of states’ ELA tests and NAEP’s reading assessments don’t quite line up, particularly since state ELA tests frequently include writing as well as reading.

You can look to the appendix for more on methods but, otherwise, let’s get to the results for fourth grade reading. See the table below.

What this table shows is that the Northeast states vary widely in what they call proficient English/Language Arts at grade 4 but not so much in the actual reading skills of 4th grade students. The percent proficient on states’ tests ranged in 2024 from 30 percent proficient in Rhode Island to 65 percent in Maine. However, when you use the NAEP Proficient standard as a constant benchmark across states, you see a smaller range. And, in fact, Maine has the lowest percent NAEP Proficient (26 percent), Rhode Island is at 33 percent, and Massachusetts is at the high end with 40 percent NAEP Proficient. If you look at where their state test thresholds for proficient would sit on the NAEP scale (the threshold for proficient column), they range from about 198 in Maine to 242 in Massachusetts and Rhode Island—a very wide range on the NAEP scale. Rhode Island looks as if it has a low percent proficient in part because their standard is so high (higher than the NAEP Proficient threshold of 238 NAEP scale points).

Growth between 2022 and 2024 was not great in any Northeast state. Those that could report change on their state tests (all but Maine, New York, and Vermont, which changed their tests between 2022 and 2024), reported small or no gains. On NAEP, every state was stagnant or actually declined. You see this in both the NAEP Proficient columns and the NAEP average score columns. Remember, the fourth graders in 2024 had been kindergartners in March 2020 when schools shut down for the pandemic and then spent much or all of the 2020-21 school year attempting to learn to read at home.

Let’s look at how the variation in the Northeast states’ proficiency thresholds in ELA at grade 4 play out against the backdrop of the distributions of student achievement in the states, as shown in NAEP. In the charts below, 2022 is represented in purple and 2024 in magenta. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale. The black towers are bands around NAEP’s NAEP Proficient threshold.

Here are all the charts. I’ll discuss them a bit below. Although you might find other elements of interest in the charts, what I’m primarily looking at are the shapes of the reading achievement distributions, how they’ve changed between 2022 and 2024, and where states have placed their thresholds for reading proficiency.

What struck me was how similar these 4th-grade charts looked to the charts for 8th-grade reading for nearly every Northeast state:

• Remember when I said that the 2024 4th-graders had had an especially difficult path to learning to read because the pandemic shutdown occurred in the spring of their kindergarten years? What I meant to say is that the 2024 8th-graders had had an especially difficult path, too—because for all the Northeast states, except Maine and New York, these charts look a good deal like the charts for 8th grade.

  • In Maine, the differences between 4th and 8th grade were relatively subtle. At 4th grade, the main changes between 2022 and 2024 were declines in the percentages of students with middle-high skills (the purple at the top right of the chart) and increases in percentages of students with middle-low skills (the magenta down the left slope). At 8th grade, Maine experienced declines in the percentages of students at relatively high skill levels and increases in the middle skill range. So, the declines at 8th grade occurred higher up the skills distribution.

  • In New York, 4th grade looked considerably better than 8th grade. In fact, New York, which had stood out as maybe having the worst declines among Northeast states between 2022 and 2024 at 8th grade, probably had the best-looking chart at 4th grade. You see in the chart more magenta to the right and purple to the left, meaning larger percentages of students with higher skills in 2024 and larger percentages with lower skills in 2022. And, unlike Connecticut, New York didn’t have any increase in percentages of students at the very lowest skill levels.

• As I said for 8th grade, Maine and Vermont have lower thresholds for proficient than the other Northeast states at grade 4, but it’s not clear to me that they are misplaced. If the states targeted efforts at the region near their proficiency thresholds, then there are lots of students in that lower-middle achievement region whose reading could be improved.

In Sum

In most Northeast states, progress (or lack of progress) between 2022 and 2024 in reading at grade 4 looks a lot like grade 8 reading: a tough period that shows up mainly when you look at their state test results in the context of their NAEP results. The exception is New York, which bucked trends in the Northeast and the nation as a whole: In New York, the reading skills of 4th-graders in 2024 looked as good or maybe better than in 2022.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Connecticut, Maine, Massachusetts, New Hampshire, New York, Rhode Island, and Vermont). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except for Connecticut (1.6 points) and New York (1.1 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Maine, New York, and Vermont had significant changes in their state English/language arts tests between 2022 and 2024, which means you can’t really assess their growth between 2022 and 2024 using their state tests. So, I’m not using their state test scores for looking at change between 2022 and 2024 and I’m only using their 2024 data to estimate their state threshold on the NAEP scale. Please note that my estimates are my estimates only and not official results from NCES, where they do a much more rigorous job.

The performance distribution figures were also produced from the distribution percentages statistic in the NAEP Data Explorer.

For a more comprehensive but short and readable discussion of issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences and REL is the Regional Educational Laboratory Program.

In this series of posts, we’re looking at (1) states’ reports of “percent proficient,” which is the percentage of students reaching the threshold of “proficient” (or whatever the state calls it) on their state test; (2) where the state’s threshold for proficient would land on the NAEP scale (given that all the states set their own threshold for proficient) and how the state’s threshold compares with NAEP’s threshold for NAEP Proficient (set as a goal for the nation as a whole); and (3) where the state threshold and the NAEP Proficient threshold fit in the wider distribution of achievement in the state.

What the NAEP perspective provides is a way to compare progress across states and to get a smooth look at the whole spectrum of student achievement and not just the percentages hitting achievement thresholds (see also these excellent charts at the Nation’s Report Card website). Seen together, the state and NAEP results can point to potential reasons state tests and NAEP can sometimes appear to be saying different things about progress in a state. It matters, for instance, where states set thresholds for proficiency compared with NAEP Proficient. In the case of states’ ELA and NAEP’s reading assessments, it’s important also to note that the content doesn’t quite line up, particularly since state ELA tests frequently include writing as well as reading.

Okay, what are the recent results on Northeast state ELA tests and NAEP reading assessments at grade 8? See the table below. State results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Connecticut 2025, Maine 2025, Massachusetts 2025, New Hampshire 2025, New York 2025, and Rhode Island 2025). NAEP results are drawn from the NAEP Data Explorer.

Two things stood out to me from this table: First, with the exception of Massachusetts and Rhode Island, states reported higher percentages of their students as proficient than NAEP reported as NAEP Proficient. What this means is that the thresholds the Northeast states are setting for proficient are below NAEP’s NAEP Proficient threshold—except in Massachusetts and Rhode Island, where the thresholds are comparable. This doesn’t mean the states are wrong in where they’re setting their thresholds. The italics for NAEP Proficient are intentional because the body that sets the NAEP Proficient thresholds, the National Assessment Governing Board (NAGB), has historically set NAEP thresholds high to fit the whole nation and to allow for progress over time.

Second, and perhaps more important, states’ reports on percent proficient would suggest progress between 2022 and 2024 but when you bring in NAEP it looks like that progress is probably limited. NAEP and the states were pretty similar in what they found in terms of changes in percent proficient but NAEP found virtually no change in average scores (with the exceptions of New York and Vermont, where NAEP showed pretty substantial declines). This was true in grade 8 math for the Northeast states too (see that post.)

You can see this better when you have charts of the whole distribution of achievement. In purple is the distribution of student achievement on NAEP for 2022 and in magenta is the distribution for 2024. The charts show the percentage of students scoring in each 10-point band of NAEP scores from 0 to 500. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale. The black towers are bands around NAEP’s NAEP Proficient threshold.

The first set of charts is for Connecticut. I’ll spend a bit more time on Connecticut than the other states to help orient us to the figures.

Three things I noticed for Connecticut: First, the towers for 8th-grade reading are right in the middle of the mountains for 2022 and 2024. The middle is where the highest percentages of students are. The top of the 2022 figure exceeds 10 percent of students and the top for 2024 is just under 10 percent of students. If you could move achievement at this middle level of achievement in Connecticut, you might see a relatively large shift in percent proficient on both the state test and NAEP. Second, though, is that you don’t see that shift between 2022 and 2024. Neither the state nor NAEP found almost any change in percent proficient between 2022 and 2024. And you can see that lack of change in the chart: The purple and magenta shapes are almost perfectly centered. The exposed purple at the top of the chart shows there were higher percentages of students in that middle range in 2022. But the symmetry suggests that by 2024 that middle area had spread evenly between lower and higher achievement. The mountain is spreading in Connecticut—more higher and more lower achieving students—but it’s happening evenly on either side of the proficiency thresholds for the state and NAEP. Third, if you look at the grade 8 math post, you’ll see the reading shapes are much taller and narrower than compared with math. This suggests there may be greater disparities in achievement in math than in reading at grade 8 in Connecticut.

How about Maine?

Two quick things on Maine: First, that long exposed magenta slope on the left (the lower achievement side) suggests a decline in achievement across a broad stretch of the achievement distribution. Second, notice the placement of the green tower on the left side of the mountain, near the top. If Maine focused attention on these students (lower-middle achieving students), they could achieve a substantial gain in percentage of students reaching Maine proficient. This may not show up on NAEP (the NAEP tower is to the right of the Maine tower) but it would still represent substantial growth in Maine.

Here’s Massachusetts:

First, as in Connecticut, the reading achievement distribution looks tighter than the math achievement distribution (the 8th-grade math post). Second, also as in Connecticut (and the nation as a whole), the distribution got a bit wider between 2022 and 2024—the exposed magenta on the lower slopes suggest higher percentages of both lower and higher achieving students.

New Hampshire?

First, the chart shows very little change between 2022 and 2024 in New Hampshire, across the distribution. However, second, you can see a possible reason New Hampshire showed a bit more progress (a gain of 3 points in percent proficient) than NAEP (1 point in NAEP Proficient): The New Hampshire proficient threshold is a bit to the left of the NAEP threshold and most of that exposed magenta in the chart is to the right of the New Hampshire threshold but not necessarily the NAEP threshold.

New York:

This is not what you want to see in 8th-grade reading skills in New York. That is a whole lot of exposed magenta at the very lowest skill levels in New York (the bottom left), lots of purple at the top (average skill levels), and only a little magenta at the bottom right. What this indicates is tremendous growth in the percentages of the lowest-skilled readers, a reduction in the percentage of mid-skilled readers, and a very small increase in the percentage of the very top readers. These changes don’t show up in changes in NAEP Proficient because that change (the purple and magenta) is occurring below the threshold for NAEP Proficient.

Here’s Rhode Island:

It’s difficult to see from the chart why Rhode Island is reporting a 3-point gain in percent proficient and NAEP estimated a 1-point drop in percent NAEP Proficient. The mountains for 2022 and 2024 (based on NAEP results) overlap very closely, suggesting only a slight decline in overall achievement. And, the state and NAEP thresholds for proficient are very close together. The small differences between the Rhode Island state and NAEP results may stem from differences in the assessments (Rhode Island’s ELA test includes writing and NAEP’s reading assessment does not).

How about Vermont?

Reading in grade 8 in Vermont looks something like Maine. They both show declines in achievement between 2022 and 2024. Here, with Vermont, you can see this in the exposed magenta along the left slope of the chart and exposed purple on the right. As with Maine, the new state test threshold is set lower than the NAEP Proficient threshold (the green tower is to the left of the black tower). But, as with Maine and New York, the Vermont threshold is placed in an area where there are lots of lower-middle achieving students; if Vermont focused efforts on this area then they could see substantial increases in percent proficient (and the increases may not be picked up by the NAEP Proficient metric).

In Sum

Reading achievement at grade 8 in the Northeast states is a more varied story than the story for math. Maine, Vermont, and, particularly New York stand out as suffering substantial recent setbacks in reading achievement. All three states have implemented new tests with new thresholds for proficient in recent years. These new thresholds for proficient are lower than those for the other Northeast states and lower than the NAEP Proficient threshold but they seem to be aimed squarely at a large number of lower-middle achieving students.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Connecticut, Maine, Massachusetts, New Hampshire, New York, Rhode Island, and Vermont). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except for Connecticut (2.1 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. Maine, New York, and Vermont had significant changes in their state English/language arts tests between 2022 and 2024, which means you can’t really assess their growth between 2022 and 2024 using their state tests. So, I’m not using their state test scores for looking at change between 2022 and 2024 and I’m only using their 2024 data to estimate their state threshold on the NAEP scale. Please note that my estimates are my estimates only and not official results from NCES, where they do a much more rigorous job.

The performance distribution figures were also produced from the distribution percentages statistic in the NAEP Data Explorer.

For a more comprehensive but short and readable discussion of issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences and REL is the Regional Educational Laboratory Program.

As with other posts in this series, I’m looking at (1) states’ reports of “percent proficient,” which is the percentage of students reaching the threshold of “proficient” (or whatever the state calls that level of achievement) on their state test; (2) where the state’s threshold for proficient would land on the NAEP scale, given that all the states set their own threshold for proficient; (3) how the state’s threshold for proficient compares with NAEP’s threshold for NAEP Proficient; and (4) where the state threshold and the NAEP Proficient threshold fit in the wider distribution of achievement in the state.

So, what are the recent results on state tests and NAEP in grade 8 math for the Northeast states? See the table below. State results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Connecticut 2025, Maine 2025, Massachusetts 2025, New Hampshire 2025, New York 2025, and Rhode Island 2025). NAEP results are drawn from the NAEP Data Explorer.

The first thing I notice in the table is that all the states except Rhode Island and New York reported similar percentages of students as proficient (see the first two columns—I’d disregard the Maine 2022 results when they administered an interim test and reported a shockingly high percent proficient). In 2024, all but Rhode Island and New York reported between 34 and 39 percent proficient; Rhode Island reported just 25 percent proficient and New York 57 percent. Is Rhode Island not doing as well as the rest of New England? Actually, it may be doing pretty well. First, they had a gain from 2022 (which is good) and, second, when you look at their NAEP results (columns 3 and 4), they don’t look so different from the other states. In 2024, NAEP reported 26 percent NAEP Proficient in Rhode Island, which is similar to results for Maine and Vermont and the same as New York. Rhode Island just has high standards. This is reflected in the columns showing the thresholds for proficient in the states and NAEP. My estimate for Rhode Island (302 points on the NAEP scale) is actually higher than NAEP’s NAEP Proficient threshold.

The second thing I notice is that NAEP’s reported two-year gains in percent NAEP Proficient are pretty similar to the states’ reports on their own tests. The states reported gains of between 2 and 5 points. NAEP mostly reported 2-point gains.

But this is just percent proficient (or percent NAEP Proficient). Let’s look for change across the whole distribution of students. Each state gets a chart below. In purple is the distribution of student achievement on NAEP for 2022 and in magenta is the distribution for 2024. The charts show the percentage of students scoring in each 10-point band of NAEP scores from 0 to 500. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale. There’s no green band for New York because it’s difficult to line up their NAEP population with their general assessment population, since so many of their students take an algebra assessment rather than the general math assessment for their end-of-year assessment. The black towers are bands around the NAEP Proficient threshold.

If you have time, you should study the charts closely. Some things I saw:

1. Almost uniformly, states’ distributions are slouching. The exception is Maine, whose distribution looks almost unchanged between 2022 and 2024. The other states are purple-topped mountains with magenta patches down the slopes. This indicates smaller percentages of students achieving in that middle range and larger percentages of students achieving relatively poorly (lower left slope) and relatively highly (lower right slope). That is, there were, both, more lower achieving kids and more higher achieving kids in 2024 than in 2022. In the case of New York, you only see that magenta patch on the left slope, meaning they just had larger percentages of lower-skilled students (not higher-skilled students).

2. The bands for the proficiency thresholds tend to be near the tops of the mountains, where there are relatively large percentages of students (the higher the point on the mountain, the higher the percentage of students in that score band). That means relatively small differences—either over time or between the NAEP and state thresholds—could result in relatively large differences in percent proficient. This could help explain the slight differences between the gains in percent proficient reported by the states and NAEP.

3. For all the changes across the distribution in a state (all the exposed purple or magenta showing higher percentages of students in a score range in 2022 or 2024), the only changes captured by the percent proficient metric are those crossing the proficient threshold. So, you can report increases in percent proficient while still having increases in the percentages of students at very low levels of achievement. And, you can have increases in percent proficient at the same time that you’re having no change in average scores (as some Northeast states saw on NAEP between 2022 and 2024).

4. New York’s NAEP chart is just what you don’t want to see: purple on the right and magenta on the left, indicating larger percentages of higher-skilled students in 2022 than 2024 and larger percentages of lower-skilled students in 2024 than 2022.

   

   

In Sum

Looking across the Northeast states, you see a pretty consistent story. Almost uniformly, states seem to be building up the percent proficient in grade 8 math but possibly not improving at the overall level. And this is because, as they build up higher achieving students, the states seem also to be accumulating lower achieving students.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Connecticut, Maine, Massachusetts, New Hampshire, New York, Rhode Island, and Vermont). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. For Maine, because the state test changed between 2022 and 2024, I used only my estimate for 2024. NCES did not estimate a proficiency threshold for New York in 2022 because the combination of general test and algebra test results in New York didn’t support the analysis needed to do such an estimation; so I didn’t make an estimate either. Please note that my estimates are my estimates only and not official results from NCES, where they do a much more rigorous job. Vermont also changed its test between 2022 and 2024. However, their threshold changed very little on the NAEP scale and, therefore, I included the comparison between 2022 and 2024 for Vermont. New York changed its threshold for proficiency between 2022 and 2024.

The performance distribution figures were also produced from the distribution percentages statistic in the NAEP Data Explorer.

For a more comprehensive but short and readable discussion of issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences and REL is the Regional Educational Laboratory Program.

Updated 6/10/2026 to remove 2022 to 2024 state test comparisons for New York, because I learned New York had changed its threshold for proficiency between 2022 and 2024.

Share

This is also part 13 of a broader series across the nation. I’m comparing results of states on their own tests with their results on NAEP. First, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests and the percentage NAEP Proficient. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8.

Second are charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of the recent 4th-grade reading results for these states. Note that we can’t calculate two-year gains on the state tests for three of the states because they changed their state tests between 2022 and 2024. These are Arkansas, Oklahoma, and Texas. We have to rely entirely on NAEP to measure change over time for these states.

The first thing I note in the table is, as is often the case, the stark differences in percent proficient between the states and NAEP. In 2024, the Southwest states ranged in percent proficient according to their own tests from 49 percent in Texas to 36 percent in Arkansas. On NAEP, the states ranged from 32 percent NAEP Proficient in Louisiana to 20 percent in New Mexico. So, it’s not even the same states that are identified as highest and lowest achieving between the state tests and NAEP. That’s because the states each set their own thresholds for proficiency, while NAEP has one standard for NAEP Proficient across all the states. As a result, Texas may report nearly half its students proficient 4th-grade readers, while NAEP reports closer to a quarter NAEP Proficient. On the other hand, Arkansas’s threshold is so rigorous relative to the other states and many of their students such strong readers that NAEP finds Arkansas among the leaders in the Southwest, even as Arkansas reports itself as last in the region in terms of 4th-grade reading proficiency.

Second, for the states where we can compare the results of the state tests between 2022 and 2024, we see pretty big differences in gains between NAEP and the state tests. But in different ways: Louisiana’s state test showed a 3-point decline in percent proficient, while NAEP estimated a 4-point gain in NAEP Proficient; however, New Mexico reported a 6-point gain in percent proficient and NAEP estimated a slight decline in percent NAEP Proficient. We saw something similar at grade 8 for New Mexico in reading. I’m not sure why that’s happening. It’s important to point out, though, that NAEP and state tests are different assessments, and this is particularly so in the case of reading/ELA. We know, for instance, that the states’ English/Language Arts (ELA) assessments test writing but NAEP’s reading assessment does not.

Sometimes it helps to look at the charts that show the whole distribution of achievement on NAEP. So, let’s go to the charts (for information on how to read the charts, please see the appendix).

Arkansas, Oklahoma, and Texas

I want to show, first, charts for the states that changed their state tests between 2022 and 2024: Arkansas, Oklahoma, and Texas. I’m going to discuss Arkansas and Oklahoma together, because their charts look similar. Here’s Arkansas:

Here’s Oklahoma:

These charts look like slight declines in percent NAEP Proficient and in average score (which is what we see in the table, too). You can see purple to the right of the NAEP threshold for proficiency (the black tower) and magenta to the left of the threshold. Purple stands for 2022 and magenta for 2024. And so, this pattern suggests declines: larger percentages of students with higher skills in 2022 (purple), larger percentages of students with lower skills in 2024 (magenta). The pattern is more pronounced for Arkansas, where we see a good deal of magenta well down the left slope (toward lower skill levels).

In Oklahoma, the shift appears to be small, from the upper right side of the mountain to the upper left side. That is, in both years there were large percentages of students near those proficiency thresholds for Oklahoma; it’s just that there were more to the right in 2022 and to the left in 2024. As in 8th-grade reading, the placement of Oklahoma’s threshold (the green tower) is right where there are lots of students and right where there seems to be a good deal of movement over time. You could imagine focused effort at that skill level leading to increases in percent proficient in Oklahoma.

Texas

Texas, like Oklahoma, has a good deal of movement near its proficiency threshold (the green tower) and NAEP’s threshold (the black tower). The magenta peak indicates an increase in the percentages of 4th-graders in the middle skill levels. Unfortunately, there’s also magenta at the bottom left slope, indicating growth in the percentages of students with very low reading skills. As in the post on 8th-grade reading, I checked to see if the increase in the percentage of lower-skilled English readers might be caused by an increase in the population of English Learner students. However, the population of English Learners ticked down a bit between 2022 and 2024 (though declines in skills appeared to be especially large for English Learner students).

Louisiana

Louisiana’s chart for 4th-grade reading is just what you want to see: purple on the left and magenta on the right. It’s just as I said for math at both grades. The chart shows larger percentages of students with lower skills in 2022 (purple on the left) and larger percentages of students with higher skills in 2024 (magenta on the right). As with the other grades and subjects, I don’t know why NAEP is estimating gains and Louisiana is not. Part of the difference here could be the presence of writing items in the Louisiana ELA assessment and the lack of them in NAEP’s reading assessment. However, I don’t know why that would cause a difference in gains between two years given that the assessments haven’t changed: NAEP never had writing items in its reading assessment and Louisiana presumably did not change its assessment between 2022 and 2024.

New Mexico

The difference between NAEP results and New Mexico’s state test results is also not explained by the chart. Similar to 8th-grade reading, New Mexico reported a large gain in percent proficient and NAEP did not. The chart of NAEP results shows almost a perfect overlap in the distribution of skills in 2022 and 2024 in New Mexico, suggesting no change rather than the big gain reported by New Mexico. Also, as with 8th-grade reading, I calculated a large drop between 2022 and 2024 in New Mexico’s threshold for proficiency on the NAEP scale (the numbers behind the green tower). This suggests there may have been some kind of difference between NAEP and the New Mexico test in 2024 that was not there in 2022.

In Sum

In the introduction to the post, I focused on Louisiana and how much better its 2024 results looked on NAEP than on the Louisiana state test. The other key finding of the post was that the other Southwest states did not fare so well, including New Mexico, which appears on its state test to have made considerable progress between 2022 and 2024. I want to close with a reminder of the very difficult circumstances facing the 4th-graders of 2024 over the years they were expected to develop foundational literacy skills. These are the students who were in the spring of their kindergarten year when the pandemic hit and students were sent home from school. Many of them then spent most or all of 1st grade trying to learn remotely. Those who were in physical schools in first grade faced all the stress and disruptions that went along with in-school learning in the 2021 school year. No one had a favorable environment for learning to read.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Arkansas, Louisiana, New Mexico, Oklahoma, and Texas). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except Louisiana (1.8 points) and Oklahoma (1.4 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. For Arkansas, Oklahoma, and Texas, because their state tests changed between 2022 and 2024, I used only my estimates for 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 12 in a series that reviews recent state test results at grades 4 and 8 in math and reading in the context of National Assessment (NAEP) results. Region-by-region, I compare states’ results on their own tests with their results on NAEP. First, I show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests—because percent proficient tends to be the focus of state tests results (“proficient” or some other similar term)—as well as the percentage NAEP Proficient and states’ average NAEP scores. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8.

Second, I present charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

In this post, we’re looking at grade 4 mathematics for the five states in the Southwest region of the IES REL program. In addition to last week’s post on 8th-grade math, there’s also a post on 8th-grade reading.

Here’s a table of recent 4th-grade math results for the IES REL Southwest states:

Usually, my first comment on this table is look at how much wider (and sometimes higher) the range is for the state results on percent proficient than for the percent NAEP Proficient. In this case, however, the two measures line up pretty well. On both the state test and NAEP, Texas appears to have the highest percent proficient (44 percent on the state test and 43 percent on NAEP in 2024) and New Mexico the lowest (25 percent and 23 percent, on the state test and NAEP, respectively). The consistency between NAEP and the state tests is because the thresholds for proficient for the state tests and NAEP are pretty close to each other across the states. In the cases where we see bigger differences ((Arkansas and Oklahoma), it’s probably because those states have somewhat lower thresholds, and so, they are reporting higher percent proficient than NAEP does.

Second, look at those gains in percent NAEP Proficient! Nationally, NAEP reported a 3-point gain in percent NAEP Proficient between 2022 and 2024. All the Southwest states show gains at least as big (although only Louisiana and New Mexico had sufficiently clear gains that they passed NCES’s statistical tests of significance). Also, it looks like Louisiana and New Mexico both made more progress as measured by NAEP than their state tests. The other three states had changes to their state tests between 2022 and 2024 such that you can’t compare their growth on their state tests.

But we can learn more if we look at the charts that show the whole distributions of kids’ scores on NAEP in each state in 2022 and 2024 (for information on how to read the charts, please see the appendix). So, here goes.

Arkansas, Oklahoma, and Texas

Let’s look first at the states that changed their state tests between 2022 and 2024. Here’s Arkansas:

What you want to see on these charts is purple to the left and magenta to the right. That’s what you see in Arkansas in 4th-grade math. What it means is higher percentages of students at lower scores in 2022 (purple at the left) and higher percentages with higher scores in 2024 (magenta on the right). You can see a little bit of magenta at the farthest left, indicating higher percentages of students in 2024 at the very lowest scores but, otherwise, this is purple on the left, magenta on the right. The largest gains appear to be in the percentages of students with the very highest scores in Arkansas.

Oklahoma

Oklahoma showed progress that was, in a way, very similar to Arkansas: purple on the left, magenta on the right, indicating gains broadly across the distribution of students from lower to higher achieving; but with a little magenta on the left, meaning a bit of an increase in the percentage of very low achieving students. What I want to point out as distinctive about Oklahoma is how tall and narrow their figures are. In each year, that peak in the middle of the figure is about 13 percent. This means high percentages of students in that middle-achieving range in Oklahoma and less disparity in math skills between higher and lower-skilled kids than you would see in many other states.

Texas

You can see in this chart why Texas might see a large gain in percent proficient (estimated 5 percentage points): There’s a solid shift to the right at the top of the chart in the range of the NAEP Proficient threshold (the black tower overlapping with the green tower showing my estimate for Texas’s own state test proficiency threshold). That purple peak on the left and magenta peak on the right shows a substantial shift from high percentages of students with skills just below the proficiency threshold in 2022 to above it in 2022. A second feature to note in this chart, though, is the presence of patches of magenta at the bottoms of the chart, on both the left and right sides. These patches suggest, as we’ve seen in the other grades and subjects in Texas, a widening disparity between lower and higher-skilled students. That is, an increase in, both, the percentages of the lowest-skilled students and highest-skilled students.

Louisiana

This chart for Louisiana is what progress looks like. This is the purple-on-the-left, magenta-on-the-right that I’ve mentioned again and again as what you want to see in these charts. This is what an 8-percentage-point gain in percent NAEP Proficient and a 6-point gain in average NAEP score looks like. I don’t have a good explanation for why Louisiana showed a smaller gain in percent proficient on its own state test. A 2-percentage-point gain, as the state test shows, is no small feat but an estimated 8-point gain, as NAEP finds, is really something.

New Mexico

Normally, it’s concerning to see a state’s distribution grow shorter over time. By that I mean a taller purple peak than magenta peak. The shorter peak is usually accompanied by a widening distribution, meaning greater disparities between higher and lower-skilled students and, often, an accumulation of students with very low skills. But that’s not what we see in New Mexico in 4th-grade math between 2022 and 2024. There is a small amount of growth in the percentages of very low-skilled students (the magenta strip along the lower left slope) but there’s a good deal of growth in percentages of students scoring from the middle skill levels to the very highest skill levels. That is, there is some widening in the distribution but it seems almost entirely to be the result of a growth in higher-skilled students.

In Sum

Without exception, the REL Southwest states did as well as the nation or better in rebounding from the low 4th-grade math scores we saw in 2022 in the immediate wake of the pandemic. Louisiana, however, really stood out, perhaps in part because of how much more progress NAEP is showing than its own state test. We saw that in math at 8th grade and reading at 8th grade, as well. In the next post, we’ll see it again in 4th-grade reading. The rest of the Southwest region did not get off so well in 4th-grade reading in 2024.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Arkansas, Louisiana, New Mexico, Oklahoma, and Texas). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except New Mexico (1.2 points). I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. For Arkansas, Oklahoma, and Texas, because their state tests changed between 2022 and 2024, I used only my estimates for 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job. For more reading on issues related to tracking changes in percent proficient, see this piece based on the work of Andrew Ho and others.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

Share

This is part 11 of a series reviewing recent state test results in the context of National Assessment results for math and reading at grades 4 and 8. It’s the second part discussing results for the IES REL Southwest region states of Arkansas, Louisiana, New Mexico, Oklahoma, and Texas (see the appendix for spelled out acronyms for IES, REL, and others). Last week, I posted a summary for 8th-grade reading in the Southwest states. In each post, I first show a table that summarizes results in 2022 and 2024 on the percentage of students “proficient” on state tests based on each state’s definition of proficient (or whatever term the state uses), as well as the percentage NAEP Proficient and states’ average NAEP scores. I use the years 2022 and 2024, because they are the most recent years with NAEP results in math and reading at grades 4 and 8. Next, I run through charts for each state that show changes in the whole distribution of students, from the lowest scorers to the highest, using NAEP data. This allows you to see where changes may be occurring that don’t show up in changes in the percent proficient or even in average scores (see also these excellent charts at the Nation’s Report Card website).

Here’s the table of recent 8th-grade math results for the IES REL Southwest states:

The first thing I always note is the difference between what the states report on percent proficient and what NAEP reports on percent NAEP Proficient. For instance, in 2024, the REL Southwest state tests ranged from 40 percent proficient in Texas to 19 percent proficient in New Mexico. NAEP shows the same states at the top and bottom but with a much smaller range: 24 percent NAEP Proficient in Texas and 14 percent in New Mexico. This is because NAEP applies the same threshold for proficiency to both states and New Mexico’s threshold for proficient is a bit higher than the Texas standard.

The second thing I note is that this table—focused on percent proficient and on average scores—shows almost no change. This is partly because Arkansas, Oklahoma, and Texas changed their state tests and so they can’t be compared between 2022 and 2024. However, if you look at just the NAEP numbers, you see don’t see much change either. The official NCES release of NAEP results reported “no change” across the board for the Southwest states because the results did not pass statistical tests of change (though, if you look at the far right columns, those decreases in average scores for Texas and New Mexico were probably pretty close to passing the statistical test and likely represent real, substantive declines).

The charts show change. When you look beyond the percentages reaching proficiency and beyond the average scores, it looks like real changes in these states between 2022 and 2024. So, let’s go to the charts (for information on how to read the charts, please see the appendix).

Arkansas, Oklahoma, and Texas

I grouped together first the states that changed their state tests between 2022 and 2024. Here’s Arkansas:

The Arkansas story in 8th-grade math appears to be different from last week’s story in reading. In 8th-grade reading, it looked like Arkansas had made progress in moving more students closer to its proficiency threshold. So, although the official reports didn’t show change, it looked real improvement for many students and boded well for the future in Arkansas in reading. The story is more mixed in 8th-grade math. Look at the magenta along the lower slopes on the left and right and the purple at the top. The magenta represents 2024, and so, the magenta on the slopes indicates higher percentages of students at both the lowest skill levels and highest skills levels in 2024. So, beneath the lack of change in percent proficient, there was actually divergence in outcomes for higher and lower achieving students—more students at the most advanced levels but also more students with very low skill levels.

Oklahoma

Oklahoma’s pattern is similar to Arkansas’s but less distinct and maybe a bit more positive? There were relatively small gains in the percentage of students with the lowest skill levels (magenta at the bottom left) and a bit greater gains in the percentage of students at the highest skill levels (magenta at the lower right). Again, magenta is 2024.

Texas

The Texas picture for 8th-grade math is not great. You can see why NAEP estimated a 4-point decline in Texas’s average score. Look at all the magenta to the left and purple to the right. It’s the opposite of what you want to see. What this shows is higher percentages of students all along the lower skill levels in 2024 (the magenta on the left slope); in 2022 there had been higher percentages of students at the middle skill levels (the purple at the top and down the right slope). Most of this change occurred below the NAEP Proficient threshold (the black tower) and so it’s not surprising that it did not show up as change in the reported NAEP Proficient. And Texas changed its state test, so we don’t know what the change in percent proficient would have been on the state test (the chart suggests a decline). You might also note that there’s no green tower for Texas’s state threshold. NCES did not estimate Texas’s threshold in 2022 because Texas has a relatively large portion of their 8th-graders taking an algebra test rather than the general 8th-grade test, which makes it difficult to align the Texas NAEP population with its general math test population. The same is true for the next state, Louisiana.

Louisiana

Now Louisiana, the best-looking chart in the Southwest. As in last week’s summary of 8th-grade reading, Louisiana looks better on NAEP than on its state test. This chart shows that “magenta-on-the-right, purple-on-the-left” pattern that you want to see, indicating larger percentages of higher-skilled students in 2024. In fact, Louisiana is the only REL Southwest state with no magenta at the lower level. That is, no growth in the percentage of the lowest-skilled students. (You may also note there’s no green tower showing the estimated state threshold for proficiency on the NAEP scale; see why above in the Texas paragraph.)

New Mexico

NAEP had reported an almost-gain in percent NAEP Proficient (estimated 1-point) at the same time as an almost-decline in average score (estimated 3 points) for New Mexico in 8th-grade math. You can see in this chart why that might be. In a way, it looks like the Arkansas chart, with purple in the middle and magenta down either slope. That shows increasing divergence between higher and lower-skilled students: an increase in the percentage of students with the highest skills (magenta on the lower right slope) and an increase in the percentage of students with lower skill levels (magenta on the left slope). Here, though, there is markedly more growth on the left (lower skills) than the right.

In Sum

This is a case in which you really need to see the NAEP charts across the whole distribution of skills in the states in order to get a feel for what happened between 2022 and 2024. The reported results of percent proficient and average scores suggested virtually no change but the charts showed, in most cases, the kind of divergence we’ve been seeing across the nation for years (growth for higher achieving students, as the number of lower achieving students also grows). The exception here was Louisiana, which showed growth at the high end without an increase in the percentage of lower-skilled students. Texas, unfortunately, showed the opposite pattern: growth in the percentage of students with lower skill levels without growth at the high end. Over the next two posts, I’ll look at 4th-grade.

About the Analysis

State test results are drawn from the Zelma tool at the Education Data Center, which also has great state summaries on Substack (see Arkansas, Louisiana, New Mexico, Oklahoma, and Texas). NAEP results come from the NAEP Data Explorer. Estimates of where states’ proficiency thresholds would land on the NAEP scale for 2022 can be found at the NAEP NCES Mapping State Standards site. See also the recently released NCES report Mapping State Proficiency Standards Onto the NAEP Scales.

For 2024, I had to make my own estimates of states’ thresholds for proficient on the NAEP scale. I used the “distribution percentages” statistic in the NAEP Data Explorer. This statistic produces the percentage of students in each 10-point band of NAEP scores. What I did was say, okay, if a state says 34 percent of its students are proficient, then where’s the point on the NAEP distribution that 34 percent of the state’s students reach? That’s the point where the state threshold would land on the NAEP scale. I couldn’t find the exact point on the NAEP scale because the distribution percentages statistic uses 10-point bands. So, I estimated the threshold within a 10-point band by assuming that students within the band were evenly distributed across the band. This is not a great assumption but probably does not lead to big mis-estimations. I checked my results for 2022 against NCES’s published results and my results were within 1 NAEP scale point in each case, except Oklahoma (1.5 points) and Louisiana and Texas, which do not have NCES estimates because their NAEP populations and state test populations do not align for these purposes because they use algebra tests as state grade 8 math tests for more than 5 percent of their students. I then averaged my 2022 and 2024 threshold estimates into a single pooled threshold estimate for 2022 and 2024. For Arkansas, Oklahoma, and Texas, because their state tests changed between 2022 and 2024, I used only my estimates for 2024. Please note that these are my estimates only and not official statistics from NCES, where they do a much more rigorous job.

How to read the chart: The chart shows the percentage of students scoring in each 10-point band of NAEP scores. The figures are shaped like mountains because higher percentages of students tend to score near the average (the middle) and lower percentages score either very low (to the left) or very high (to the right). The black tower is a band around NAEP’s NAEP Proficient threshold. The green towers are bands around the point where I’ve estimated the state’s threshold for proficient to be on the NAEP scale.

NAEP is the National Assessment of Educational Progress. NCES is the National Center for Education Statistics. IES is the Institute of Education Sciences. REL is the IES Regional Educational Laboratory Program.

See this short, readable discussion of issues associated with tracking change using percent proficient.

Updated 5/25/2026 to correct appendix comparison of NCES estimates of state thresholds and my estimates, as well as to add discussion of ramifications of Louisiana and Texas having more than 5 percent of 8th-graders taking an algebra test rather than the general state test as their end of year test in 8th grade.