Skip to content

Fourth graders lost about half a grade level of learning during the pandemic

Earlier this year the Department of Education conducted a "special administration" of the long-term NAEP test for 9-year-olds. Why the long-term test instead of the main NAEP test? Because by chance they had conducted a long-term assessment in 2020, which meant that a 2022 retest would tell us a lot about how the COVID-19 pandemic affected elementary school students.¹

So how did our fourth-graders do? In a word, badly. Here are the topline results:

The usual rule of thumb is that ten points on the NAEP equals one grade level. This means that in reading, kids in fourth grade this year were about half a grade level behind the fourth graders of 2020. In math they were about three-quarters of a grade level behind.

But it gets worse. The best students lost only 2-3 points in both reading and math while the worst students lost 10-12 points. Low-income students (-6 in reading, -8 in math) lost more than non-poor students (-3 in reading, -5 in math). Students in the West (-1 in reading, -5 in math) fared considerably better than students in every other region (-7 in reading, -8 in math). Black, white, and Hispanic kids all lost about six points in reading. In math, everyone lost a lot but Black kids lost a catastrophic 13 points. Oddly city kids didn't lose anything in reading. Aside from that, kids in suburbs did a little worse than kids in cities, towns, and rural areas.

That's the basics. I'm sure there will be plenty of analysis over the next few weeks, but for now the basic results are all we have. I also suspect there's going to be a very great deal of shrieking and moaning about how our children are all doomed to a future of failure, but please don't join in. These things tend to wash out in later school years, and I very much doubt that by the time our fourth graders graduate from high school they'll be noticeably different from any other graduating class.

¹The long-term assessment is a special version of the NAEP that's been designed to stay the same from year to year, thus making it highly useful for comparisons over time. However, over the course of two years it's no different than the main NAEP test. It's just chance that it happened to be the only test given in 2020.

The main NAEP test was given to 9-year-olds in 2019. It seems to me that 2019-22 would have been a perfectly useful comparison period, but I guess 2020-22 gives a slightly better focus on precisely the pandemic years.

27 thoughts on “Fourth graders lost about half a grade level of learning during the pandemic

  1. jdubs

    As a parent of school age kids who have taken many state standardized tests....it's not clear to me that the incredible amount of time the students spend preparing for tests is very beneficial for anything other than doing well on the metric.
    Covid definitely took time away from test prep. I'm not sure anyone is actually worse off because of that.

    1. cephalopod

      What is all this test prep that is being done? My kids spend almost no time on test prep (most of what my kids report is their teachers encouraging them to get enough sleep and make sure to eat a healthy breakfast). They are in an urban school that tends to outperform the district on tests, so there isn't much at stake for them, I guess.

      1. jdubs

        With only direct knowledge of schools in Houston, Austin, Dallas and Denver, kids spend quite a fair amount of time on practice tests, lessons on test taking strategies and studying material from past years tests.

        Certainly reminds me of executives gaming their bonus metrics instead of actually running the business well.

      2. Atticus

        I agree. I have school age kids and my wife id a teacher. They don't spend any time on test prep. (Other than learning/reviewing the actual material.)

    1. MontyTheClipArtMongoose

      I was going to guess one of the Demokkkrat racists, like Andrew Jackson or Woodrow Wilson, or that miscegenating cretin Thomas Jefferson.

    2. Ken Rhodes

      Nah! Nothing like Trump, who never went to war with Canada or Mexico over imagined abuse of white Americans.

      OTOH, it's practically a verbatim translation of Putin. ... who, BTW, resembles der Fuehrer in quite a few other ways, as well. Now THAT'S something we need to be aware of!

        1. MontyTheClipArtMongoose

          I wonder if he used to play Pretend with Don, Jr., & that's why El Jefecito posted the "my dad has a big donger" image on Twitter.

    3. Spadesofgrey

      Please, nope. Trump is a jew boy. It's Zionism with Rothschildian control, as is Dugin who is a jew. Come on libtards put it together. Kill him. Do a dna test. Understand your destiny.

      Why do you think they fell for Ana de Rothschild scam????

      Hitler???? Please. NOpe.

  2. Joseph Harbin

    What the result seems to show is that shifting from on-campus to remote schooling for a year has a cost. That should surprise nobody. Everything has a cost. Going on summer vacation has a cost. Fixing my broken tooth this week has a cost. Protecting the health of kids and the public from a deadly virus during a time with no vaccine and imperfect information has a cost.

    If you plotted my son's grades you'd see a dip in his performance during the 2020-21 school year. Was moving to remote schooling for a year the right thing to do? Yep.

    1. MontyTheClipArtMongoose

      How do homeschooled kids do on these tests? Or, are homeschooling parents not required to have their charges taken them?

      1. Joseph Harbin

        Not really sure. It may depend on the state. In Calif., all kids are expected to take standardized tests. Home-schooled kids generally do well, they say, but that may have more to do with being the type of family that can afford to home school their kids. "My son was 1st in his class!" "Well, isn't he special!"

  3. rick_jones

    These things tend to wash out in later school years, and I very much doubt that by the time our fourth graders graduate from high school they'll be noticeably different from any other graduating class.

    How to square that assertion with:

    Here's the key thing to look at: gains in elementary school don't matter unless they're sustained throughout the entire 12 years of primary schooling.

    And the assertion in https://jabberwocking.com/new-study-says-american-students-are-improving-are-they/ is they weren’t/aren’t. So gains made in fourth graders are lost by high school, but losses will be washed-out?

  4. middleoftheroaddem

    Many private schools remained open to in person learning during the same period. I SUSPECT the learning loss, if there is even a loss, at private schools will be less than that of public schools.

    1. Spadesofgrey

      Dude, no they didn't

      If the state health departments told them to shut, they did. Most schools went partially in 2020-21. Do you need backhanded???

  5. ScentOfViolets

    What I have heard -- and this is, of course, purely anecdotal -- is that test peformance has a bimodal distribution. I could pretty up the 'why' some, but basiclly it comes down to parents who think their kids should be learning something at school vs parents who think that school is state-sponsored babysitting. Or so I've been told 😉

  6. Atticus

    This makes me grateful I lived in Florida during the pandemic. The decline in test scores was completely predictable due to e-learning. This is the main reason all Florida schools were open for in-person learning starting in August 2020. That 2020/2021 school year districts still offered e-learning for those that chose that option. In our district (Hillsborough County) about 40% chose e-learning to start the year but by the end of the year it was less than 10%. It was obvious kids weren't getting covid in school and the benefits of being in the classroom far outweighed any risk.

  7. skeptonomist

    "The usual rule of thumb is that ten points on the NAEP equals one grade level."

    I don't believe this, if the points are on tests applied to all ages (that is not somehow normalized to age). Kids learn different things at different rates. They learn reading very fast at fairly young ages, while learning math is a long process which can go on past college. So I would expect absolute reading scores to increase little from say, age 8 to age 18, while math scores could increase a lot (depending on what is covered in the test). At any rate it would be helpful to see the normal progression of reading and math scores with age before making sweeping pronouncements about these results.

Comments are closed.