NAEP spin

Fourth- and eighth-grade math scores continue to rise and reading scores are up in fourth but not eighth grade, according to the just released Nation’s Report Card.

Black and Hispanic students are improving faster than whites, observes Education Trust.

Since 1996, the percentage of our nation’s fourth-grade students performing below the Basic achievement level in math has been cut in half, from 39 percent to 19 percent, with even stronger improvement among poor and minority students (from 73 percent to 37 percent for African Americans, 61 percent to 31 percent for Latinos, and 60 percent to 30 percent for poor students). At the same time, higher performers also posted significant gains, increasing the ranks of students at the Proficient and Advanced levels.

In reading, the gains are coming at the bottom and middle levels; eighth-graders aren’t improving, notes Time.

This kind of stagnation, along with the disappointing results in 8th grade overall, will further fuel the current debate among educators over how America teaches reading. It appears that the recent emphasis on phonics and reading mechanics, encouraged by the Bush Department of Education, is helping in early years, but something different is needed to take students beyond an elementary level.

No Child Left Behind defenders are claiming credit for the progress. Critics say scores were rising faster before NCLB kicked in.

I agree with Kevin Carey’s analysis on The Quick and the Ed:

I’ve said it before and I’ll say it again: NAEP results, particularly in elementary math, absolutely disprove the notion that public education is unreformable and nothing can be done for disadvantaged students. And while there’s a lot of talk about how NCLB’s focus on bringing up low-performing kids is pulling down the top and short-changing the gifted, I’ve yet to see any compelling evidence that this actually true. Our society is relentlessly focused on providing all manner of opportunity to people with an excess of talent, money, and social capital, and no federal law — particularly one narrowly focused on education — is going to change that.

More here on the likely consensus: We’re moving slowly in the right direction.

Update: “The Eclectic Linda” Seebach, now retired from the Rocky Mountain News and blogging, mocks the Minnesota spin on NAEP results: Yes, Minnesota ranks above average but only because the state’s students are “way whiter than average.” Minnesota’s black and Hispanic students do not outperform similar students elsewhere.

About Joanne


  1. Do you ever *read* the NAEP “reading” questions?

    Please. The NAEP is not a reading test. It’s predominantly a writing test. Check out the extended response questions here.

    I can never figure out why people are so bowled over by the NAEP. I keep hoping it’s because they blindly read the results without looking at the test, but it always worries me that there might be people who actually think an extended essay is relevant to a reading comprehension test.

  2. If you want even more of a shock copy and paste one of the reading selections into Word and have it do a readability analysis. There was a story about the Mir Space Station on the 4th grade test several years ago that was on an 8th grade reading level.

  3. MiT raises a good point. (Of course, these crude readability calculators should not be the final decider with respect to designing rerading comprehension test.)

    I suggest you run some of your 4th – 8th textbook passages through the the Word readability analyzer and you’ll see that the textbooks that we expect children to read are also at a ridiculously high readability level.

    So, these difficult reading passages in NAEP do seem to represent the kind of material that children will be expected to be able to comprehend whenthey read to learn.

    Nonethekless, two wrongs don’t make a right.

  4. The 4th grade reading sample here:

    averages 15.6 tokens/sentence and has a Hayes LEX
    value of about -18. This would put it at about the
    reading level of an old 1920s Elson-Gray 6th grade
    reader. It is a bit easier than Sherlock Holmes
    stories in terms of grammar and word choice (although
    much easier in terms of tracking plot).

    It is probably reasonable for the 4th grade reading
    test to contain some texts that are higher and some
    that are lower than the reading level expected. We
    do, after all, want to know if some kids are reading
    *above* grade level.

    Still, the sample is above 4th grade by 2 to 4 grade levels
    in terms of grammar and vocabulary.

    -Mark Roulo

  5. So what? Maybe the text could be more readable. It hardly matters. It’s supposed to be a reading test, and the required response is a 2-3 paragraph essay.

  6. Sister Howitzer says:

    “Please. The NAEP is not a reading test. It’s predominantly a writing test. Check out the extended response questions here.”

    What proportion of the questions are extended response? I can’t find any information about the distribution of the question types. I don’t have a problem with them if there are just a few.

    I put the text sample from View here into this analyzer, and got a Flesch-Kincaid Grade Level of 4.74.