High school wasteland: Demographic changes do not explain test-score stagnation among U.S. high school seniors

Source: NCES

Source: NCES

The National Assessment of Educational Progress, run by the U.S. Department of Education, is the only test that is administered in schools across every state in the nation. Fourth and eighth graders across the country have shown meaningful progress on it since the early 1990s, especially in math. The 2013 results in these younger grades even showed small improvements from 2011. (Source: NAEP Mathematics and Reading 2013).

But the 2013 testing results for twelfth graders, released May 7, 2014, are woefully stagnant. The scores for high school seniors haven’t improved at all since 1992, when reading tests were first administered. Indeed, today’s reading scores are actually lower than they were in 1992. The math results, which date back to only 2005, show a modest increase right after that first year. But it’s been complete stagnation since.

It’s hard to make sense of this data. How do you explain why there are improvements in fourth and eighth grade, but not twelfth?

This article also appeared here.

This article also appeared here.

One explanation could be demographic changes. Today there are many more minorities in twelfth grade. The Hispanic population, which has typically scored lower, has exploded. Hispanics have tripled from 7% of high school seniors in 1992 to 20% in 2013. The white population, which has traditionally scored higher, has declined from 74% to 58% in the same time period. More students are diagnosed with a disability today (11% in 2013 vs. 5% in 1992). More students are English language learners.

Furthermore, the high school graduation rate has jumped from 74 percent to 81 percent. That means that the weakest students who used to drop out of high school and, were previously not around in twelfth grade to be tested on the NAEP, are now taking the test.

“Our twelfth grade population is our population. And we don’t explain away test scores based on demographics. But it’s useful to keep in mind that we are seeing increases in subgroups that have traditionally performed lower,” said John Easton, acting commissioner of the National Center for Education Statistics and director of the Institute of Education Sciences.

The pool of high school students being tested in 2013 is clearly a weaker pool than the one tested in 1992. That might be masking improvements that we otherwise would have seen had demographic changes not occurred.

But here’s the thing. When you look at top achieving students in the top 75th and 90th percentiles, their scores are FLAT. (See the NCES charts below). High achieving students aren’t improving at all. So you can’t blame the infusion of more low performing students in the testing pool for the disappointing test scores. Even if we hadn’t introduced a greater number of weaker students into the mix, the scores of our high school students would still be stagnant.

Indeed, when you drill down by percentile, it’s the weakest students who are showing modest improvements. If not for their improvements, the national average would have declined!

Source: NCES

Source: NCES

Screen Shot 2014-05-07 at 8.52.32 AM

Related stories:

Connecticut and Arkansas show unusual gains on test of high school seniors; Massachusetts shows sharp decline

Washington DC and Tennessee post huge gains in math and reading in 2013 while nation shows small improvement


POSTED BY Jill Barshay ON May 7, 2014

Comments & Trackbacks (9) | Post a Comment


>How do you explain why there are improvements in fourth and eighth grade, but not twelfth?

Another explanation is that it’s easier to “teach to the test” in primary and middle school mathematics because the problems are predominantly calculation-based, which can be drilled. Analogous to using techniques like “phonics” to teach students to read in the early grades, the weakness of which is then exposed in the “4th grade slump”, the folly of teaching mathematics as skills rather than thoughtful problem solving is eventually exposed in the NAEP 12th grade results.

Jill Barshay

@CCSSIMATH You make a compelling argument. Do you think the new Common Core PARCC tests will steer U.S. curriculum more toward thoughtful problem solving? And is it possible to move to thoughtful problem solving if students haven’t mastered their operations (adding, subtracting, multiplying, dividing)?

Joyce Martin

Why can we not involve the role of parents and caregivers in these discussions? No matter what you do to improve our schools the family still plays an extremely important role in a student’s performance. It goes way beyond actual income. It is the attitude toward education and it’s role in their children’s future that is lacking. Living with adults who for one reason or another did not engage in school or become lifelong learners is a true indicator that children may not be expected to demonstrate the drive and determination needed to be successful. How are we going to achieve a change in the mindset of the parents and guardians? That’s what I’m trying to figure out.


“Fourth and eighth graders across the country have shown meaningful progress on it since the early 1990s, But the 2013 testing results for twelfth graders, released May 7, 2014, are woefully stagnant”
“Indeed, when you drill down by percentile, it’s the weakest students who are showing modest improvements.”

Both of the above conclusions are related; learning is not a linear activity. By improving middle school math scores, you will not be able to improve algebra and geometry scores. The correct model is one of steps with the higher steps being harder. High school tests algebra and geometry skills, and the results are all or nothing. You can improve your score to certain level, but cannot improve it beyond that. Hence, the ability to increase the scores at lower level, and the performance of the lower percentiles. In fact, the lower percentile scorers are actually cleaning up on the easier answers.

Jill Barshay

@Vijay Thank you for your sharing your thoughts. I suspect you are right. The weakest high school seniors might be getting better at adding, subtracting, multiplying and dividing (using their calculators), but not getting the basics of algebra and geometry.

Here’s an example of that kind of basic math question:
360 X 0.3 =
A. 10.8
B. 108
C. 120
D. 980
E. 1,080

Here is a link to sample test questions.

The education policy question is what to do about it? You’ve got a lot of kids in algebra classrooms that aren’t ready to learn algebra and that limits how deeply you can teach the subject to the top students. Do you go back to tracking? Change the curriculum?


It’s senioritis. Seniors don’t care much about their tests when they’re already ready to get out of high school.

Jill Barshay

@Holmes This question was asked by a reporter during a press briefing May 6, 2014. Cornelia Orr, executive director of The National Assessment Governing Board, said it was “urban myth” that seniors are blowing off the test when they sit down to do it.

On the NAGB website, there is a whole section about how motivated 12th graders are to take the NAEP.


Unmotivated seniors is a lazy answer. What evidence exists that eighth graders and fourth graders are highly motivated, but lose all interest in senior year.

Felix Carpio

–I don´t see why the pessimism. If we consider that data before 1998 was not accommodated, and show unreliable sudden up and downs, and must be discarded, we can then conclude that Reading Skills have not decreased in spite of the increasing numbers of minority students being assessed. That means significant improving. Whether the always-bitching ever-pessimist ones like it or not.
–And the Math results are even better, they show improving no matter what, no matter if more disadvantaged minority students are in game. We know, according to data not shown here, that minorities have been improving dramatically since the 1970s. So cheer up and stop bitching.

Your email is never published nor shared.