Americans Stink at Math (But We’re Much Better Now)

By 07/25/2014

1 Comment | Print | NO PDF |

Elizabeth Green’s story for Sunday’s New York Times Magazine, “Why Do Americans Stink at Math?” is a must-read. Green illustrates our national struggles with math in numerous and at-times painful ways–in particular, read about how customers preferred McDonald’s 1/4-pound hamburger over A&W’s 1/3-pound patty because they thought it had more meat. Her piece is entertaining and seamlessly brings in education topics like teacher preparation, the structure of the school day, poorly aligned textbooks, Common Core, etc. It’s easy to forget she’s writing about math.

But for all the time Green spends documenting the ways Americans stink at math, she never mentions that we’ve gotten much better. That’s unfortunate, because the math results on the National Assessment of Educational Progress (NAEP) are one of the brightest spots in education. Between 1973 and 2012, the long-term trend scores of 9-year-olds rose 25 points. Due to Simpson’s Paradox, where the size of the group can mask aggregated data, the scores of white students, black students, and Hispanic students all gained more than the national average. Scores improved across all performance levels and achievement gaps narrowed. The same trends hold true for 13-year-olds. Across both ages and all groups of students, math achievement in 2012 was higher than it had ever been.

It’s worth noting that the scores for 17-year-olds have been flat overall, although the scores of white, black, and Hispanic students have all risen and achievement gaps have narrowed over time.  Still, the results of 9- and 13-year-olds would have been the most relevant for Green to include because her article mainly focuses on the basic math skills students learn in elementary grades.

No one knows for sure why math achievement has risen so rapidly, but it’s likely some combination of standards-based reforms, rising education expenditures, and falling class sizes. It may also be due to the curricular and instructional changes Green documents; I just wish she’d done a little more math.

–Chad Aldeman

This first appeared on Eduwonk

Comment on this article
  • Stanley Ocken says:

    The author writes:
    Between 1973 and 2012, the long-term trend scores of 9 -year olds rose 25 points.

    It is astonishing that no further information is offered, and that the headline states …”But we’re much better now.”

    The most elementary lesson in interpreting such information is to ask: 25 points out of what? Was the change statistically significant (this just means that the change was not due to random sampling effects, but rather was caused by something else). Was the change educationally significant (did students actually demonstrate a non-trivial gain in knowledge)?

    The trend score rose from 220 to 245 out of a total of 500.
    This would be equivalent to a change from 45 to 49 on a standard grading scale. In a technical sense, this is statistically significant (at the weakest level), namely p < .05 . Whether it is educationally significant is impossible to know from the given data.

  • Comment on this Article

    Name ()


         1 Comment
    Sponsored Results

    The Hoover Institution at Stanford University - Ideas Defining a Free Society

    Harvard Kennedy School Program on Educational Policy and Governance

    Thomas Fordham Institute - Advancing Educational Excellence and Education Reform