NAEP Report: A Closer Look at Trends in the Achievement Gap
A new report from "the nation's report card" (and my own Education Week story yesterday) emphasizes progress in closing achievement gaps for black and Hispanic students between the early and mid-1970s and today.
While this is mostly true (except for one category, the Hispanic-white gap for 9-year-olds in math), it doesn't tell the whole story. (Thanks to Bob Rothman from the Alliance for Excellent Education, who posted a comment on my story that got me thinking.)
Much of that narrowing of the achievement gap was actually accomplished by the mid to late 1980s, the data indicate. It's ebbed and flowed a bit since then, but in most cases, the gaps are no smaller today than they were two decades ago. In fact, they're sometimes larger, though not by amounts deemed statistically significant.
For example, the black-white achievement gap for 13-year-olds in reading reached its narrowest point in 1988, at 18 points, compared with 23 points in 2012.
In math, the black-white achievement gap for 9-year-olds was 25 points in 1986, the exact same figure as for 2012.
The only instance I could find where the numerical achievement gap was smallest today was the black-white gap for 9-year-olds in reading. It reached 23 points in 2012. In 1988, the figure was 29 points (though this difference is not considered statistically significant).
Of course, achievement gaps are not the only thing to be concerned with. The hope is that all students will make progress over time. The good news here is that whites, blacks, and Hispanics all have seen increases in their average scores since results were first available on the long-term trends report. But the results get more complex when comparing the results from the mid- to late-1980s to today. In reading, average scores for 9-year-old blacks are improved today over 1988, but for 13-year-olds and 17-year-olds the difference was not statistically significant.
The new data come from the National Assessment of Educational Progress long-term trends report. Keep in mind that this assessment is different from the main NAEP in reading and math administered every two years. You can learn more about the differences here. I should also note that the NAEP study does not report out data for Asian/Pacific Islander students or for American Indian (including Alaska Native) students because of limits on the data available. They were included in the national samples, however, and some data for them can be found on the NAEP Data Explorer website.
Given how much data we're talking about with the NAEP long-term trends report, it's hard to identify perfect trend lines. Any analysis is complicated by the multiple factors involved: We're talking about:
• Two subjects, reading and math;
• Two gaps, black-white and black-Hispanic;
• Three age levels (9, 13, and 17); and
• Up to 13 different assessment years.
Of course, any talk of progress in closing gaps can quickly introduce an element of politics, as people may wish to use the data to argue that a particular evolution in education policy explains the changes, such as the push for tougher accountability measures. So don't be surprised in coming days if these data are used to defend a number of different agendas. I'll stay out of that debate here, but will do my best to shed a little more light on what the data show.
Now, let's do the numbers!
Below I've reproduced several graphics from the NAEP report that provide a detailed look at changes in the gaps over time. Take a look and draw your own conclusions about what it all means. But be sure to keep an eye out for those small asterisks. They signal years in which the results are considered different by a statistically significant margin.
In the chart below, you'll notice the gap was smallest in 1988.
The Hispanic-white gap in 1988 was exactly the same as in 2012, this next chart shows.
In math, this next chart shows the black-white achievement gap was the same amount, 25 points, in 1986, 1994, 1996, and 2012.
As this final chart shows, the Hispanic-white gap in 2012 was 17 points. This is NOT considered a statistically significant change from 1975, though it is measurably larger than for several points in time, including the 1999 results.
I'll close by briefly tackling one other complication raised by the NAEP data. I was a little puzzled about how it could be that average national scores for 17-year-olds were about the same in 2012 as they were back in the early 1970s, even as whites, blacks, and Hispanics all saw progress. For help in making sense of this, take a trip over to the Change the Equation blog.
"The reason for this apparent impossibility?" the blog post says: "Black and Hispanic students, who unfortunately lag behind their white peers, make up a much bigger share of the population now than they did in 1973. That brings down the total score."
Anyway, as always there's lots to mine in the new NAEP report. But it's complicated stuff, and there are plenty of caveats, so take your time!