« A Survey on Science | Main | Community College Proposal Coming from Obama »

White-Black Achievement Gap: Stagnation Among Older Students


This morning, the top statistical wing of the U.S. Department of Education is releasing a study on the achievement gap between white and black students. There are more details in EdWeek’s story on the subject, which describe the mixed bag of results. A couple points worth special attention:

—Much of the progress in closing the achievement gap, at least in individual states, appears to be occurring at the 4th, rather than 8th grade level. Fourteen states, plus the District of Columbia, narrowed the gap in math from the 1990s to today. They are: California, Connecticut, Delaware, D.C., Florida, Georgia, Louisiana, Massachusetts, Michigan, Mississippi, New Jersey, Pennsylvania, South Carolina, Texas, and Virginia. But among 8th graders, only four states narrowed the gap in math: Arkansas, Colorado, Oklahoma, and Texas. There was a similar drop-off, though not as pronounced, from 4th to 8th grade in reading.

Should be interesting to see how states spin these scores. Some of them may choose to focus on having a more narrow overall gap between white and black students than the national average—while ignoring the fact that the disparity in their states hasn't lessened since the 1990s. Look to state and local coverage for a breakdown over the next 24 hours.

The mediocrity of older, American students in math has received a lot of attention in recent years. While younger students have made steady progress in math on the NAEP in recent years, scores among 17-year-olds have basically been a flat-line for decades. This report seems likely to prompt a re-examination of why this stagnation is occurring. Meanwhile, Congressman George Miller sees the continued achievement gap among 13-year-olds in reading as a factor contributing to students dropping out of high school.

—As with any study of NAEP scores, this one will almost surely yield discussions about No Child Left Behind’s impact, particularly since a major goal of the law was improving the performance of low-achievers. But a review of these results shows that finding a clear NCLB-related theme is difficult. Since 1999, the achievement gap among 9- and 13-year-olds closed, but not significantly; in reading, it narrowed significantly for students in both age groups. Among 4th graders, the gap has remained stagnant since 2003, but it narrowed in reading by a three-point, significant margin. The 8th grader achievement gap has closed significantly since 2003, but it’s been unchanged in reading since then.

So tell me: Can you detect a clear breakdown from the NCLB era in all that?


Yes, if you look at main NAEP, you can see pre- and post-NCLB differences, taking 2003 as the real start of NCLB's impact. Sorry for the length, but here are some details:

(The numbers in parentheses indicate the size of the white-black score gap; whites outscored blacks at all points.)

In Math, both the rate of gap narrowing and the rate of score gain for blacks has slowed in the 2003-07 period compared with the 2000-03 period, that is from the pre-NCLB period to the NCLB period, on the main NAEP. [See figures 2 and 4, page 7.]

1. Grade 4 math, main NAEP: The gap was the same in 1990 as in 2000 (31), narrowed steadily from 1992 (25) to 2003 (27), but has not narrowed since then (the 1 point closure to 26 points in 2007 is not statistically significant). The rate of gap closing has slowed from 2000 to 2003 (3 points in 4 years) compared with 2003-07 (1 point in 4 years). Also, the rate of gain for blacks was much faster from 2000 to 2003 (13 in 3 years) than from 2003-2007 (6 in 4 years).

2. Grade 4 math, main NAEP: There was more closure from 2000 to 2003 (5 points in 3 years) than 2003 to 2007 (4 points in 4 years). The rate of absolute gain for blacks was faster in 2000-03 (9 points in 3 years) than in 2003-07 (7 points in 4 years).

In Reading, both the rate of gap-closing and the rate of absolute score improvement slowed in both grades 4 and 8 from pre-NCLB to post-NCLB, on the main NAEP tests. [Figures 17 and 18, page 31.]

1. NAEP main, grade 4: The gap did close 2 points from 2005 to 2007. It had remained constant from 2002 to 2005. The largest gap closing periods were 2000-02 (5) and 1994-98 (7). (At various times the gap had also widened.) Similarly, the largest gains for blacks were 2000-02 (9) and 1994-98 (8). Thus, the rate of score improvement for blacks and the rate of closing the black-white gap have both slowed compared with the 2000-02 and 1994-98 pre-NCLB periods.

2. NAEP main, grade 8: The gap remained statistically unchanged from 1998 through 2007 (wavering between 26 and 27). It had closed 4 points from 1994 to 1998. The NCLB period is not worse than that immediately before it, but it is worse compared with the mid-1990s. Similarly, black's actual average score was statistically unchanged from 242 in 1998 to 244 in 2007. The same 2-point gain (242 vs 244) between 2005 and 2007 is statistically significant, but note that the scores were 244 in both 2002 and 2003, meaning that from 2002 or 2003 to 2007, the NCLB period, there has been no overall score gain for blacks. However, there were improvements in the pre-NCLB period: from 1994-98, the gap closed 4 points while the absolute score was up 7 points. As with grade 4 reading, NCLB represents regression compared with the mid-1990s.

The page numbers are from the reportyou can download from http://nces.ed.gov/nationsreportcard/pubs/studies/2009455.asp

FairTest has previously analyzed NAEP results under NCLB (see www.fairtest.org) as has the Civil Rights Project - with similar results, which is not surprising given this study relies on the previous NAEP reports. We found similar slow-downs, for the most part, for Hispanics.

Thanks, Monty, for the detailed explanation. I am especially intrigued by your point 2 - that NCLB impact represents a regression compared to the mid-1990s. So, something we were doing prior to NCLB was beginning to make headway, then we interrupted it with NCLB and its side effects? Is that a fair interpretation?

Monty Neill loves to play the game of comparing 2003-2007 to 2000-2003 to show pre and post nclb effects. Since nclb became law in January of 2002, this has always been a head-scratcher for me.

I won't go through the true analysis which compares 1998-2007 to 1989-1998, contrasting the period of the full flowering of standards based reforms to the previous period. That's for another day.

Let's just look at the conclusions in the executive summary today.

4th and 8th grade math scores in 2007 for both black and white students were "higher than in any previous assessment, going back to 1990."

"This was also true" for 4th grade reading scores for both blacks and whites in 2007.

Nationwide gaps were "narrower" in 2007 in 4th and 8th grade math and 4th grade reading "than in previous assessments."

In the long term, math scores for both 9 and 13 year old blacks and whites were "higher in 2004 than in any previous assessment."

For age 9 reading, scores for both black and white students "were higher in 2004 than in any previous assessment, going back to 1980."

We have work to do on 8th grade reading where the numbers have been largely flat over the entire period and are only slightly better now. And we must accelerate all gains. But the gains we have experienced recently are palpable and historic.

Poor ol' nclb. It gets hit either way. Since both black and white scores have gone up, the Act gets slammed for not closing the gap as much as it could have, had white scores been flat. But you can be sure if white scores or top decile scores had been flat or gone down, the critics would have blasted the Act for causing a focus on the bubble and/or lower performing students. (Well - they say that, too, in spite of the data!)

Should Mr. Neill and his ilk have their way in weakening or destroying the standards based reforms culminating in nclb, which I fervently wish will never happen, I ask just as fervently that all who have read or contributed to these comments commit to come back in 10 years and look at the naep data together again. Sadly, I predict it would be an undebatable and very sad sight.

I apologize for not including a link to the source in the released report for my quotations. Here it is:


Let's focus on the two issues that should jump out as key.

1. Student performance grew the most in the 90s when the economy grew.

2. The ability to read for comprehension, to "read to learn," is essential for high school and beyond, And, as predicted, 8th grade Reading scores show that NCLB may improved decoding skills or forced everyone to work harder, but it has failed to address reading comprehension.

In fairness to "poor ol nclb," it makes sense that math instruction has improved. In 1990 or before, how many elementary or even middle school teachers had training in math instruction?

Or kids need numeracy, just like they need to be thoughtful consumers of math and stats, and problem-solving, but I am more worried about Reading stagnation in older grades than pleased with algebra increases.

But then again, I want to be consistent and to be fair to NCLB supporters. In what alternative universe could we have expected reading comprehension to have increased during an era when cell phones, iPods, video games, and other digital miracles emerged? In the long run, those technological changes should do great good. In the short run, wasn't it inevitable that reading comprehension would be undercut.

When we get past his blame game of primitive "accountability," then we can discuss ways that schools can address real challenges in a respectful way.

Check out the book "Between the Rhetoric and Reality",'Dorrance Publishing':9=2009. It may quite possibly hold the clue towards decreasing the country's horrendous Black/White Academic Achievement Gap!

Comments are now closed for this post.


Most Viewed on Education Week



Recent Comments