School Choice & Charters Opinion

Deconstructing NAEP Reading Scores for Cities

By Diane Ravitch — June 01, 2010 4 min read
  • Save to favorites
  • Print

Dear Deborah,

This year the federal testing program (the National Assessment of Educational Progress, or NAEP) expanded to include 18 urban districts. The testing of urban districts is known as TUDA (or, Trial Urban District Assessment). TUDA was launched in 2002, in response to a request by the Council of the Great City Schools.

In 2002, the following districts voluntarily participated in testing a sample of their students in reading in 4th and 8th grades: Atlanta, Chicago, the District of Columbia, Houston, Los Angeles, and New York City. In 2003, both reading and math assessments were administered in 10 cities, with the addition of Boston; Charlotte, N.C.; Cleveland; and San Diego. Austin, Texas, joined the program in 2005.

The new participants in 2009 testing include Baltimore; Detroit; Fresno, Calif.; Jefferson County, Ky.; Miami-Dade (County); Milwaukee; and Philadelphia.

NAEP is recognized as the gold standard of achievement assessment: its tests are excellent, combining multiple-choice questions with short-answer questions and constructed-response questions so that students actually have to explain what they know or how they would solve a problem or write a short essay. NAEP is a no-stakes exam; no teacher or student gets a reward or a punishment based on its results. And no one can prepare for NAEP because no one knows who will take it. For all these reasons, NAEP tends to be more trustworthy than state exams, which typically produce inflated scores.

In 4th grade, the highest-performing district is Charlotte, followed closely by Austin. The districts that have made the biggest gains since 2002 are Atlanta and Washington. The new additions to TUDA include some very low-performing districts, such as Baltimore, Detroit, Fresno, Milwaukee, and Philadelphia. The districts in which 50 percent or more of 4th grade students are “below basic” are Atlanta, D.C., Chicago, Baltimore, Fresno, Los Angeles, Milwaukee, Philadelphia, Cleveland, and Detroit. In Detroit, a depressing 73 percent of students in 4th grade are “below basic.” Nationally, 34 percent of students in this grade are “below basic.”

A demographer would probably see a close correlation between NAEP scores and poverty.

Eighth grade reading scores have been stagnant nationally. The overall reading scores for the nation in this grade were precisely the same in 2009 as they were in 1998. Among the TUDA districts, two saw significant improvements in this grade: Atlanta and Los Angeles.

Four urban districts are at the national average for this grade: Austin, Charlotte, Jefferson County, and Miami-Dade. This is very impressive. These four districts seem to be doing something very right: Their profile for 8th grade students looks very similar to the national profile and is significantly higher than the large-city average. What is especially impressive about Miami-Dade is the strong performance of its Hispanic students in 8th grade; the white-Hispanic gap in Miami is only 12 points, compared with 28 points in New York City, 30 points in Houston, 31 points in Austin, and 32 points in Los Angeles.

Cleveland, Milwaukee, and D.C. are the three cities with large numbers of students using vouchers and charter schools. Charters are folded into the public school numbers, though they can sometimes be disaggregated. Nick Anderson of The Washington Post analyzed the D.C. numbers and concluded that Washington was one of the few districts (other than L.A. and Atlanta) that made gains in 8th grade, if one excluded the charters. The regular D.C. public school system made gains, he found, but the scores of D.C.'s charter schools declined, giving the impression of stagnant performance. By excluding the charters (which are supposed to be “public schools”), Anderson found that D.C. had made significant gains in reading in both 4th and 8th grades.

Another interesting point: New York City often describes the closing of achievement gaps, but the gaps remain distressingly large: in 8th grade, 40-41 percent of whites and Asians are proficient readers, compared with 12 percent of blacks and 13 percent of Hispanics.

One other city bears mentioning, and that is Chicago. Chicago is important because it seems to be a model for Arne Duncan’s national strategy of opening charter schools, closing public schools, and placing greater emphasis on testing. Duncan became superintendent in Chicago in 2001. Fourth-grade students made a significant gain on NAEP from 2002-2003, but have seen no improvement since then. Eighth grade scores have been completely flat since 2002. Scores for black, Hispanic and low-income students in 8th grade are unchanged since 2002. In 4th grade, there are fewer students “below basic” (down from 66 percent in 2002 to 55 percent in 2009), but there has been no improvement whatever at 8th grade.

There is a lot to digest here. Here are a few thoughts: Beverly Hall, the superintendent in Atlanta since 1999, is making a difference. Miami-Dade showed some impressive results in its first appearance in the TUDA. The much-ballyhooed successes in Chicago were a mirage. In the future, I hope that the National Assessment Governing Board displays the public-charter differences so we can learn more about their relative performance over time. Bottom line: Some improvement, and, with some notable exceptions, continued stagnation in 8th grade.


The opinions expressed in Bridging Differences are strictly those of the author(s) and do not reflect the opinions or endorsement of Editorial Projects in Education, or any of its publications.