Students in urban schools are doing better in reading and math, even in Atlanta, which has been embroiled in a cheating scandal on state exams.

Federal officials said there was no evidence that the cheating had carried over to the National Assessment of Educational Progress — called the "Nation's Report Card" — and that Atlanta fourth- and eighth-graders have made substantial gains since 2002.

The national test is administered by independent officials rather than by the school district. Atlanta is one of 21 urban districts that volunteered to be part of the federal testing program, which is congressionally mandated to gauge how students are performing using a uniform measure.

Federal officials warned against comparing the urban districts that participated in the national test because they vary widely in student makeup, teacher experience and culture. Still, the urban districts' results mirror results released in a national report last month — students made progress in math but their reading scores have mostly remained stagnant in the last two years.

Since 2002, though, reading scores have climbed steadily in most participating cities for fourth- and eighth-graders.

"Urban schools in general are getting better. But we are determined to make them better still," said Michael Casserly, executive director of the Council of the Great City Schools. "We are not satisfied but we believe that we are on the right track — and the new NAEP data bolster our confidence."

In Atlanta, 24 percent of fourth-graders are proficient in reading, compared to 11 percent in 2002 and 21 percent two years ago, according to the report. Those scores outpace the national average for urban districts, which is 23 percent proficiency.

Eighth-grade math students hit 16 percent proficiency, up from just 6 percent in 2003 and 11 percent two years ago. But that trails the urban district average of 26 percent.

"The travesty of all of this is there are more and more indicators that suggest the system did not have to cheat," Atlanta schools Superintendent Erroll Davis, who took over in July, told The Associated Press. "The educational achievement levels are not near where we want them to be, but we are continuing adding value at rates faster than other systems are adding value."


— In Boston, 27 percent of fourth-graders passed muster in reading, compared to 15 percent in 2002. And 33 percent of eighth-graders were proficient in math, compared to 18 percent in 2003.

— Chicago saw 20 percent of fourth-graders score at the proficient level in math, compared to 10 percent in 2003, while 20 percent of eighth-graders passed muster, compared to just 9 percent eight years prior.

— For Detroit, where the troubled school district is being run by the state, there were small gains since 2009 — the first year the district participated in NAEP — but the numbers are still lagging behind many other districts: 69 percent of fourth-graders scored below basic in reading, and just four percent of eighth-graders passed muster in math.

In July, Georgia investigators found widespread cheating in nearly half of Atlanta's 100 schools on state standardized tests dating back to 2001. Tens of thousands of students were affected by what experts say is the largest test cheating scandal in U.S. history. Investigators said nearly 180 educators gave answers to students, changed answers on tests after students had turned them in or ordered subordinates to cheat. Teachers who tried to report the cheating were retaliated against and punished, creating a culture of "fear and intimidation" in the district, investigators reported.

The educators face possible criminal charges and could lose their teaching licenses. So far, eight teachers and three school administrators have lost their certification with the state. Many of the educators had resigned or retired when the report was released over the summer, but the ones remaining have been placed on leave and are in the process of being fired.

But the national test scores show that most Atlanta students were learning despite the cheating on the state test, experts said.

"The NAEP results represent the district as a whole, not this school or that school where there might have been cheating occurring," said Cornelia Orr, executive director of the National Assessment Governing Board, which administers the test. "I don't think you can assume that because it went on with a certain group of students in certain schools, gains were not across the board."

The national test doesn't come with the same pressure as the state tests, which are used to determine whether a school meets federal benchmarks, experts said. And the students who take the test are chosen by federal officials to create a sample that represents the entire district.

The testing problems in Atlanta schools first came to light after The Atlanta Journal-Constitution reported that some scores were statistically improbable. The state released audits of test results after the newspaper published its analysis.

A state probe also has led to an investigation by the U.S. education department's Office of Inspector General and the Georgia Department of Education, which says the district could owe thousands in federal money for low-income schools that have high test scores.

Follow Dorie Turner at http://www.twitter.com/dorieturner.