More than 90 percent of Georgia schools were cleared of concern about testing irregularities in a review conducted by the governor's Office of Student Achievement and the state's testing vendor, CTB-McGraw Hill.

"This is good news," state school Superintendent John Barge said. "It shows that the measures we put in place are having a positive impact."

The review, the third annual one conducted since the cheating scandal that engulfed Atlanta Public Schools in 2009 after reporting by The Atlanta Journal-Constitution, examined the percentage of incorrect answers that were changed to correct ones on the 2011 Criterion-Referenced Competency Test. The test, covering English, language arts, reading and math, is given to Georgia students in grades 3 through 8.

Schools were grouped in four categories:

-- clear of concern, with less than 6 percent of classes flagged for incorrect answers changed to correct answers

-- minimal concern, with 6 percent to 10 percent of classes flagged for incorrect answers changed to correct answers

-- moderate concern, with 11 percent to 24 percent of classes flagged for incorrect answers changed to correct answers

-- severe concern, with 25 percent or more classes flagged for incorrect answers changed to correct answers

The report does not say any school cheated on the CRCT.

In Atlanta Public Schools and the districts in Cobb, DeKalb, Fulton and Gwinnett counties, a total of 20 schools were in the moderate- or severe-concern categories.

The three schools -- all in metro Atlanta -- that fell into the state's severe concern category were: International Student Center in DeKalb and Marietta Center for Advanced Academics and Autrey Mill Middle School in Fulton County.

Just under 38 percent of classes at the International Student Center were flagged for answer changes, the highest percentage in the state.

"It’s important for parents to know that the data released today does not necessarily indicate that any wrongdoing occurred at any school," said Walter Woods, spokesman for DeKalb County Schools.

"It is a statistical study of the number of wrong-to-right answer changes at each school; and the identification of those schools with multiple classrooms above the state's mean of wrong-to-right answer changes," Woods said. "But the school district will meet all of the state's expectations in conducting a thorough investigation to determine why these classrooms were flagged."

Woods noted the district has put in place a series of measures to enhance test security, including installing cameras in the rooms where tests are stored, limiting the number of staff members who have keys to those rooms, and making sure tests are monitored.

MCAA Principal Jennifer Hernandez said an outside investigation found no evidence of procedural or personnel wrongdoing there.

"Parents should be confident that the 2011 assessments reflected the high level of instruction students receive at the school," she said.

Principal Jimmy Zoll said he was “very confident there was no cheating" at Autrey Mill. "As always, all staff members who administer the test are trained and the administration monitors the testing environment,” he said.

A pair of APS schools that were at the heart of the cheating scandal, Stanton Elementary and Parks Middle, were in the moderate-concern category, but spokesman Keith Bromery said the district's renewed focus on test security is paying off.

"Over the past two years, APS has instituted a number of test security measures, including sealed containers for testing materials with tamper-proof security seals and locked safe rooms in all schools that are accessible only by the principal and the testing coordinator," Bromery said.

"We also instituted the two-person rule for counting all testing materials. We believe these measures, along with mandatory annual ethics training, have combined to make the testing environment at APS as secure as humanly possible," he said.

The state report, which examined erasures on 125,000 test documents per subject for each grade level, relied upon a mathematical formula that varied with grade level and subject to determine whether a class should be flagged.

State Board of Education members, briefed on the report Thursday, voted to place state monitors in severe-concern schools when the CRCT is administered this year.

The board also voted to have the state conduct on-site audits as necessary, rotate teachers in minimal-, moderate- and severe-concern schools during this year's CRCT, and make random checks in moderate-concern schools when the CRCT is given.

The first state review of answer changes on the CRCT in 2009 found 80 percent of schools fell into the "clear of concern" category. In 2010, that percentage rose to 87 percent, and it edged up to 90 percent in 2011.

Meanwhile, the percentage of schools in the severe-concern category stands at .2 percent, down from 4 percent in 2009 and .5 percent in 2010.

About 7.4 percent of schools were in the minimal-concern category, down from 10 percent in both 2009 and 2010. Schools in the moderate-concern category fell to 2.6 percent in 2011 from 6 percent in 2010 and 3 percent in 2009.

Fulton had three schools in the moderate-concern category and one in the severe-concern category, but Diane Jacobi, a North Fulton County PTA member, said she is not concerned about cheating at schools in her district.

Jacobi said the four Fulton schools flagged have students who probably changed their answers after finishing the test in less than the allotted time.

"It does not surprise me that students at those schools have the time to go back over their answers," she said. "That's the way our children are being taught."

Metro area schools flagged

A total of 20 schools in Atlanta and Cobb, DeKalb, Fulton and Gwinnett counties were in the state's moderate- or severe-concern categories because of the percentage of classes flagged for wrong-to-right answer changes on the 2011 CRCT.

APS:

Brown Middle: 11.3 percent of classes flagged, moderate concern

Charles R. Drew Charter: 12.5 percent, moderate

Young Middle: 12.1 percent, moderate

East Lake Elementary: 11.1 percent, moderate

Cleveland Elementary: 16.1 percent, moderate

F.L. Stanton Elementary: 12.5 percent, moderate

Parks Middle: 15.3 percent, moderate

Cobb:

Kennesaw Charter: 11.9 percent of classes flagged, moderate concern

Rocky Mount Elementary: 11.9 percent, moderate

King Springs Elementary: 12.8 percent, moderate

DeKalb:

Pine Ridge Elementary: 11.1 percent of classes flagged, moderate concern

Leadership Prep: 11.1 percent, moderate

DeKalb PATH Academy Charter: 11.7 percent, moderate

International Student Center: 37.8 percent, severe

Fulton:

Gullatt Elementary: 16.7 percent of classes flagged, moderate concern

Autrey Mill Middle: 25 percent, severe

River Trail  Middle: 22.9 percent, moderate

Roswell North Elementary: 11.1 percent, moderate

Gwinnett:

Lovin Elementary: 13.7 percent of classes flagged, moderate concern

Anderson-Livsey Elementary: 13.7 percent, moderate

The full report can be found at www.gaosa.org.

Source: Governor's Office of Student Achievement

Timeline

2008-2009: The AJC reported statistically unlikely gains in test scores at some Atlanta schools on the Criterion-Referenced Competency Tests.

February 2010: After a state-commissioned erasure analysis flagged 58 Atlanta schools, the state ordered the district to investigate.

August 2010: The state rejected the Atlanta district’s results as inadequate, and then-Gov. Sonny Perdue named two former prosecutors to lead a criminal investigation.

July 2011: The state review determined that cheating occurred in more than half of Atlanta’s elementary and middle schools in 2009. Investigators named about 180 teachers and administrators, some of whom confessed to altering test papers. A Fulton County grand jury is considering criminal charges.

March 25, 2012: The AJC reported that standardized test scores across the country were suspicious and are unlikely to have come without human intervention such as tampering.