Commission reports evidence of cheating on tests

109 educators cited; investigative panel calls for cultural change from superintendent

A long-awaited investigative report into alleged cheating in Atlanta Public Schools, released Monday, may have raised as many questions as it answered.

A commission appointed to study suspicious erasures on standardized tests recommended possible sanctions against 109 educators, including 38 principals or other administrators, who may have taken part in inflating scores. But officials declined to publicly identify the educators nor did the report try to explain how similar efforts could have occurred without coordination.

Investigators hired by the commission said widespread cheating appeared to be limited to 12 schools, far fewer than the 43 flagged as most suspicious by state officials. The commission's investigators used a different standard to assess test data.

The commission's report cited no evidence that either Superintendent Beverly Hall, who created an aggressive accountability system tied to test scores, or other top officials orchestrated or condoned cheating.

Finally, the status of the report itself remained unclear late Monday. By a 5-4 vote, the Atlanta Board of Education decided not to formally accept the report from its own "blue-ribbon" commission. The board did send a copy to state officials, who had set Monday as an irrevocable deadline for completing the investigation. But at least one member of the state Board of Education said the unapproved report may not be acceptable.

Hall said the panel's findings involving the 109 employees will be turned over immediately to the district's lawyers for review, further investigation and, if warranted, disciplinary action. Officials cited the continuing investigations in deciding not to identify the accused educators. Investigators based their conclusions mostly on data analysis, noting in their report that "there were no self-admissions by any central office staff, district office staff, or school staff of any wrongdoing."

"Our school system must clear this up, the sooner the better, " Hall said. Of anyone who is found to have cheated, she said: "We will ferret them out, and the consequences will be severe."

The school board directed Hall to outline how she plans to address the report by Aug. 16. Already, she had announced plans for intensive tutoring and other academic help for children at the 58 schools affected by the cheating scandal.

AJC report raises questions

The report's release, and the accompanying confusion, followed a nearly six-month investigation into possible cheating last year at 58 Atlanta schools on the Criterion-Referenced Competency Test, the exam that helps determine whether schools are meeting national and state education standards. The state ordered the investigation after The Atlanta Journal-Constitution reported on statistically improbable increases in CRCT scores at schools in Atlanta and elsewhere in Georgia.

Hall, who has won national accolades for the district's skyrocketing test scores over the past decade, largely escaped criticism from the blue-ribbon panel, made up mostly of people with financial, civic and other ties to the superintendent and the district.

But panel members made clear they want Hall to respond aggressively to their recommendations, including cultural changes to ease pressure on schools to continually raise test scores.

"We know that student achievement and measurable outcomes are critical, " said Gary Price, the panel's chairman. "But that has to be balanced by positive ethical behavior."

78 worked at 12 schools

The panel said its investigators conducted 292 interviews --- nearly half with school district officials present. Most were with teachers and school administrators; just 12 students and parents were interviewed, limiting investigators' ability to study whether individuals' CRCT scores matched their overall academic performance.

Investigators zeroed in on 109 district employees through data and statistics, as well as through "qualified allegations" --- described by the panel's report as backed by plausible though indirect evidence.

The panel said 78 of the employees --- 30 administrators and 48 teachers --- worked at just 12 schools. The findings, Price said, suggest "schoolwide institutional issues" that warrant wholesale changes on those campuses.

At 13 other schools, the panel said, 25 employees --- eight of them administrators --- appear to have acted individually, as did six other employees in some of the remaining 33 schools under investigation.

The Governor's Office of Student Achievement, which analyzed erasures on test papers statewide, flagged 58 Atlanta schools with suspicious changes. The agency cited "severe" concerns about 43 of the schools, where at least one-fourth of all classrooms had questionable numbers of answered changed from wrong to right. The others were said to be of "moderate" concern, meaning that at least 11 percent of classrooms were under suspicion.

The agency set guidelines for Atlanta and other districts, instructing them to establish a chain of custody of test papers to identify the individuals who had an opportunity to change answers. The agency did not tell districts to repeat the analysis of test data.

Yet Atlanta's blue-ribbon panel hired a consulting group, Caveon Test Security, to reanalyze data. The firm used its own algorithm to identify statistical indicators of testing irregularities. In 12 schools, the firm said, there was at least a 1-in-10 million chance that score increases were legitimate. In 13 schools, Caveon found a slightly lesser probability of irregularities.

In the remaining 33 schools, there was as much as a 1-in-10,000 chance of wrongdoing, and all 33 showed significant declines in CRCT results this year. Yet the panel recommended only that the school district further assess those schools.

State officials frustrated

The panel's presentation exposed a rift on the city school board. Several said they wanted more time to digest the report's findings and evidence before they voted to accept it. They said they were concerned about unjustly implicating employees.

"You want to make sure it says what you understand it says, " board member Yolanda Johnson said. "It's just really about being clear on the information."

But that decision may jeopardize the city system's standing, at least in the short term.

State officials, frustrated by several delays, imposed the deadline of midnight Monday last month and said they would sanction the system if it was not met.

Several state officials attended Monday's presentation, and left frustrated by the board's action.

"It's just uncalled for, " James Bostic, a state school board member whose district includes Atlanta, said afterward. "The board didn't accept it. I don't know why [the state] would accept it. It's a great disappointment."

Penalties could involve the loss of Adequate Yearly Progress status for 2009 and 2010, which tarnishes the district's academic reputation. Federal funding also could be withheld.

City board Chairwoman LaChandra Butler Burks, who was on the losing side of a motion to formally accept the report rather than just transmit it, said she would try to work with the state to avoid those sanctions if possible.

Some board members suggested a 48-hour window of review, after which another vote could be taken. But as of Monday evening, no such vote had been scheduled. Butler Burks was uncertain of the board's next steps.

The board's decision clearly surprised the panel. It was revealed Monday that small-group "briefings" with board members were scheduled over the past few weeks to talk about the report.

The five board members who declined Monday to accept the report --- indicating they wanted more time --- did not attend those briefings, according to Butler Burks. She, Emmett Johnson and Reuben McDaniel III did. All three were in the minority voting to accept the report, as was board Vice Chairwoman Cecily Harsch-Kinnane, who was on vacation but received a briefing via telephone.

"It is our expectation you will take a body of work and accept it and then you will decide what to do with it, " said Ingrid Saunders Jones, a panel member and senior vice president, with the Coca- Cola Co.

Too many erasures

The scrutiny of Atlanta's CRCT scores began after the AJC published an analysis in December 2008 that showed improbable gains on tests taken first in the spring and then in the summer by students struggling to master core skills such as reading and math.

A subsequent state investigation named four schools statewide that turned in questionable results for tests taken in summer 2008, including Atlanta's Deerwood Academy.

The state eventually threw out those scores, which were from fifth-grade math retests. The audit found evidence of an abnormal number of erasures at those schools on those retests in which the wrong answer often was replaced by the right one. In some cases, educators involved in the probe denied wrongdoing. Others confessed to erasing students' answers to increase scores.

Last October, the newspaper published a second investigation that showed 19 schools statewide reporting extraordinary gains or drops in state test scores between 2008 and 2009. A dozen of those schools were in Atlanta, including schools where students went in the course of one year from among the bottom performers statewide to among the best. According to the AJC analysis, the odds of making such a leap were less than 1 in 1 billion.

With test monitors and other security measures in place, many of the flagged schools saw substantial declines in CRCT scores this year. At some elementary and middle schools, Atlanta students failed at two, three or four times the rate they did last year. Average scores on some tests plummeted dozens of points.

How the board voted

Atlanta Board of Education members voting to accept the report:

LaChandra Butler Burks, Emmett Johnson, Reuben McDaniel III and Cecily Harsch-Kinnane

Board members voting to delay acceptance:

Brenda Muhammad, Khaatim Sherrer El, Nancy Meister, Yolanda Johnson and Courtney English