Atlanta Public Schools rehires 4 ex-principals named in cheating report

Digging Deep. In July 2011, a special state investigators' report alleged 178 Atlanta educators either participated in widespread test cheating or should have known about it. An Atlanta Journal-Constitution analysis in December 2008 first reported statistically improbable increases in scores on the Criterion-Referenced Competency Test at an Atlanta school. The following year, the AJC published another analysis that found suspicious score changes on the 2009 CRCT at a dozen Atlanta schools. The AJC will continue following this important story.

Four former Atlanta principals who led schools that had questionable standardized test scores are back on the job in administrative positions outside the classroom, according to records obtained by The Atlanta Journal-Constitution.

They were among the 38 principals named in the state's 2011 investigative report into test cheating, but they are the only ones to be reinstated. In all, the investigation claimed 178 educators either participated in test cheating or should have known about it.

Explore» Read previous coverage of Atlanta schools cheating scandal

But personnel files obtained by The Atlanta Journal-Constitution through the Georgia Open Records Act show that the Atlanta school district’s follow-up investigation did not produce evidence to substantiate the charges against these four principals.

Lawyers for the four said their rehirings should clear their names.

Explore» Profiles: Who was named in the indictments

“There’s just simply no evidence whatsoever that any of them participated in any way or knew of any cheating in their schools,” said Michael Kramer, who represents three of the principals.

The state investigation of cheating on the state-mandated Criterion-Referenced Competency Test said that each of their schools had alarmingly high erasure rates, with so many wrong answers changed to right answers that there was nearly no chance that the changes occurred naturally.

The former principals' lawyers said the statistical analysis was flawed, and Atlanta Public Schools couldn't present facts or witnesses indicating these principals cheated. The former principals wouldn't comment, their lawyers said.

“The statistics were misused to create an overinclusion of educators, particularly administrators, who themselves did nothing wrong,” Kramer said.

The four former principals — Marlo Barber of F.L. Stanton Elementary, Anthony Dorsey of Fickett Elementary, Arlene Snowden of Capitol View Elementary and Cheryl Twyman of West Manor Elementary — are all working in district offices outside of schools, according to their personnel files.

But Superintendent Erroll Davis said their employment doesn’t necessarily resolve his concerns about how they supervised their schools, with the exception of Twyman, who Davis said acted responsibly to find out about misdeeds.

Davis said teachers’ job protection rights, sometimes referred to as tenure, mean the school district must clearly prove wrongdoing before a teacher can be fired.

“The only confidence I have at the moment would be a confidence that either a case couldn’t be made or one was made and found insufficient,” Davis said. “You’re innocent until proven guilty.”

Davis initially wrote letters to the four principals notifying them that their employment contracts wouldn’t be renewed. But settlements later sent to three of the principals said the school district was “unable to introduce admissible evidence sufficient to establish that employee engaged in or countenanced wrongdoing.”

Twyman didn’t receive a settlement, but a letter to her signed by Davis said the Atlanta school district’s investigation concluded there was insufficient evidence to establish that she engaged in or failed to report wrongdoing.

A high number of changed answers doesn’t always indicate cheating, said Warren Fortson, who represents Snowden. Teachers told students to mark out answers they knew were wrong — using the process of elimination to arrive at the right answer — and then erase those marks later, he said. Some teachers handed out extra pencils to students who needed more erasers.

“For a lot of these people, the only thing they have against them is the statistical data,” Fortson said. “It wasn’t just teaching to the test. It was teaching how to take tests.”

But one of the state’s lead investigators in the probe of suspected test cheating, former Georgia Attorney General Michael Bowers, said erasure rates provide substantial evidence of cheating, even in cases where no witnesses come forward.

“Should a principal be held responsible based upon statistical evidence that shows strong evidence of cheating at their schools? Absolutely,” Bowers said.

About 150 of the educators named in the investigation have either resigned, retired or lost their appeals to keep their jobs. Thirty-five of them face criminal charges related to alleged cheating. None of the four ex-principals reinstated in APS jobs is among those charged.

At two of the schools that these principals led, F.L. Stanton and Capitol View, an erasure analysis cited in the state investigative report indicated that more than 70 percent of classrooms in each school were flagged for high numbers of changed answers.

The state investigation’s report said witnesses told investigators that Barber stayed late at F.L. Stanton during the week of standardized testing in 2009. She refused to answer any of the investigators’ questions. The school district’s settlement with Barber said it couldn’t introduce evidence establishing that she engaged in wrongdoing.

Barber was rehired Feb. 8 as an interim administrator on special assignment in the school district’s east region office, with a $110,954 salary, comparable to what APS principals earn.

At Capitol View, the report said Snowden denied cheating and told state investigators that erasures could be explained by teaching strategies, such as asking students to recheck their work. She was rehired March 1 as an interim administrator on special assignment in the school district’s north region office, with a $107,283 salary.

At Fickett, parents sent a petition with 325 signatures to the school board last year asking that Dorsey be reinstated because of his track record of success at the school and because there wasn’t evidence that he cheated. Two teachers confessed to cheating, and the state investigation’s report said Dorsey “created an environment that encouraged cheating” by informing teachers of the exact number of students that needed to meet expectations in order to achieve testing goals.

The school district rehired Dorsey on Jan. 16 as an interim administrator responsible for federal grants and programs, with a salary of $99,918. He made $102,001 as a principal in the 2008-2009 school year, according to his personnel records.

Romeka White, co-president of the Fickett Elementary PTA, said she doesn’t believe cheating occurred at Fickett.

“I’m sad he wasn’t able to come back to our school, but I’m glad they determined he wasn’t the issue. If your personnel is good, you want to keep those people,” White said.

At West Manor, the report said erasure data showed a “strong indication of cheating on a broad scale,” and Twyman was accused of failing to properly monitor testing. But no confessions or witnesses were mentioned in the state report.

Twyman was rehired as a principal mentor starting Sept. 24, 2012, with a salary of $101,435. Her salary as a principal was $94,306 in the 2011-2012 school year, according to her employment contract.

“I am excited about being cleared and I am ready to move forward with helping all children of Atlanta Public Schools,” Twyman wrote in an email to the AJC last year.