Atlanta teachers and principals for years methodically altered answer sheets for students taking state tests, boosting scores and transforming struggling schools—and the district as a whole—into what appeared to be a spectacular urban success story, a unveiled this week by the Georgia governor’s office says.
“Thousands of schoolchildren were harmed by widespread cheating in the Atlanta Public School System,” the report says. “Many of the accolades, and much of the praise, received by APS over the last decade were ill-gotten.”
The investigative team drew on more than 2,100 interviews and examined more than 800,000 documents as part of its nearly yearlong probe into improbably high test scores across the 48,000-student district.
Apparently one of the largest cases of alleged cheating on state exams to date, the scandal shines a harsh spotlight on the much-lauded district and is expected to fan debate over the validity and effectiveness of high-stakes testing-and-accountability systems.
The findings also bring a disturbing coda to the career of former Superintendent Beverly L. Hall, who retired in June after 12 years in that job, and was at the helm of the district when the cheating was alleged to have occurred.
The report says that Ms. Hall and her top staff must have known that cheating was taking place, because of the unusually high gains and the word of whistleblowers. Ms. Hall has not responded directly to allegations, but through a statement from her lawyer, she has categorically denied any intentional wrongdoing.
In 2009, Ms. Hall was honored as the national superintendent of the year by the American Association of School Administrators. The Council of the Great City 69´«Ă˝ singled out Ms. Hall for her contributions to urban education, and she is the first school administrator to receive the Distinguished Public Service Award from the American Educational Research Association.
“I have had nothing but the highest regard and respect for Beverly, so I personally find this devastating,” said Daniel A. Domenech, the AASAâ€s executive director. “If it’s true, it’s incredible. If it’s not true, what a shame that such a dedicated educator would wind up her days in this fashion.”
â€No Shortcuts to Success’
U.S. Secretary of Education Arne Duncan, who has praised Atlanta for its reform efforts, said in a statement to Education Week that “this unfortunate incident highlights the need for transparency and accountability throughout our education system.”
Mr. Duncan added: “There are no shortcuts to success, and there are schools and districts across Georgia and across the country that are facing the same expectation to perform that are making genuine progress without cheating.”
The Atlanta school board that it was extending the contract of interim Superintendent Erroll B. Davis Jr. for a year, to provide stability in the wake of the scandal. Prior to joining the district, Mr. Davis was the chancellor of the University System of Georgia. The district also plans to institute annual ethics training for employees and set â€trigger points’ that would result in investigations of schools whose test scores increase by a larger than expected percentage.
The 800-page report says cheating occurred at 44 of the 56 schools that investigators examined. The investigators focused on schools where erasure analysis on test sheets showed that it was virtually impossible for so many wrong answers to be changed to correct responses without some kind of outside intervention.
The report also says students who would have been due remediation because of low scores missed getting the help they needed.
But the most sensational parts of the document are the narratives that recount, in the words of teachers and principals, how cheating allegedly was carried out in individual schools.
At times, those sections read like a spy novel. One middle school teacher said he took photos of the unsuspecting test coordinator’s office, so that he could alter test materials and return them, leaving the office looking undisturbed.
A testing coordinator at one elementary school who admitted to test tampering said that she witnessed the principal altering test sheets while wearing gloves, for fear of leaving fingerprints.
At another elementary school, veteran teachers told investigators they doctored the answer sheets of students taught by newer teachers until they could trust those teachers to know about the cheating scheme. One year, a group of the school’s teachers took answer sheets to the home of another teacher and held a weekend “changing party,” the report says.
The investigators accuse Ms. Hall and her top staff of having “created a culture of fear, intimidation, and retaliation,” in which witnesses feared, with good reason, that they would face punishment for speaking up about the cheating. Teachers claim they were told that if they didn’t meet certain testing targets, they would face penalties.
Some of the employees interviewed described humiliations that went beyond bad performance reviews. The report said one elementary school principal forced a teacher to crawl under a table in a faculty meeting because that teacher’s students had low test scores.
“Dr. Hall failed to balance the data-driven environment she created with an equal focus on the importance of integrity in achieving these goals,” the report says.
Richard H. Deane Jr., Ms. Hall’s lawyer, said that the report’s conclusion that the former superintendent must or should have known about the test tampering is based entirely on supposition and negative inferences “from selective, circumstantial evidence.”
A string of other districts have undergone accusations of test-tampering, including recent allegations in the District of Columbia, Baltimore, and Philadelphia. And news surfaced this week that the federal Education Department had joined in the investigation into cheating in District of Columbia schools. But the widespread allegations in Atlanta—82 staff members told investigators they had altered tests, and investigators believe their report implicates 178 in all—makes this case perhaps the largest.
Results Brought Praise
“It’s by far the biggest, most prominent, most widespread instance [of cheating] I’ve seen,” said Gregory J. Cizek, a professor of educational measurement and evaluation at the University of North Carolina at Chapel Hill, who assisted the state investigators. The state’s vigorous probe is also noteworthy, Mr. Cizek said.
The test-tampering on the state’s Criterion-Referenced Competency Tests, or CRCT, did give the appearance of impressively high scores in the district, the report says.
For example, at one elementary school, the percentage of students “exceeding expectations” in English/language arts rose from 28 percent in 2006 to 79 percent in 2007. Another posted a 42-percentage-point gain in the number of students exceeding state standards in math from 2004 to 2005.
District officials attributed those eye-popping improvements to comprehensive districtwide reform efforts, and the high achievement brought national recognition as well as funding from philanthropies such as the Annie E. Casey Foundation, the Bill & Melinda Gates Foundation, and the GE Foundation. (The Gates Foundation was a recent funder of Education Week’s parent nonprofit organization, and the GE Foundation currently helps support the newspaper’s coverage of the STEM subjects.)
Atlanta schools have also shown gains on the National Assessment of Educational Progress. The state investigation did not suggest that those tests, which are administered by NAEP staff members to a random sample of students, had been compromised.
As allegations of test-tampering mounted, led by stories in the Atlanta Journal-Constitution, the district responded with its own of the 2009 state-test administration. It found some problems, but no cheating admissions or signs of a centrally coordinated effort to manipulate scores.
Then-Gov. Sonny Perdue called that report “woefully inadequate” and launched his own investigation, which was continued by the current governor, Nathan Deal. Both are Republicans.
Improving Test Security
UNC’s Mr. Cizek said state investigators looked at the average number of wrong-to-right erasures on tests statewide, then compared them with Atlanta’s results school by school.
On a 40-question test, only one or two wrong answers on average should be erased and then marked correctly, Mr. Cizek said. In contrast, there were some entire classes in the district where an average of 15 wrong answers were erased and the correct answers filled in—a difference nearly impossible to attribute to chance, Mr. Cizek said.
The report about Atlanta comes amid national concerns about high-stakes testing. A committee of the National Research Council recently studied test-based incentive systems and found that approaches implemented so far have had little or no effect on student learning. (“Panel Finds Few Learning Benefits in High-Stakes Exams,”, June 8, 2011.)
People in all fields tend to respond very narrowly when performance is based on a single measure, said Michael Hout, the sociology chairman at the University of California, Berkeley, and the chairman of the NRC committee. Often, teachers start “teaching to the test,” he said.
He said that while the misdeeds in the Atlanta district are “Enron-like in scope,” the problem can be addressed by measuring school performance in multiple ways.
Mr. Cizek said districts and states should not abandon accountability in the wake of testing misdeeds, but they should get serious about test security. Current processes for spotting problems are “anemic,” he said.
“You have to assign this to an independent body that doesn’t have a stake in the results,” Mr. Cizek added, rather than the state department of education. “Assign the cop duty to someone else.”