ATLANTA — Thousands of Atlanta Public School students affected by a cheating scandal that led to criminal charges against educators have fallen behind in reading and English language arts compared to classmates whose tests showed little or no evidence of manipulation.
That's according to a new study by Georgia State University researchers released this week. It's believed to be the first look at how students were impacted by one of the largest standardized cheating scandals. The district requested the analysis shortly after Superintendent Meria Carstarphen came to Atlanta in 2014.
But researchers and district officials acknowledge it's difficult to determine whether the cheating is the sole cause of any individual student's struggles.
Researchers focused on 7,064 students with high numbers of erasure marks on a 2009 statewide exam — indicating wrong answers changed to correct responses and potential interference. District officials said about 3,700 of those students remained enrolled as of fall 2014, including about 1,800 who had the highest number of erasure marks indicating more interference.
The data indicated that teachers or administrators who changed the answers focused on students not expected to do well. It also showed the test-score cheating disproportionately affected black students.
The district's students are 75 percent black, but researchers found 98 percent of students with 10 or more erasure marks were black.
Georgia State economics professor and lead author Tim Sass said the analysis controlled for student demographics and performance on a statewide exam after cheating had been revealed. The district also noted in its own summary that the study doesn't factor in teacher experience, quality or other "culture variables" that could have affected students.
The report found mixed effects on students' performance in math and little or no evidence that student attendance or behavior in class worsened for students in that category.
But their performance in reading and English language was one-fourth to one-half behind the average achievement in those subjects for middle school students nationwide. Sass compared that to students in a first-year teacher's classroom versus students being taught by someone with five or more years' experience.
In the past, district officials instituted broader programs aimed at any student struggling or at risk of dropping out. Using the Georgia State study as a guideline, officials now want to review whether those programs have helped students affected by cheating and consider new, individualized plans where needed, said Bill Caritj, chief accountability and information officer hired to the new position after Carstarphen became superintendent.
"There is no silver lining I can see in this," Caritj said. "I think if you ask teachers and administrators, every single one of them would say when you cheat children there is no silver lining."
Richard Quartarone is a member of a parents' education advocacy group called Southeast Atlanta Communities for Schools. He said the district should get credit for doing the study but thinks it's difficult for any researcher to sum up deeper issues affecting Atlanta students, including poverty.
"I think it's difficult to understand the psychological impact to the kids, the community, the city," he said.
A state investigation found cheating began as far back as 2005, including educators who pointed students toward correct answers during testing or erased and changes students' testing sheets after they were turned in. Investigators said they found 44 schools and nearly 180 educators were involved.
In 2013, 35 educators were indicted on charges including racketeering, making false statements and theft. Many pleaded guilty, but 12 went to trial. A jury convicted 11 of those former educators of racketeering in April.
The cheating came to light after The Atlanta Journal-Constitution reported that some scores were statistically improbable.
Former Superintendent Beverly Hall was among those charged and insisted she was innocent. Hall died in March of breast cancer, never going to trial after arguing the disease made her too sick to help with her defense.