COLUMBUS — When Columbus City Schools Superintendent Gene Harris took over in July, 2001, about half the district’s students graduated from high school.
The rate now is 85.2 percent, using the same method to calculate the rate as then. Even under a new method implemented this year, it has climbed to 75.8 percent.
“I would describe that, if it were an accurate representation, as miraculous,” said C. Todd Jones, a former deputy assistant secretary of the U.S. Department of Education who now is a member of the state board of education. “That would be a miracle if it were true.”
But is it?
A Columbus schools internal audit released on Thursday calls the achievement into question and recommends that the state auditor investigate the graduation rate’s accuracy.
Ms. Harris said that such a probe would be premature, and that a committee headed by her deputy superintendent, John Stanford, can effectively monitor whether schools are accurately reporting graduation data.
The internal auditor found that some reports were inaccurate.
District officials were changing student grades, changing attendance data so that students could meet the requirement for a diploma — sometimes years after the absence occurred — and failing to properly code dropouts in the computer system.
“We were seeing students who to me looked like they had quit attending school, possibly in the 2008-2009 school year, but [the district] had not classified them under a dropout code,” district internal auditor Carolyn Smith said. “There was some cleaning up of that [data] this year. But they didn’t actually get removed — coded as a dropout — until 2012.”
In other words, students dropped out, but no one told the state department of education.
In 2010, The Dispatch reported that the district’s dropout numbers had declined dramatically.
Columbus reported only 20 eighth-grade dropouts, down from 481 just two years before. In a typical year, 2,000 students had been dropping out, but in 2010, it was just 500.
Education officials wanted to know: What was Ms. Harris’ secret?
“Those are great numbers,” Peter Cunningham, an assistant secretary of the U.S. Department of Education, said at the time. “If anyone has cracked the code on dropouts, they need to tell us and tell their peers across the country. We’d love to know what Columbus is doing right.”
One researcher who studies dropouts said at the time that sudden declines such as the one Columbus reported “would be pretty unusual” and highly improbable.
“I can’t put my finger on it, but I have to suspect something is going on with the data system,” said Chris Swanson, vice president of research and development at the EPE Research Center, the research arm of the publisher of Education Week. “That is a big drop for a single year, especially if things have been fairly steady up until that point.”
The internal auditor’s report released on Thursday does point a finger.
Wherever the internal auditor looked, she found manipulation of student data.
Administrators routinely withdrew and then re-enrolled students who had never really left school. That “broke” their enrollment, erased their absences, and kept their low test scores out of schools’ state report-card tallies.
Ms. Smith interviewed principals who said that administrators at the district’s data center had specified that they should withdraw students who scored poorly on tests.
She found that the district broke state and district rules on when and how to withdraw students.
She found high-school students who had been temporarily demoted from 10th to ninth grade, thereby kicking their Ohio Graduation Tests out of the state tally.
And then there were the unreported dropouts, changed attendance data, and changed grades that could affect the graduation rate.
If the internal auditor’s suspicions are correct, “It would mean that graduation rates have not been improving as claimed,” Mr. Jones said.
State Auditor Dave Yost’s office said that he hasn’t determined whether to examine the district’s graduation rate.