GothamSchools — daily independent reporting on NYC public schools

For the second day in a row, the city’s comptroller has released an audit questioning the validity of the city’s education data. And for the second day in a row, political jockeying initially overshadowed the report’s content.

At a press conference this morning, Comptroller Bill Thompson, who is running for mayor, said the audit of testing oversight revealed that the Department of Education had allowed “an environment ripe for cheating.” ”We found that the Department of Education has engaged in sloppy and unprofessional practices that encourage cheating and data manipulation,” he said.

But the report did not find new instances of cheating.

The audit focuses on the role played by testing monitors in overseeing the math and English Language Arts, or ELA, tests given to elementary school students in 2008. These monitors, employed by the DOE, make unannounced visits to schools on testing days to ensure that protocols are being followed. Thompson’s audit deems the monitoring system “inadequate.”

The report suggests that the DOE is not thoroughly monitoring its monitors. “DOE does not keep track of the monitors assigned to visit schools and the submission of checklists,” the audit states. It adds that, in many instances, monitors respond to checklist questions that they couldn’t possibly know the answers to — citing a case of a monitor who arrived after the test had begun, but marked “yes” that the tests had been safely stored.

The DOE today shot back, saying that the comptroller’s auditors had not actually witnessed the process they were supposed to be investigating. According to the DOE, of 17 tests the auditors witnessed, none was being monitored.

“They didn’t follow all the procedures because they were not at the schools as monitors; they were there to assist the Comptroller’s auditors,” said spokesman David Cantor in a statement. (Cantor had been ejected from the comptroller’s press conference.) “The Comptroller also recommends several improvements to the monitoring process that the DOE already implemented in time for the 2009 tests, which he could have seen for himself if he hadn’t declined the DOE’s invitations to do so.”

The audit also highlights the fact that the city no longer uses a particular strategy to detect cheating: erasure analysis, which counts the number of times each student erased a wrong answer and bubbled in a correct one. “The old Board of Education abandoned this practice in 2001 because they determined it was a waste of money,” Cantor said.

But Thompson said erasure analysis should be reinstated. The education department said today it would take Thompson’s suggestion “under consideration.” Currently, the DOE only does a complete review of test scores if it receives allegations of improprieties.

In his audit, Thompson’s auditors performed erasure analysis on a sampling of tests, finding two fourth-grade students whose tests had “an excessive pattern of erasures.” The DOE’s statement said that the audit’s concerns over the two tests were “baseless,” and that the two students had accidentally filled in the wrong row of bubbles.

The department found 13 instances of cheating between 2006 and 2008, according to the audit.