It was a heart-warming story: A school where nine in 10 students were poor enough to receive a free or reduced lunch, and yet where nine in 10 students met or exceeded most state testing standards.
As recently as 2009, Atlanta’s East Lake Elementary School was honored as a “No Excuses” school and deemed not only to be making the critical Adequate Yearly Progress, but to be doing so in “distinguished” fashion.
Then came the state’s analysis of suspicious wrong-to-right erasure marks on test answer sheets, including red flags for 42 percent of East Lake Elementary’s classrooms. Tighter monitoring was on order during the 2010 Criterion-Referenced Competency Test, and the results, as reflected in the school’s test scores, were devastating.
Of 15 CRCT exams (three subjects apiece for five grades), scores fell from the 2009 levels in 13. In the third and fourth grades, they fell by double digits across all three measures — reading, math and English/language arts — including a 30-percentage point drop for third-grade reading.
Six months after the erasure analysis was unveiled publicly, and after a subsequent investigation was completed, we still have no idea. And we still have no idea in large part because, despite this statistically glaring reversal of fortunes, those investigating possible cheating in Atlanta Public Schools didn’t deem East Lake Elementary suspicious enough to warrant their attention.
They interviewed just three of the school’s 39 employees, and none of its students or their parents. A separate analysis identified 12 other schools as even worse, so they spent little time or energy trying to find out what happened at East Lake and 45 other schools where adults appear to have cheated to make their students — and thus themselves — look good.
The steep drop-off at East Lake was also seen at many of the other 45 schools: At Cascade Elementary School, scores fell in all 15 tests from 2009 to 2010, by an average of 9 percentage points; at Fickett and D.H. Stanton, there were drops in all but two of the 15 tests; at Boyd and Dobbs and Heritage, all but four. And so on.
The 12 more-scrutinized schools truly were worse than the others. But a rigorous and statistically conservative — very conservative — analysis had already winnowed the 58 Atlanta schools from more than 1,700 public elementary and middle schools across Georgia.
These 58, and 43 in particular, were already considered the worst of the worst in terms of suspected cheating. The point of the investigation was not to find and focus on the worst of the worst of the worst.
And the process of paring the list to just 12 was inherently flawed: It indexed the 58 only in relation to one another rather than to schools with answer sheets that didn’t have loads of suspicious erasures.
The state and the analysts hired by the investigating panel are in a he said, she said situation about whether access to more sheets was ever requested. But it hardly matters — only 58 schools’ tests were examined, and that very plainly makes an inadequate basis for essentially clearing most of them right off the bat.
Yet that’s what the investigators did, and one wonders why.
One wonders if the point of the exercise became protecting the brand of Atlanta Public Schools by limiting the fallout. One wonders whose interest that really serves. But not whose interest it doesn’t serve: that of APS kids.