By most accounts, Atlanta schools Superintendent Beverly Hall runs a very tight ship. She demands results and she produces results.
However, results of a different type now threaten to undermine much of what Hall has accomplished. Strong evidence of widespread, perhaps systematic cheating on state-mandated tests within Atlanta public schools now calls into question not just the effectiveness of the Atlanta system but its basic integrity.
How Hall responds to that evidence will determine both her legacy and her continued ability to lead the district.
Sparked by an earlier AJC investigation that found evidence of cheating on state tests, Gov. Sonny Perdue ordered a probe of all state-mandated tests taken last year by Georgia students in grades 1-8. Every single test was inspected for evidence that wrong answers had been erased and changed to correct answers.
Of course, students often change answers in the course of taking a test. The state’s analysis found that on average, a student might change one or two answers on each test from wrong to right.
However, according to an erasure analysis conducted by CTB-McGraw Hill, the vendor that provides Georgia’s statewide tests, results at several hundred Georgia schools showed evidence of erasures well beyond what you would ordinarily expect. The company estimated the odds of such excessive erasures occurring naturally at one in a thousand.
The analysis also found that at 74 schools statewide, including 43 in Atlanta alone, more than a quarter of the classrooms tested showed evidence of erasures well beyond the ordinary.
At Parks Middle School in Atlanta, for example, almost 90 percent of classrooms tested showed evidence of an abnormal number of answers being changed from wrong to right.
State officials have stressed that the statistical evidence does not constitute proof of cheating in any particular classroom, a point that Hall repeated in a telephone interview Thursday. Smaller class sizes in Atlanta public schools also increase the odds that test results of a particular classroom might have been skewed.
Those are important caveats. Unfortunately, at best they can only mitigate the overall findings. They cannot explain the sheer scale of anomalies found at Atlanta public schools. They cannot explain, to cite just one example, how an average of 27 of 70 answers in one fourth-grade math class were changed from wrong to right.
The evidence that something has gone seriously wrong seems inescapable.
And as a strong supporter of public education, and as a father of two children who thrived in the Atlanta public schools, I do not come to that conclusion lightly.
In the interview Thursday, Hall reiterated her belief that poor, urban students are not fated to fail.
The considerable progress shown in Atlanta public schools over a decade of her leadership is often cited as proof of that fact, and she expressed sincere concern that the accomplishments of Atlanta students might now be tainted.
On the other hand, “cheating is never acceptable,” Hall said, pledging “an independent review of every classroom, every teacher and every principal” where problems might exist
Nationwide, the growing emphasis on standardized testing as a means of holding teachers and principals accountable is controversial. Hall has embraced that approach with a passion, using an intensely data-driven approach to demand measurable improvement from principals and teachers alike.
But as a consequence of that high-pressure environment, it now seems almost certain that some employees turned to cheating to produce results they could not achieve by legitimate means.
Until now, the Atlanta Board of Education has given Hall the freedom and support that any good superintendent needs to restructure a stubborn bureaucracy. But to protect both Hall and the district, the board now needs to take a strong leadership role in ensuring an independent, aggressive investigation of these allegations.
This is not merely a case of cheating the system, of misleading bureaucrats. The biggest victims are the students themselves. Their parents were reassured that their children were performing adequately and did not need additional help.
In too many cases, it appears, that simply wasn’t true.