Learning Curve: Erasing trust in APS

For example, the state audit found evidence of answer-sheet tampering in 90 percent of Parks Middle School classrooms. Has APS ever followed Parks students through to high schools to compare scores?

Responding to evidence of cheating uncovered by an AJC investigation in 2008, the state reviewed every 2009 answer sheet to measure how often kids changed wrong answers to right by virtue of erasures on the sheets. Because every test sheet was checked, the state was able to develop a reliable index of how often test answers were changed from wrong to right and flag schools that had inordinate occurrences of answer changes, right down to the classroom level.

The findings suggest that one in five Georgia public schools may have tampered with student answers. But the worst incidences occurred in Atlanta, where the data suggest that cheating went on in 37 of 55 elementary schools.

Concerns over APS scores are not new. In 2001, the AJC reported that 30 of 68 Atlanta elementary schools had gains of 30 or more percentage points in one or more CRCT subjects. Ten of them had gains of 40 or more points. Nine years ago, Atlanta school officials credited those impressive leaps on its school reforms and better classroom strategies, a rationale offered now by Hall in defense of the 2009 scores.

While committing to a vigorous review of the erasures by independent experts, Hall said, “I also have to say for the record that hard as it is for some people to believe, when children are held to high standards, when teachers use appropriate strategies, when there are high expectations, a lot of support, intervention strategies, children do improve and some of them improve dramatically.”

But research shows that very dramatic improvements are unusual. In 2001, even APS board member Jean Dodd expressed doubts, saying, “Over a period of 30 years, I taught every grade of elementary school. I had just not ever seen scores like that before.”

One of the schools that showed remarkable improvement in 2001 was Dunbar Elementary. In the state’s 2009 audit, nearly 70 percent of Dunbar’s classes had test erasures that far exceeded the state average, the 10th worst rate in the state.

At Atlanta’s Gideons Elementary, an average of 27 of 70 answers on each fourth-grader’s math test were changed from wrong to right in one classroom. Yet, the state report shows that Georgia fourth-graders on average each only erased and changed 1.87 answers from wrong to right.

Hall admitted such data seems damning, but urges restraint. “Clearly, we have already been tried in the court of public opinion,” said Hall. “But I really look at the data. You are looking at the summary data. On first blush, it indicates a major problem. But I know enough about data analysis to know that is not certain. We are going to have to look class by class.”

If that examination confirms a culture of cheating, Atlanta schools will need a shake-up that bounces desks off the floor and rattles posters off the walls. Such educational malpractice not only hurts the children, but victimizes the next teacher in the chain who can’t understand why a student who scored proficient in reading the year before now can’t sound out a sentence.

Under federal laws, students may have qualified for tutoring based on their actual scores, says Kathleen Mathers of the Governor’s Office of Student Achievement. “Students were deprived of those opportunities.”

Last week provided a chilling glimpse into how easily cheating occurs. In affidavits admitting that they cheated on the 2008 CRCT, the principal and assistant principal of DeKalb’s Atherton Elementary School recounted glancing at the summer retests of their students and realizing that their school would not make the all-important Adequate Yearly Progress.

So, Principal James Berry pulled out a pencil, told his assistant principal, Doretha Alexander, to call out the answers and began to erase the students’ answers. Not only did all 32 students pass the retest, 26 scored at the highest level.