In responding to what appears to be epic levels of cheating at her schools. Dr. Beverly Hall has to answer this question: Did she allow principals to present her with miracles that she failed to question hard enough?
Did APS ever look to see whether students and classes with sudden, unexpected surges in their CRCT scores maintained those levels of proficiency in high school?
For example, the state audit found compelling 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?
Consider this post from someone who says they are an APS high school principal: (Whether he/she is a principal is not critical; it’s the comment that I think is credible and worthy of exploration.) (Also, here is a new map of the possible cheating sites statewide.)
CAN’T UNDERSTAND HOW PARKS MIDDLE SCHOOL CAN BEAT OUT INMAN AND SUTTON OR EVEN BUNCH. LOOK AT THE FEEDER SCHOOL DATA AND THEN LOOK AT HIS DATA. THE FEEDER SCHOOLS ARE DOING POORLY AND THEN THEY GET TO PARKS AND BECOME SMART. THEN CHILDREN FROM PARKS COME TO MY SCHOOL WITH LEVEL THREE AND CAN’T READ.
Many of you have commented that you suspected cheating for many years and that complaints were made. Indeed, AJC reporter Paul Donsky wrote this story in 2001. Now, it seems prophetic. If only Dr. Hall had acted on this issue in 2001. Her reputation and that of her system may not have been on the line today. (Nor did the state respond to the allegations at the time, even though it could have sought an erasure analysis in 2001, which, by the way, is a minimal review and does not catch all instances of cheating.)
Of the 68 elementary schools tested last year, 30 had gains of 30 or more percentage points in one or more CRCT subjects. Ten of those had gains of 40 or more points. And Dobbs is one of 17 Atlanta Public Schools that did well enough to come off the 2000 failing list.
Atlanta school officials say new reform efforts and old-fashioned hard work by teachers and students helped push the district’s scores up.
But amid the cheering, there are some questions. A rise in scores is expected the second year any standardized test is given, as was the case with the CRCT last spring, because students and teachers become more familiar with the test. But sudden 40-, 50- and even 60-point spikes are not common, testing experts say.
“Either somebody is doing a terrific job at something . . . or there’s something inappropriate going on, ” said Gregory Cizek, associate professor of educational measurement and evaluation at the University of North Carolina.
Some of the numbers are astonishing:
> At Dunbar Elementary School near downtown Atlanta, three-quarters of fourth-graders passed the reading portion of the test, compared to about one-quarter last year.
> At M.A. Jones Elementary in west Atlanta near the Atlanta University Center, 88 percent of fourth-graders passed in math compared to 34 percent the year before, a 54-point increase.
> At Thomasville Heights Elementary School in southeast Atlanta, 73 percent of fourth-grade students passed in reading, compared to 19 percent the year before.
Atlanta does about-face
Atlanta’s results are notable because many schools that posted huge gains have student populations that are almost entirely minority and low-income, groups whose test scores historically have lagged far behind. At several schools that posted huge gains, including Thomasville Heights and Cook Elementary schools, nearly all students live in public housing.
Atlanta’s CRCT scores remain below the state average but are in line with neighboring systems in the metro area. However, Atlanta’s scores are rising much faster. For example, 72 percent of Atlanta fourth-grade students passed in English in 2001, a 15 percentage-point jump from the year before. In Fulton County, 80 percent passed in English, a 2-point jump. Clayton County posted a 71 percent pass rate, an 8-point jump. DeKalb County had a 70 percent pass rate, a 3-point jump.
For Atlanta Public School officials, the scores help validate reforms that have been put in place in recent years, such as an intensive reading program designed to boost literacy in early grades. Kathy Augustine, deputy superintendent for instruction, said the district also made sure teachers knew the topics covered on the CRCT. Principals checked regularly to see that those areas were covered in class.
Atlanta school officials say the rise in CRCT scores should not raise eyebrows, pointing out the district’s SAT scores jumped 16 points this year.
Still, Atlanta school officials say they double- and triple-checked the CRCT results, at times examining scores of individual students, to make sure the numbers added up.
That wasn’t enough for Atlanta school board member Jean Dodd, who stormed out of a school board meeting Sept. 17 after expressing concern about the validity of the scores. Her comments came during a portion of the meeting that was closed to the public. The meeting was held to decide the size of Superintendent Beverly Hall’s bonus, which is largely tied to test score results. She received a $47,520 bonus.
“Over a period of 30 years, I taught every grade of elementary school, ” Dodd said in an interview. “I had just not ever seen scores like that before, and so I just . . . made my concerns known.”
Third-party audit sought
Gary Henry, who has studied state testing for years and serves as director of Georgia State University’s Applied Research Center, said Atlanta’s results should be independently verified.
“There will be some folks celebrating Atlanta’s turnaround, and others shaking their heads at these results until you have a third party at arm’s length validate these results, ” said Henry.
The most likely organization to do that is the state Office of Education Accountability, formed last year to grade schools as part of the state’s education reform efforts. OEA Director Davis Nelson said he will be drafting a policy in the coming months to determine when to launch such investigations. Dramatic test score jumps could be one trigger of an inquiry.
Atlanta school officials defend the scores and say there’s no need for such an audit.
“We are proud and our students are proud of the scores, ” said spokeswoman Pat Bowers. “And we believe the results should be allowed to speak for themselves, particularly when reinforced by other test scores over the last year.”
Cheating and manipulation on standardized tests has occurred in other cities, including New Orleans, New York and Fairfield, Conn., where officials in 1996 discovered an unusually high number of wrong answers erased. Two Atlanta high schools were investigated and ultimately cleared last year after allegations arose of cheating on the high school graduation test.
Georgia State’s Henry said there are several possible explanations for Atlanta’s test scores.
“When you look at these numbers, ” he said, “the first reaction is that your jaw is going to drop.” However, he said a much higher percentage of last year’s fourth-graders were enrolled in pre-kindergarten programs than the previous year’s fourth-grade students, making them better prepared. Also, last year’s fourth-graders were also the first in APS to benefit from a new district reading program that targets early grades.
But, he added, “no single factor is likely to explain” the results.
(Also, here is a new map of the possible cheating statewide.)