In a meeting with AJC editors and reporters, Atlanta Superintendent Beverly Hall pledged cooperation with a new state probe. Still, she said she’s unsure cheating occurred.
Q: Going back over the last year and a half, starting with reports in The Atlanta Journal-Constitution about unusual gains at schools like Deerwood Academy, your spokesmen said you stood by the numbers. Do you have any regrets?
A: I did not say I didn't think there was a problem. What I said was, we wanted to go in and do a detailed review. When [the state] presented the evidence, we felt let's go back to Deerwood and see exactly what happened before we call it cheating. Our deep investigation showed there were lots of processes and protocols that weren't followed. In the summer administration of the test, people were very lax. We would take disciplinary action but we did not believe the principal at the time was implicated.
Q: You've had a number of warnings over the years that there were anomalies, going back to 2001, when the CRCT scores in 30 schools jumped at least 30 points on average, 10 schools jumped 40 points on average. The district said at the time there was no need for further analysis. Why has there been this resistance to look deeper?
A: I'm not sure it was resistance. We felt if it was an anomaly, that eventually scores would begin to fluctuate up and down. In organizations like schools, and I have run several, people change, teachers come and go, principals come and go — 90 percent of the principals who were there that first year are no longer even in APS today. At some point, if this was something that was a conspiracy to cheat, someone would be telling. You can't keep a secret with 3,600 teachers, I don't think.
The truth of the matter is, the scores held. I asked to participate in the NAEP [National Assessment of Educational Progress] — this is an independently monitored test. They look at the class rosters, they select the students to be tested, they walk in with the test booklets — two students do not have the same booklets. They collect the booklets and they walk out. So if indeed what was happening on the state tests was the result of people cheating, it would very much show up on the NAEP. We would not see progress. We would begin to see the opposite. Lo and behold, the NAEP has been on a steady trajectory, moving in the same direction as the state tests. So that was really what gave me added confidence that the gains we were seeing were real.
Q: But on the NAEP, the roster comes from the school, so students could be removed beforehand.
A: We have to send the roster [to them]. They then match the roster we send to the state. If they see anything that appears to be a discrepancy, they contact us. We then have to resubmit it. They verify that children on the register are sent to them.
Q: The other issue is that the gains have leveled off since 2007 — Atlanta is still lagging behind Georgia, other urban districts, other large cities. The way you are depicting it doesn't show there is still a gap; there's a huge gap, 50-60 points, between black students and white students in Atlanta.
A: I think what the NAEP shows is, if we continue to improve at the rate at which we're improving compared to other urban systems, we will eventually close the gap with both the state and the nation. When I came to Atlanta, these scores were very, very low. We were taking a risk, we knew we could be way below basic. But we felt we had to have another way of assessing how the system is doing compared to systems nationwide.
The CRCT shows a parallel. In 2010 in eighth grade, with [state] monitors in the room, they went up 2 percentage points in reading — 88 percent meeting or exceeding standards to 90. The challenge for us is to have [students] improve at a rate that eventually allows that gap to be totally closed.
Q: [The state flagged 58 schools as needing further investigation; Atlanta's investigation identified 12 of those as rating serious concern.] You agree 12 schools are a problem. How do you feel about the other schools — are they cleared with question marks?
A: Legally, we need to look at those schools very carefully as well. There are two top priorities: One to hold accountable anyone who cheated, because they've done a terrible disservice to the children and Atlanta Public Schools, and that is a disgrace. The second is to make sure any child who could've been involved in this will receive whatever they need to get them on track.
Q: Earlier you talked about your sense there were no conspiracies, but is there another scenario where it is not a conspiracy, and it's not necessarily just individual actors, but a system that does not adequately police itself?
A: One of the things Caveon [Test Security, the consultant hired to investigate Atlanta schools,] said very clearly in its report, when they compared our security processes and protocols and activities around securing the tests, that we were better than just about every other district they had looked at. They found that there were discrepancies in the following of the processes.
I think you have individuals at certain schools who did not go to training or they went to training but we don’t have their signatures, so it’s suspect. As a system, we have a sufficient amount of training, notices, warnings, etc. [This year] we actually had more monitors than [the state] required.
Q: Your sense of whatever issues are there had to do with some bad actors, that it wasn't systemic?
A: That's what I'm saying until proven otherwise. I'm open to the investigation going in. The state investigator has subpoena power.
Q: But how could such similar actions be taken by so many people at so many schools and there not be any coordination?
A: Can you tell me what coordination?
Q: Well, 256,000 erasures that were detected in the erasure analysis done by the state?
A: I heard Kathleen Mathers [executive director of the Governor's Office of Student Achievement] say yesterday that erasure analysis does not mean cheating. We had the best expert we could find. I don't pretend to know. I don't have a clue when it comes to erasure analysis. We have asked for access to everything to allow us to determine what those erasures mean. I submit to you no other district involved with this hired one of the firms the governor listed as the top forensic audit firms, KPMG, to go in and do a forensic audit. We said do anything you want. We will give you anything you need. And we did.
Q: Do you think interviewing 12 students and parents [for the investigation of 58 schools] was sufficient?
A: They [KPMG] were the experts. Far be it from me to tell them how to do their work. I was not involved with that investigation. We turned it over to people that I believed were, until they came back with what they did, the best in the business. Now, let the state investigator come in. I'm sure he or she is going to do a very thorough job, and wherever that leads us, we're going to go there.
Q: Will you allow the state investigator to speak to APS employees without an APS official?
Q: That was obviously what was happening with KPMG — there were district officials present for most of those interviews.
A: I sent a letter out last week informing every employee that they are to cooperate fully with any investigation.
Q: Consider a different scenario, you came in and said: get on the bus or get off the bus. A number of principals are not here today who were here 10 years ago. Is it possible that that created an environment were a principal was so concerned about keeping their job that they felt they had to go in and massage the numbers to make their school look better?
A: It's always possible that somebody would need to cheat. But the vast majority, I personally believe, have high moral standards.
Credit: Jason Getz / Jason.Getz@ajc.com