Elissa Gootman’s story in the Times today on the non-functionality of the $80 million ARIS data warehouse system is important because it lays bare what teachers have known for months: ARIS was supposed to give parents and teachers radically more access to student achievement data, but in practice it suffered from frequent malfunctions and currently is providing zero information.
I’ve been talking to people familiar with the new version of ARIS, slated to be released in November. They tell me this new software is a huge improvement over the ARIS released last year. Even with new software about two weeks away from debut, there remain two major unanswered questions about the Department of Education’s effort to build a massive data warehouse.
The first question is whether the contractor, IBM, made mistakes that could have been avoided — and whether some portion of the taxpayer dollars slated to go to IBM, which total $80 million, should be paid back. The second question is whether a program like ARIS is necessary at all.
On the question of IBM’s role, the Department of Education’s position is that the difficulties ARIS faced in its first edition last year were unavoidable and to be expected of any new data system. “There’s always going to be some kind of issues,” Andrew Jacob, a DOE spokesman, just told me over the phone.
Indeed, software systems are notoriously and frustratingly bug-ridden and slow to launch. I can say that with authority now, working amongst techies here at The Open Planning Project.
But school officials early on promised more than ARIS delivered. At a long press conference last year in the basement of Tweed Courthouse, Schools Chancellor Joel Klein talked about the software as transformational and said that eventually not just teachers but also parents would have access to the system. So far, parents have no access at all*, and the software has been so disappointing that all of the schools I’ve spoken to about it have given up on ARIS. (The DOE insists some principals and educators did use the software last year.)
The glitches appeared soon after the software’s release. GothamSchools’ own Kelly Vaughan tells me ARIS crashed even during a presentation introducing it at her school, Mott Hall III, last year. Those who could log onto the program said that, even when accessible, ARIS was disappointing.
John Galvin, the assistant principal at IS 318 in Williamsburg, said he was excited by the idea of a warehouse to aggregate information from the DOE’s many databases. He also liked the idea of software to help him zero in on the precise areas where students needed to improve. But Galvin said ARIS quickly disappointed him. “You really at some point were supposed to be able to access individual kids’ scores, compare classes to one another,” he said. “You could never do any of that stuff. You could look at an individual kid’s scores, but you couldn’t manipulate the data in any kind of sophisticated way.”
The second question the Times story raises is whether ARIS should exist at all. There is already a fairly well-organized campaign arguing that it should not, especially amid a budget crunch that could become a budget crisis. A group led by the Center for Immigrant Families and Time Out From Testing has written a letter asking the city to axe the ARIS contract, and their push is gaining steam: last week, the United Federation of Teachers signed on. This group challenges not only the IBM contract, but the notion that the city’s schools need capacity to track data beyond what the state already provides.
Alongside this campaign is a substantial underground of educators who have responded to ARIS’s malfunctions by creating homemade versions of the same kind of software. This do-it-yourself software is usually made by a single tech-savvy teacher who has a good handle on Microsoft Excel or Google’s online spreadsheet tool. The teacher will compile the schools’ data, make up some process for other teachers to input their students’ scores, and then write their own formulas to aid analysis.
Galvin said he uses a spreadsheet he built with Microsoft Access to examine students’ test scores. He said the software helped his data inquiry team last year move 25 of 26 students they zeroed in on as under-performing in math move up a proficiency level on the state test. The Times article talks about a program called DataCation that one city school built and 20 others have purchased for about $13,000.
I tried to reach the Department of Education’s chief accountability officer, Jim Liebman, for comment this morning, but a spokesman, Andrew Jacob, said he was not available.
Jacob said that ARIS’s launch was hobbled by two main challenges: privacy concerns and the difficulty of gathering together the information from about five data systems that were built in the 1970s. He said IBM had only eight months to develop the “core” of ARIS, and he insisted that many principals and data inquiry teams used the software last year.
Responding to Galvin’s concern about the capabilities ARIS offers, Jacob said the new software coming next month will be more sophisticated and will respond to concerns raised by principals last year.
*CLARIFICATION: Delaying ARIS’s release to parents until this school year was always the Department of Education’s plan, Jacob said.