States' data systems more complex and costly than thought

Driven largely by No Child Left Behind, state education agencies are developing student information systems that will, among other things, compile attendance rates, test scores, socioeconomic background, and other data that could be useful in improving the delivery of education and tracking results.

But gathering data from individual school districts and then building an electronic warehouse that everyone can access has proven to be a monumental task, as well as an expensive one. A recent New York Times article found some states shelling out millions of dollars to consultants. North Carolina, for example, is projected to spend upward of $250 million before its database is fully functional.

“I’d be fascinated to learn what on earth that money went toward,” says Chrys Dougherty, director of research for the National Center for Education Accountability (NCEA), a nonprofit collaborative of the Education Commission of the States, the University of Texas-Austin, and Just for the Kids, which is sponsored by NCEA.

In 2005, NCEA surveyed states to find out how they were doing with these data collection systems. Of the 49 states that responded, only eight had in place most of the elements NCEA considers essential to an effective student information system.

These components include enrollment and program participation figures, transcript information, and a system of identifying students that maintains their anonymity.

Jim Lair, cofounder and CEO of the Virginia-based Center for Data Quality, says it’s surprising how complex seemingly simple tasks like assigning an identifier to each student becomes when it’s attempted across a state.

“The issue is being able to know where all the pieces of the puzzle are, and in some cases, in disparate systems,” says Lair, whose company reviews client databases for accuracy and structural integrity. In the last several years, his firm has expanded its clientele into the education arena.

“The education system uses a serious amount of code values for different aspects of student life, courses, and curriculum that are hardly ever the same from state to state and even district to district,” he says, “Standardizing the data is a massive undertaking because there is no common meaning for things like ‘school.’ Is it a building, is it a series of buildings, is it a classroom?”

Adding to the problem is a tendency by some states to set their sights too high, Dougherty says. “They want a Taj Mahal so they overdesign and make it too complex.”

He advises states to start small and start basic. “You can eventually get to what you want without a lot of cost.”