Link Student and Teacher Performance? Not Without the Data

John Mooney | May 10, 2010
Gov. Chris Christie has ambitious plans to improve NJ's public education. But the data to meet his goals is proving difficult to obtain

Merit pay, tenure reform, greater teacher accountability: They are all part of Gov. Chris Christie’s promises for improving public education in New Jersey. But there’s something missing that could prevent any of those promises from coming true: data.

For the better part of a decade, New Jersey has sought—and often struggled—to complete a statewide student database that would allow educators and the public to track and analyze how students and teachers are doing.

Known as NJ SMART, the system for much of that time has stumbled over everything from lack of funding to the complex challenges of just labeling 1.3 million children with standard identification numbers.

In the last year, however, the system by all accounts has made important strides, with 10 years of test scores loaded and ready to gauge student and school-wide performance. But when it comes to some of the teacher evaluations that are now the focus of Christie’s agenda, the system is still not there. It could be years away from delivering the level of sophistication the plans require.

State Education Commissioner Bret Schundler said almost as much on Friday when he unveiled the state’s reform plans. He described “literally millions in hardware and software costs” needed to complete NJ SMART.

“It is not where it has to be, or where we want it to be,” he said. “But it’s moving forward.”

With only a small handful of states much ahead of New Jersey, the state is hardly alone in being slow to attain the requisite levels of database management.

But highly sophisticated technology is integral to the kinds of sweeping improvements that Schundler has announced, with many of them depending on the state’s ability to match up student performance with individual teachers and other staff.

As part of the state’s bid for $400 million in federal Race to the Top funding, Schundler described a system that would measure precisely whether an individual teacher’s students are learning and by how much. For example, he described an incoming class where only 20 percent of students were proficient readers. “If you are taking that class to where 60 percent, what that says is that you as an educator are incredibly effective,” Schundler said.

With those concrete gains identified and catalogued, he said, teachers could be rewarded with merit pay bonuses. Or the data could be used for tenure decisions as well, with Schundler proposing a teacher’s student performance be weighed before awarding the job protection.

He said such performance grades could even replace seniority in determining whether a teacher is promoted or laid off.

“Should it be about seniority or teacher effectiveness?” he said. “If you have to do a reduction in force, I argue you shouldn’t be letting the best teachers go.”

But what Schundler did not say is the extent to which the state’s current database system falls short of those capabilities.

Under the state’s initial application for Race to the Top funding last winter, for instance, New Jersey’s lowest marks came on the evaluation of its database capabilities.

Separately, a state-by-state review by the Data Quality Campaign, a Washington D.C.-based clearinghouse, found New Jersey actually had most of the data elements, but not the necessary tools to manage the data, including the ability to link teachers and students, as well as analyze and access that data.

It’s not easy work, and New Jersey has applied for $17 million under a separate federal program to build the data warehouse capable of making those connections.

But even if the state gets the money—it has been rejected for the grant three times before—the system demands countless hours of training and logistics to match up not only personnel and students, but also the coding for class schedules and curricula.

There’s little disagreement to the value of the data.

Nathan Parker, superintendent of Summit schools, said better state data will allow districts to not just evaluate teachers but determine the effectiveness of specific programs for specific students.

“Schools have been crying out for good data,” he said. “For instance, it could really help us in figuring out what works in early literacy. And right now, we can’t get that.”

But Parker said he realized it’s not easy to do it well, nor inexpensive for a time when dollars are short. “We have a $90,000 person who just manages our student test data,” he said.