NC public schools task force considers moving on from EVAAS growth metric

For more than a decade, the Education Value-Added Assessment System (EVAAS) has been the engine behind North Carolina’s school growth scores. Now, state educators are weighing whether to recommend moving on.

At a June 18 virtual meeting, the State Board of Education Task Force on Accountability for Public Schools reached consensus to have the Department of Public Instruction (DPI) research alternatives to EVAAS, the proprietary growth model produced by SAS Institute that the state has used to grade schools since 2013. 

It was the clearest signal yet that one of the system’s central measures could be replaced, although that step would take the approval of the General Assembly.

EVAAS is the model that North Carolina uses to calculate growth in student achievement from year to year. That’s a crucial piece because achievement growth is linked to a school’s overall grading. But members of the task force said the model is nearly impossible to explain or reproduce.

“Nobody does truly understand how it works — the formula is proprietary,” said Alan Duncan, vice chair of the State Board of Education, who led the meeting. He noted that a superintendent “can’t sit there and calculate it or replicate the calculations.”

Paul Kinsey, director of testing and accountability at Dare County Schools, said principals are left waiting on a number they cannot account for. “What we do is at the end of the year scores come back, we hold our breath until the magic EVAAS number comes out,” he said.

Kim Mackey, a Wake County teacher and member of the North Carolina Association of Educators, said her deeper concern was fairness, arguing the model is built so that some schools or educators will always come up short.

“I would like for whatever we use as a growth measure to pass the classroom test, where if I were to say to my students at the beginning of the year, no matter how much you all grow, some of you will be labeled as just not having effectively grown,” she said. “I think parents, my colleagues, and even my principal would have an issue with that.”

Duncan pointed to last year, when some schools raised their proficiency scores but still did not meet growth “because EVAAS is on a bell curve.”

Chris Brandt, an associate with the Center for Assessment, the nonprofit advising the task force, explained why: EVAAS defines expected growth as the growth that would occur under average instruction, he said, so by design the score of the average school is zero.

“You can’t have all schools above the state average any more than you can have all students above the median,” Brandt said. A value table, by contrast, is criterion-based, meaning “anyone could meet growth or exceed growth.”

Moving beyond EVAAS could be challenging. Even if the task force recommends dropping EVAAS from school grades, the tool would remain the basis for measuring individual teacher and educator effectiveness, a separate use the task force isn’t reviewing.

Supporters also pushed back on the criticism, saying EVAAS measures students against themselves, not against other schools.

“It’s an expected growth model, not a ranking model,” said Janet Mason, a State Board of Education member. “It’s more of: did your students grow as much or more statistically than expected based on their prior achievement?”

John Hood — a task force member and member of the board of the John Locke Foundation, which publishes Carolina Journal — said he was not persuaded the alternatives were any better.

“I don’t see a compelling reason to move away from our value-added model,” Hood said, adding that he was “not really persuaded that the other options are a net gain.”

He warned that swapping out a long-used model carries its own costs.

“There is a downside to that from the standpoint of parental confidence, familiarity by the educators and administrators using it, and confidence among policy makers that this makes sense,” Hood said. “We all know that this whole area is fraught with political and activist peril, and you wouldn’t want to have a change that was made simply for technical reasons be perceived as something else.”

Still, members agreed to keep looking. The task force asked DPI to bring back alternatives — including value tables and student growth percentiles — at a future meeting. Mackey noted that South Carolina has already dropped value-added measures like EVAAS.

No formal votes were taken. The task force next meets Aug. 20 and plans to begin writing recommendations in December.

“NC public schools task force considers moving on from EVAAS growth metric” was originally published on www.carolinajournal.com.