Regents grade ‘scrubbing’ helped push students to graduation, study finds
April 13, 2016
By: Monica Disare
New York City teachers who boosted their students’ Regents exam scores by a few points during the Bloomberg era didn’t benefit — but their students did, according to a new study.
The study, published by the nonprofit National Bureau of Economic Research, found no evidence that teachers who manipulated scores did it to increase their own pay, or even to help their schools avoid closure. But students whose scores jumped were 27 percent more likely to graduate from high school — indicating that the teachers were often helping to spare students from missing out on a diploma.
The researchers took a close look at test results from 2004 to 2009, when teachers were allowed to grade their own students’ Regents exams. Students must pass five to graduate, and a 2011 investigation found a surplus of scores that just hit the passing mark of 65. Teachers were also asked to re-grade exams near the cutoff for several years, which means they had an opportunity to catch mistakes.
“Here it appears to be teachers trying to help marginal students,” said Thomas Dee, a professor at Stanford and the lead author on the study, “and that sort of ties into a broader tension in education policy debates.”
Rules changes have largely ended that kind of score manipulation, sometimes called “scrubbing.” Teachers no longer grade their own students’ exams, in most cases, and exams that just miss a passing grade no longer get a second look.
But the changes have done little to answer broader questions about which students deserve a high school diploma.
Some argue it makes little sense to deny students a high school diploma for coming just a few points short of the passing score on one of the tests. The global history exam in particular has been known for tripping up students, who retake the test repeatedly in order to pass. But others argue creating too many exceptions lowers the state’s graduation standards.
It’s an issue that New York’s education policymaking body is grappling with now, after the state moved to toughen its graduation requirements over the last few years. The state’s Board of Regents has been expanding “safety net” options for students who fail exams. Officials have already authorized a few alternate graduation “pathways” that allow students to bypass the fifth Regents exam in favor of another test, and now also allow students to appeal lower test scores.
In a way, teachers who re-graded exams with near-passing scores created their own, informal appeal system before those options were available.
The analysis found that not all students got an equal bonus, though. Teachers were more likely to offer a couple of extra points to students with better prior test scores or good behavior.
The practice also had a disproportionately positive impact on the graduation rates of students of color. The graduation-rate gap between the city’s black and white students would have been about about 5 percent larger without the test score manipulation, according to the study.
Though students of color were more likely to be helped in the aggregate, because more black and Hispanic students earned scores just below the cutoff score, teachers were more likely to boost the grades of white and Asian students that scored near a passing grade, the report said.
The researchers found high rates of test manipulation even in 2001 and 2002, before the federal No Child Left Behind Act and the city’s letter grading system had ramped up pressure on schools to improve their graduation rates. The researchers say this offers more evidence that teachers were more motivated by student success than their own.
But the researchers used statistics, not interviews, to draw their conclusions.
Without teacher voices, saying that educators weren’t motivated by the city’s accountability system “a pretty grand inference,” said Aaron Pallas, a professor of sociology and education at Teachers College at Columbia University.
The researchers used the same methods to analyze whether the the city’s experiment with paying teachers for their students’ success on Regents exams, which lasted from the 2007-08 to 2009-10 school year, had any effect on giving students extra points. The study finds no evidence that it did.
“There’s something much more altruistic going on that’s motivating teachers here,” Dee said.