Melissa Bailey Photo
Will each school get a number grade, or a letter from A to F?
Neither, said Garth Harries. The schools’ first report cards, which are due March 15, will look more like a 3‑D matrix.
Harries (pictured), matrix mastermind and school reform czar, painted the picture at Monday’s meeting of New Haven’s Board of Education.
Harries recommended, and the board unanimously approved, criteria by which the district’s 47 schools will be “graded” and placed into three tiers as part of the city’s nascent school reform drive.
Those tiers are part of a new system where schools will be managed differently according to their performance. Top-performing schools will be placed into Tier I and given more autonomy. A handful of lowest-performing schools in Tier III will be closed and reopened under new leadership, possibly as charter schools, with more flexible work rules.
Not all schools will be put into tiers right away: Schools Superintendent Reggie Mayo plans to start with a few schools this fall.
By union contract, Mayo has until March 15 to grade an initial batch of six to eight schools and place them into three tiers. Of those, one or two so-called “turnaround schools” will be closed at the end of the school year, and teachers will have to reapply if they want to continue working there.
Starting next school year, all the 47 schools in the district will be graded annually.
How will schools be graded?
Grades will rest on three factors: students’ absolute performance on standardized tests, improvement on those tests, and “school climate” factors.
Mayor John DeStefano, Jr.
The new guidelines came with the blessing of the reform committee, a group of teachers, parents and administrators who are drafting key components of the school reform effort. Teachers union vice president David Low, who sits on the committee, said the group agreed that the district is doing its best to come up with a makeshift grading system without key data.
“It’s incomplete data, so it’s not a perfect world,” Low said, “but we have to go ahead” and meet the March 15 deadline, he said.
Data is most incomplete in the “school climate” category, which aims to evaluate how engaged students are in their school. The school reform team is drafting new surveys to measure the “climate” of schools, but those surveys won’t be ready by March. So the district will use two existing measurements: average student attendance, and something called the Teacher Child Rating Scale, which is a survey teachers fill out about how their kids behave in class.
In the future, student improvement will be measured by a more sophisticated measurement based on Colorado’s “growth model,” according to Harries. For now, the school system has to use the data it has.
What will a school’s first report card look like in March?
It won’t be a simple number grade, Harries said.
It will look more like a matrix. Or a cube. Or a 3‑D vectorspace with one axis jumping out at you from your computer screen.
That picture arose during a roughly hour-long discussion before the school board.
Harries presented the proposed criteria to the board. They remained largely unchanged from the ones he presented two weeks ago—.(JavaScript must be enabled to view this email address) for a back story.
One notable change was made to reflect feedback from the school board: a “qualitative factor” section that gives the superintendent wiggle room to look at more than just numerical scores. Mayo had said he wanted that wiggle room, but he hadn’t spelled out what factors he’d be using. The draft Monday clarified that he will look at the “quality of school leadership team,” the “readiness of school team to engage in collaborative planning,” and “community and other resources available to support school-level planning efforts.”
Education watchdog Alex Johnston, a school board member, took pause at the qualitative factors. He pressed for a quantitative system that’s transparent and easy to understand. The district should take into account qualitative factors, like how ready school leadership is to take on reform, but those factors should be taken into account after schools are graded into tiers, not before, he argued.
From Harries’ description, board members weren’t sure whether all schools would be graded the first time around, or whether Mayo might just pick six to eight schools and grade only those.
Johnston urged that all schools be graded along with the initial batch of six to eight, so that the public — and school communities — can see how Mayo arrived at his decisions.
“We need an objective way to tier schools,” he said. “Otherwise, you, Dr. Mayo and your team, you kind of go back behind close doors” and emerge with a list of schools for the pilot program. “I’m not sure that’s transparent enough for people,” Johnston said.
Mayo acknowledged the need for an “objective” measurement, but said he needs added “flexibility” given the time constraints of getting the pilot program up and running.
Johnston asked Harries about another concern: The guidelines lay down the criteria for grading, but they don’t say how much each factor would count. Would student test performance make up 30 percent of a school’s grade, or 80 percent? Johnston said he was uncomfortable approving criteria without knowing how much each factor would count.
He asked Harries if each school would get a final number grade, like in New York City, where schools get a grade from A to F.
No — at least not this time, said Harries. The grading system for future years is still evolving, but Harries laid out a proposal for the initial batch of report cards due March 15.
He said the district wouldn’t use a weighted average to compute the three factors into a final composite score.
Instead, Mayo would eyeball a few lists and look for outliers, Harries said.
The schools would be ranked in two pools — K‑8 schools and high schools. For each category, Mayo would come up with a ranked list. He’ll lay those three lists next to each other and look for extremes. He’ll identify schools that fit easily in one of three tiers — a couple schools that sit high on all three lists, a couple that sit right in the middle, and a few more that sit at the bottom across all three categories.
High schools would be measured across only two categories: The district doesn’t have student improvement data for high-schoolers because they don’t take the Connecticut Mastery Test.
Elementary schools would be graded on all three categories, with three levels on each.
That means schools would be graded on a three-by-three matrix, or a cube, Johnston observed. Where a school lands inside that cube determines its tier.
Harries accepted the illustration.
He did not commit to creating a giant laser cube when he presents the initial round of school grades to the board, but he did say he’d make the big “matrix” public, so anyone can see what went into grading the schools. That means each school could find out how it fared in each category.
The board approved the guidelines, with an amendment ensuring that the matrix be made available to anyone who’d like to look at it.
During discussion preceding the vote, Mayor John DeStefano jumped into a somewhat esoteric debate on whether the district needs to define a weighted average for each component of the grade, or whether it’s OK to do as Harries suggested — map them out and see where they fall.
DeStefano said he’s OK with the latter. At least for the first round of grades, a Tier I school would smell like a Tier I school, so to speak. You would “know it when you see it,” he said.