Nonprofit, independent journalism. Supported by readers.

MinnPost's education reporting is made possible by a grant from the Bush Foundation.

How one high-flying charter school operator is responding to some grim test scores

Long seen as a national model, Harvest Network’s schools saw math proficiency drop 20 percentage points this year, while reading scores dropped 9 percentage points.

Like other school leaders, Eric Mahmoud, the president of Harvest Network of Schools, got the results of this year’s Minnesota Comprehensive Assessments (MCAs) about a month ago.

They showed significant drops in achievement at Harvest Network’s three schools.

“As a matter of fact, it was hard to get out of bed,” Mahmoud said. “But when I did get up I knew it was time to take a deep dive into the numbers.” 

During the six years that the news media and state officials have been highlighting positive outliers, Harvest — where 99 percent of students are black and 95 percent live in poverty — has been among the Twin Cities’ top 10 high-poverty, high-performing schools.

Article continues after advertisement

The ability to reach those students has made Harvest a national model, with a book published on its methods and plenty of media attention — including an entire episode of Soledad O’Brien’s CNN series “Black in America,” which screened at the Convention Center. Harvest’s success also earned Mahmoud a place in the National Alliance for Public Charter Schools Hall of Fame.

Yet this year, overall math proficiency at Harvest Prep, Best Academy and the Mastery School is down 20 percentage points, to 53.3 percent. Reading scores dropped 9 points, to 33 percent. For the first time, the school made the top-10 list only in math.

To be sure, the three schools still outperform the state math average for black students and Minneapolis Public Schools averages in reading and math both for African American and impoverished students. But Mahmoud and his colleagues say the dropoff is unacceptable.

“We had an explosion here,” Mahmoud said. “We are doing the forensics to try to figure it out.” 

Seven hypotheses  

Harvest is hardly alone in that effort. Overall, reading scores are up a single point over last year, for a passing rate of 60 percent. Math scores dipped 2 percent, leaving 60 percent of students proficient.

Gaps persist largely unchanged between white students and children of color, impoverished students, those with disabilities and kids learning English. In some subgroups and subjects, passage rates barely make it into the double-digits.

And Harvest is not the only historically high-flyer that, while still ahead of the pack, saw its test scores fall back toward earth. At seven of the top 10 high poverty schools, fewer than half of students read at grade level. Only three of the top 10 math odds-beaters had passing rates of more than 60 percent.

(Here we pause for the Kramer Disclaimer: Two of the schools on those lists are part of the Hiawatha Academies network, which is led by Eli Kramer, son of MinnPost founders Joel and Laurie Kramer. None of them was consulted during the preparation of this report.)

Doubtless few school leaders are celebrating this year, but Mahmoud had particular reason to pore over the subsets of data the MCAs delivered. In an effort to improve the skills that will help Harvest graduates complete college, Mahmoud has pushed his staff to learn complex teaching methods that drive critical thinking and other higher order learning.

Article continues after advertisement

Eric Mahmoud
Eric Mahmoud

After he dusted himself off, Mahmoud did two things. The first was to develop a list of seven hypotheses as to what had happened. The second was to ask Dave Heistad, the dean of Minnesota’s assessment and evaluation community, to analyze the schools’ scores. He also asked Heistad to drill down to figure out what was going on in the classrooms where gains were made.

Now the Director of Research, Evaluation and Assessment for Bloomington Public Schools, Heistad was involved in the design of many of the state’s best data collection and analysis tools.

According to Heistad’s report, at least two of Mahmoud’s hypotheses likely were significant factors in Harvest’s 2015 losses: a dramatic increase in new special ed students and the elimination of the network’s longer school year.

The demographic shift was probably the single largest factor, according to Heistad. Last year, the percentage of all Harvest Network students in special ed shot from 8 percent to 19 percent. That and an increase in the number of English-language learners (to one in five) brings the schools in line demographically with Minneapolis Public Schools. It also makes the coming years a test of whether the strategies that propelled Harvest onto the list of odds-beating programs can keep them there. 

But last year, it meant trying to reach more than 100 new students whose disabilities had already been diagnosed, but possibly undertreated, quickly.  

“That’s a good thing,” said Mahmoud. “Those are the students we want. What we didn’t anticipate is that they would have to take the same test as everyone else.”

This year Minnesota began requiring virtually all special ed students to take the regular test without modifications. And every single one of them had to do so online, which presented challenges for a population used to adapting paper tests for kids with learning disabilities.

A ‘concierge service’ to stay in touch

Add to that another problem Mahmoud is happy to have: a high number of students new to the schools. The three schools enrolled 400 new students this year, 200 of them in newly added seats, bringing the percentage of new students up dramatically. In both third and fourth grades in the original Harvest program, for instance, the number of new students was nearly triple what it was last year. 

Not surprisingly, scores were much higher among students who had attended the network two or more years. Just 28 percent of first-year third graders passed the math test and 13 percent reading, for instance. Meanwhile, all students with Harvest two or more years passed math and 68 percent reading. 

Struck by the difference in performance between new students and those who have stable attendance, Mahmoud started a “concierge service” to stay in touch with families to make sure they come back. This fall’s predicted retention rate from last spring: 85 percent. 

Article continues after advertisement

Unstable housing is the number one obstacle, so 12 Harvest employees have made contact with 100 families each at least three times this summer. If a family’s circumstance included, say, a move from Minneapolis to Brooklyn Center, the school makes sure they know it sends a bus to the northern suburb. 

Fighting the summer slide

The other big issue the data pointed to for both Mahmoud and Heistad was the impact of losing summer school. Because the state cut the funding that paid for the extended year, last year was the first in seven that Harvest had to cut its school year — by five weeks. 

Comparing Harvest’s spring 2014 assessment data to last fall’s, Heistad identified a pronounced dip. It’s typical to expect kids on average to experience a “summer slide” of about 1 percentage point.

Because they do not spend their summers traveling, attending camps and other enrichment programs the slide is more pronounced for impoverished kids. Research is divided on how summer after summer of slumps compound, but the bottom line is it’s harder to make up

Harvest’s students saw fall-over-spring dips of 3 percent to 12 percent. To put that in perspective, Mahmoud noted that they expect growth of 7 percent to 9 percent per academic year.

Heistad was out of town this week but the assessment czar who preceded him, Jim Angermeyr, now retired, said students frequently make up summer losses by the following spring’s assessment season. But the loss of several weeks of programming has to be significant, he said. 

This year’s legislature reinstated extended year funding, but—inexplicably, because high-performing charters are the schools most likely to have longer school years—now reimburses charters one-fourth the funding available to mainline district schools.

Mahmoud said he has already moved to plug the time gap. “If we have to cut whatever, we have to have that time,” he said. “Fortunately we had the data early enough we could react to it.”

As a consequence, when Harvest students start school later this month they will have a school day that’s two hours longer. And those that arrive with behavior challenges and learning disabilities will see a beefed-up special ed staff.

Article continues after advertisement

New standards also create challenges

The last big issue is one named by school leaders all over the country who are facing flat or declining assessment data. New, more rigorous standards for what students should know are challenging teachers. 

As he was this time last year, Mahmoud is convinced the new standards—designed to require critical thinking, problem-solving and other complex skills—are imperative. Experience with intellectual challenge is crucial for making sure that the schools’ alums persist through college.

Harvest’s teachers are coming back early again this year. Among the tasks they will be asked to do is to work with a new tool school leaders developed that will help them identify the best approaches.

“It was tough to present the news,” Mahmoud said. “Our teachers work their behinds off. You want to give nothing but good news to inspire them.”

He also asked Heistad to crunch the numbers for individual classrooms, not to punish teachers but to identify those who were very successful. So when he presented the grim overall results to his staff, he was able to call out the school’s stars.

“We don’t have to go to New York,” he told them. “Right here, with new students, with more special ed students, with summer slide, here are your peers who are turning it around.” 

Can they turn it around in every classroom? They don’t have a choice, Mahmoud said.