When a recent study from Stanford University’s Center for Research on Education Outcomes (CREDO) found Pennsylvania charter schools are falling behind their traditional counterparts — even while charter schools are making gains nationally — media outlets across the state declared the inadequacy of the state’s charter schools.
Headlines such as “Pa. charter students’ skills fall far short, study reveals” in The Tribune-Review and “Study: Pa. in Bottom Three for Charter School Scores” in the Pittsburgh Post-Gazette appeared in news outlets across the state. And while the study did find Pennsylvania charter schools are underperforming, charter school advocates are saying that’s not the whole story.
Robert Fayfich, executive director of the Pennsylvania Coalition of Public Charter Schools (PCPCS), wrote a letter in protest of The Tribune-Review article in which he wrote, “important facts were omitted.”
“The article summarizes a 95-page report and includes generalizations that, if taken as a full analysis of a situation, will lead to false conclusions or bad policy,” he wrote.
He pointed out that Pennsylvania has some very high-performing charter schools. Those schools should be replicated and the underperforming charter schools should be shut down, he wrote.
The CREDO study did not name the Pennsylvania charter schools that were included in the study, and the study used test results from 61,770 of the 119,000 charter school students in the state.
The study compared charter school students with their “virtual twin,” a computer-generated student reprsenting traditional public school students. Each charter school student was matched with their “virtual twin” taking into account test scores as well as characteristics such as race, gender, and eligibility for free or reduced lunches. In each following test period where results were available, the “virtual twin” was reviewed to make sure the match still applied. The report said this study method “has been assessed against other possible study designs and judged to be reliable and valuable by peer reviewers.”
Fayfich wrote PCPCS does not completely dismiss the results of the CREDO study, “but we put them in context to what is really happening in education.”
“There is a significant difference between running data through a computer algorithm in California and working face to face with children and parents in Pittsburgh,” he added.
Others were not so accepting of the study’s methodology. President of The Center for Education Reform Jeanne Allen wrote a Letter to the Editor of The Pittsburgh Post-Gazette essentially scolding the publication for citing what she called a “flawed” study.
“The reality is we cannot make conclusions about Pennsylvania charter schools or charters anywhere else without randomized control trials — the gold standard for research — that use actual student-by-student data over time,” she wrote. “The CREDO report, which produced some unfavorable figures for Pennsylvania charters, fails to use such methods. The report instead employs statistical gymnastics to make spurious comparisons of charter student achievement across state lines while altering data to ensure all students ‘start’ at the same level.”
Bob O’Donnell, former speaker of the Pennsylvania House of Representatives and a driving force behind the state’s charter school law, told Media Trackers while he hasn’t read the study, those like it are often limited. “It is impossible to make these kinds of comparisons on a statewide basis,” he said.
“You define a mission, and whether or not the mission is successful is essentially determined by your client,” O’Donnell told Media Trackers. “[Pennsylvania charter schools] have thousands of people on the waiting list.”
PCPCS data shows 119,000 students attend charter schools in Pennsylvania while 44,000 students are on waiting lists.
Devora Davis, one author of the study, told The Notebook the cyber charter schools in particular were negatively impacting the state’s charter school performance. All eight of the cyber charters included in the study did worse than traditional public school counterparts.
“The cybers are getting some of the lowest-performing students, and they don’t have a lot of time to work with them,” Fayfich told The Tribune-Review, adding that students who struggle the most in traditional public and charter schools turn to the online alternative.
And while opponents of charter schools say the lack of regulation weighs down student performance, O’Donnell said that is absolutely wrong.
“The whole theory was [charter schools] were to be schools of choice. They were to be unregulated,” he told Media Trackers. “But they’re grossly overregulated. Every time someone gets a half-baked idea, they impose it on the schools.”
O’Donnell added that charter schools are most effective when a student enters the program early and sticks with it.
“If you have a student coming into high-school algebra, and he has a fourth grade math level, you will fail to teach him algebra,” O’Donnell pointed out.
In order to boost charter school performance, he said, charter schools need less regulation, as it was intended.
“No more unfunded mandates; deregulate the schools; focus on children, not schools,” he said. “There’s this assumption every kid can learn in the same way, in the same setting, at the same time. There’s a set of standards based on that idea. We need to get rid of those standards.”
Nationally, the study found that roughly a third of charter school students perform better than traditional public school students while the remaining two-thirds of charter school students perform no better or worse. The study also found that minority students are making the most substantial gains in charter schools.