Today, the Center for Research on Education Outcomes (CREDO) released a ground-breaking report on virtual schools which found student outcomes at virtual schools lag dramatically behind both traditional public schools and peer charter schools.
While the headlines will no doubt focus on the report’s finding of lessened academic growth for the average virtual school student (180 days lost in math!), here are a few things that struck me as I read through the report:
It’s worse than we thought
There have been a number of other reports on the academic performance of virtual schools, almost all of which have concluded that the vast majority of students in virtual schools were not being served well. The report makes that point even clearer: compared to the academic growth of their socio-demographically identical peers, students in virtual charter schools experience a typical academic loss of -.25 standard deviations in math across all tested students. Researchers almost never see a quarter of a standard deviation in social science research. Those are breathtaking estimates of the number of learning days lost.
The high mobility of virtual school students could contribute to the low achievement.
Across the study’s sample, only half (53%) of virtual school students return to any virtual school for a second year, and only a quarter (25%) return for a third year. This is a sure-fire recipe for low achievement supported by decades of research in educational psychology.
When broken out by subgroup, virtual schools do even worse.
The study finds Hispanic students lose 79 days in reading and an astounding 209 days in math. Special education students lose more than 216 days in reading and 288 days in math. You get my point.
Virtual networks don’t produce better results.
The study also analyzes the performance of 21 networks (virtual school operators with more than three schools). All but one had negative days of learning for reading; all but one had negative days of learning for math. It seems the benefits of scale do not apply to virtual operators.
And finally, there are clear implications for authorizers, the organizations that approve and monitor virtual schools under the current model. First, we must set a high bar for every charter school—brick and mortar or virtual—and then make sure each school meets these high expectations. The study’s results reveal that there are many virtual schools that are dramatically failing to meet their expectations and should be closed. This will be difficult, but as authorizers, we must honor our duty to protect students and the public interest.
Going further, the report calls into question whether virtual schools should be included in the charter school model at all. By definition, charter schools must be open to every student. The report’s overwhelming results indicate virtual schools are failing the average student.
In the coming months, my colleagues and others will push state legislatures to address the report’s findings. The report shows that our current model is unacceptable and new policy measures must be adopted that create incentives for children to succeed in virtual schools.
At NACSA’s annual authorizing conference last week, innovation was discussed often and passionately. Authorizers should continue to create more quality seats for children and allow innovative schools to open and thrive. There is no place, however, for any school that fails students in public education—especially as dramatically as these results suggest. Today’s report should sound an alarm to every authorizer and policymaker around the country that many virtual schools are not producing the kinds of outcomes many hoped they would under the current system.
Check out Education Week’s commentary post, which includes additional highlights from the report.