The timing was good. A growing number of families choose the small learning models and millions in public funds support them through school choice programs. An added bonus: The project would have been among the first from an independent research organization to measure the impact of microschools on student learning.

A year later, the researchers came up empty.

In fact, Jonathan Schweig, a Rand senior scientist, concluded that determining the impact of attending a microschool on student outcomes is 鈥渘early impossible鈥� because of a lack of useful assessment data. Parents often flee traditional schools and opt for unconventional models precisely because of a perceived over-reliance on testing. 

Nonetheless, Schweig added, testing 鈥渋s important to understand how students are doing relative to other similar students.鈥�

Co-researcher Laurel Covelli was scheduled to present preliminary findings from the study, shared exclusively with 社区黑料, at this week鈥檚 Harvard University conference on Emerging School Models. But instead of solid conclusions, she鈥檒l focus on the challenges of studying a sector that even supporters say is . 

Microschools 鈥� small learning centers that meet in homes, renovated storefronts and churches 鈥� have been popping up across the country since the pandemic. In fact, recent research from Tulane University, shows that they represent  of private schools serving students on vouchers or education savings accounts.

According to the National Microschooling Center, nearly 40% of the small schools serve students using state-funded school choice programs. Microschools now number nearly 100,000, with the sector expected to expand  further when states opt in to the new federal tax credit program for school choice. There are anecdotal accounts of parents who say their children thrive in the close-knit, personalized settings, as well as examples of start-ups that fizzled. But objective achievement data can be much harder to find, even as their growing numbers are sparking calls to hold them to the same academic standards as public schools.

In May, Lisa Tarshis, part of the Primer microschool network, said she was looking forward to Rand鈥檚 findings. 

鈥淎s a movement, we have to be successful in what we’re doing 鈥� and we have to have data to show that what we’re doing is working,鈥� Tarshis said. As head of the Primer Foundation, which provides financial aid to families and support to schools in the network, she said she鈥檚 a 鈥渂ig believer in full transparency鈥� and the power of the free market. 鈥淎ny business that isn’t meeting the needs of its customers will go out of business. The same should be true for schools.”

One leading supporter of the sector said she isn鈥檛 surprised Rand wasn鈥檛 able to do a 鈥渃omprehensive comparison.鈥� 

鈥淢any families tell me that they chose a microschool or similar model because they wanted less testing and more focus on the whole child,鈥� said Kerry McDonald, who recently released a new book on microschools and often contributes to 社区黑料. But she doesn鈥檛 consider the lack of data to be a 鈥渟etback鈥� for the movement, noting that microschools increasingly serve students with autism and other special learning needs 鈥渨hose academic abilities may not be adequately reflected on traditional tests.鈥�

Missing data

The Rand team initially approached the project as if they were evaluating charter or private schools, intending to examine how microschool students performed in reading and math on the MAP Growth tests. Because the tests from NWEA, an assessment company, are widely used in public schools, a comparison should have been possible.

But the data didn鈥檛 exist.

Rand gave NWEA a list of 271 microschools, compiled with the help of the Las Vegas-based National Microschooling Center. NWEA could only find 10 with fall and spring scores. Without both, Rand couldn鈥檛 determine if microschool students learned any more or less in a year than their counterparts in traditional schools. 

From the 10 schools they did identify, preliminary data shows that microschool students grew, on average, a third of a percentage point less in math and one half of a point less in reading than non-microschool students.

But the bigger takeaway is the absence of a large enough sample to produce meaningful results.

鈥淢icroschools are not little charters. A lot are really new and don’t yet have two test events per student,鈥� Schweig said. Another issue is that because the schools are so small 鈥� with a median size of 22 students 鈥� parents might register for MAP assessments on their own and are never linked to a specific school.

But the primary reason for the lack of data, he said, is that 鈥� a large number of microschools are not opting in to do these assessments in the first place.鈥� 

Schweig and his co-authors hinted that they were running into trouble in March. In an , they said that information on students鈥� backgrounds and academic performance was often unavailable. The 鈥渓ack of data poses a threat鈥� to the validity of any studies on the impact of microschools, they wrote. 

Rand isn鈥檛 the only research organization trying to measure student achievement in microschools. Mathematica is currently analyzing data conducted with Rock by Rock, a project-based learning company based in New York City. The goal is to pair reading data from i-Ready, another commonly used public school assessment, with videos and student work to offer a more well-rounded view of what kids are learning. 

Schweig said the Mathematica project may have more success because microschools had to administer i-Ready three times a year in order to participate and get a discount on a Rock by Rock subscription. Jeffrey Imrich, CEO and founder of Rock by Rock, said he didn鈥檛 know if the researchers were having any complications. A Mathematica spokesperson declined to comment on the status of the study.

鈥楶arents are happy鈥�

Some microschools assess students with traditional tests. According to a from the National Microschooling Center, a third of leaders say they use standardized assessments to measure student progress, but observations and portfolios of student work are more common. There鈥檚 also a shift toward accreditation, largely because some states require it from microschools that serve students using ESAs. In August, KaiPod, a microschool network, a partnership with accrediting agency Cognia to release accreditation standards specifically for the sector.

Adina Victor, vice president of non-public school services at Cognia, said they鈥檙e not trying to make microschools 鈥済ive up鈥� what makes them attractive to some families. But if leaders don鈥檛 gather student achievement data, their school may suffer.

鈥淚f you want state funding, there is a level of accountability that comes with that,鈥� she added. 

Comparisons between microschool students and their counterparts in district schools do exist. They鈥檙e just not always easy to find. 

Prenda, one of the first microschool networks, published an 鈥溾�� for the New Hampshire Department of Education based on 2022-23 i-Ready data. It showed that Prenda students outpaced other New Hampshire students and the nation. But Kelly Smith, founder and CEO, said it was a one-time requirement for a contract with that state. 

Most states don鈥檛 require students on ESAs to take the same annual tests as kids in public schools. Indiana is an exception, but the state only reports data from accredited microschools. Most are not. 

In West Virginia, microschool students receiving the Hope Scholarship, an ESA, are required to take a standardized test or submit a portfolio of work. Most parents choose the second option, according to the state treasurer鈥檚 office, which runs the program. But officials only track whether students meet the requirement, not specific results. 

Michael Parsons, a microschool founder in the state, said he鈥檇 like to see some kind of 鈥減ublic-facing dashboard鈥� with data on what test microschool students take and how their results compare to national norms. He gives students the Iowa Assessments, commonly used among homeschoolers, and said 鈥渋t’s a disservice to students whose first time sitting down for a standardized test is when their admission to college depends on it.鈥� 

Left-leaning think tanks, skeptical of private school choice, argue for stronger testing requirements. A February questioned the quality of education in microschools because many founders are not professional educators and often develop their own curricula. 

EdChoice President Robert Enlow, and other advocates, tend to point fingers back at public schools. In , Enlow said families are leaving traditional schools because they鈥檙e 鈥渘ot performing up to snuff.鈥� 

Others argue that parents are the best judge of whether their children are learning.

鈥淭he number one measure that we have to demonstrate that these programs are working 鈥� is that their parents are happy,鈥� Lindsay Burke, now with the U.S. Department of Education, said last year in an . At the time, she was still an education expert with the conservative Heritage Foundation.

Rachel Good, another Rand study participant, founded the in Chattanooga in 2022.  A former public school special education teacher and reading specialist, she believes students need 鈥渇oundational academics,鈥� but she wants them to do more than 鈥渞egurgitate鈥� facts on a test. 

鈥淚 want to inspire these kids to find their voice,鈥� she said. In public school, 鈥渨e didn’t have time to teach any meaningful social and emotional awareness. That’s something you’re going to use every day of your life.鈥�

The lack of knowledge about how students perform on academic tasks has implications, not just for the public, but for foundations supporting school choice, said Schweig, with Rand. The Walton Family Foundation funded the study as part of a larger grant supporting choice research.

鈥淪ome schools are doing great things,鈥� he said, 鈥渂ut there is value in understanding and helping schools that are not doing great things do better.鈥�

Disclosure: Walton Family Foundation provides financial support to the Rand Corp. and 社区黑料.