Students attending microschools are progressing at the same rate as their peers nationally and sometimes a whole lot faster than expected, even though parents often choose such programs because their kids have disabilities or other significant challenges, new data shows.
From fall 2025 to spring 2026, about 400 students from 27 schools made gains in math and reading on the widely used i-Ready tests, with more than half exceeding expected growth based on national norms in reading and 47% doing so in math.
Get stories like this delivered straight to your inbox. Sign up for The 74 Newsletter
Some microschools, especially larger ones, were “knocking it out of the park,” with “every student achieving over and above what was expected.” said Daniel Hamlin, dean of education at the University of Texas at Austin, who conducted the study on behalf of the National Microschooling Center, an advocacy organization.
Nearly half of the students had special learning needs, 28% had mental health needs and 14% had a chronic medical condition.
“You have a very high percentage of kids who would be classified in a public school as having a disability or some kind of other significant life circumstance, like a severe medical condition,” he said.
I-Ready creates a projected score for each individual student, based on the data from public and private school students nationwide. In the microschool sample, results were weaker among students in poverty, those in the elementary grades and Black students in math.
The study has multiple limitations. It’s small and the students who took the tests might not be representative of the microschool population in general, estimated at 1 million to 2 million students nationally. Families who choose microschools may also be different in several ways from those in public schools, and the schools weren’t randomly selected. Still, the research is significant because Hamlin was able to collect data from the same kids in the same microschools over the course of the school year. The Rand Corp. tried something similar last year and ran into multiple obstacles.
“This is a valuable and timely study in a space where we still have far too little empirical evidence,” said Jonathan Schweig, the senior scientist who led Rand’s project.
In that study, out of a sample of 271 microschools nationwide, he couldn’t find enough programs that had both fall and spring assessment data to reach any meaningful conclusions about how much progress students made in a year. The perception that there’s too much testing in the public schools is often one of the reasons families leave them for microschools.
Schweig added that it’s still difficult at this stage to “make broad claims about microschools given the sector’s diversity.”
Hamlin, who is slated to present the data Thursday at an annual Harvard University conference devoted to emerging education models, elaborated on why the sector is difficult to research. As parents piece together different learning experiences for their children, often with state-funded education savings accounts, students sometimes “fall into multiple categories.” They may homeschool part of the week, take an online course and spend a few days at a microschool.
“It’s really hard to know exactly how many students are in a microschool because there’s just so much overlap between all these different emerging school models,” he said.
One-on-one learning
The i-Ready results are part of a larger report that includes survey responses from both parents and students. Small classes, individualized attention and a safe school environment were the top three reasons parents gave for choosing a microschool.
“Finding a school with well-trained teachers who are able to recognize strengths and challenges without passing judgement is really hard,” said Lara Landrum, whose daughter, a sophomore, is in her third year at the Columbia Academy for Learning and Enrichment, or CALE, in Columbia, Missouri. While it’s possible for students to feel like they belong in a large school, she said, “the intimacy of a microschool allows for this naturally.”
Like many of these non-traditional programs, CALE opened during the pandemic, operating out of a former medical office. Now housed in a church, the school has expanded from the elementary grades to offer a full junior K through 12th grade program.
Roughly 80% of CALE’s 135 students previously attended public schools. Some, said CEO Jill Brown, struggled with reading disabilities and needed “structured literacy support,” others had gaps across multiple subject areas and some kids were working so far above grade level they needed more advanced assignments.
“We also have students who experience anxiety in larger school environments and benefit from the smaller, more supportive setting at CALE,” Brown said.
In the survey, those who moved their child out of a public school, about 20% of the sample, said these students are making more progress and have a tighter circle of friends in their current microschool. The school also feels safer and is better at communication, they said. The majority of students reported having one-on-one time with a teacher nearly every day, small group learning and making close friends.
Given the growing interest in these small private schools, Hamlin recommends several areas for additional research. Future studies should have a larger sample and compare microschool students to those who are demographically similar in district, charter and other private schools.
Because there are different types of microschools, like nature-based programs, models that combine virtual and in-person learning, and those that serve specific populations, he suggested analyzing results to better understand which are more likely to boost student performance.
Considering many microschool students with disabilities are using ESA funds, researchers could do a “cost-benefit analysis” comparing public schools to microschools. More rigorous studies will also follow microschool students over a longer period of time, Hamlin said.
He acknowledged that working with the National Microschooling Center, a group that’s advancing the movement, to recruit schools for research, “does end up creating some limitations because you get a self-selected sample.”
“But from my vantage point, we have to start somewhere,” he said.