Ian Kingsbury, PhD
July 2026
Summary
In this July 2026 report, Ian Kingsbury reviews how virtual schools are evaluated and argues for a broader view of their performance. He focuses on limits in existing research, differences among the students these schools serve, and outcomes that matter to families beyond standardized tests.
The report describes a central problem with comparisons between virtual and traditional schools: students who choose virtual education may differ in ways that available data cannot capture. Kingsbury also points to concerns about how economic disadvantage is measured and how unfamiliar testing locations may affect virtual students’ results.
He argues that evaluations should consider families’ reasons for choosing a school and outcomes such as attendance, anxiety, engagement and school fit. The review describes virtual schooling as an option some families seek after bullying, emotional challenges or academic difficulties in conventional schools.
On school management, Kingsbury describes mixed evidence and cautions against treating nonprofit status or operator type alone as a reliable guide to quality. He explicitly says the review does not prove every virtual school is effective and urges caution about sweeping conclusions from test-score comparisons alone.
Key findings
Kingsbury argues that test-score comparisons can overlook differences in the students virtual schools serve and the reasons families choose them. His review calls for attention to student well-being, school fit and parental preferences, while acknowledging that it does not establish that every virtual school is effective.
Questions and answers
What kind of report is this?
A methodological review of virtual-school research and an argument for evaluating a broader range of outcomes.
Why might test-score comparisons be incomplete?
Students may enter virtual schools for reasons that available data do not capture, making comparison groups difficult to match.
What other outcomes does Kingsbury discuss?
Attendance, anxiety, engagement, student well-being, parental preferences and school fit.
Does the review establish that every virtual school works?
No. Kingsbury explicitly rejects that conclusion and calls for a more balanced evaluation of the students served and their needs.
Should virtual schools be judged by test scores?
Why do virtual schools have low test scores?
This report reviews existing research. It concludes that test-score comparisons alone are an incomplete basis for judging virtual schools, because of selection bias, measurement problems, and the distinct needs of the students those schools serve. It does not offer an effectiveness verdict on virtual schools as a whole, and it does not claim that every virtual school is effective. Related EFI papers: selection bias in online schooling, test anxiety in virtual schools, and parental satisfaction at virtual charters.
Read the paper
Read the full paper (PDF): Virtual Schools Reconsidered: Methodological Limits and Broader Outcomes
Cite this study
Kingsbury, Ian. (July 2026). Virtual Schools Reconsidered: Methodological Limits and Broader Outcomes. EFI. https://www.efinstitute.org/wp-content/uploads/2026/08/EFI-REPORT_Virtualschool_IK.pdf
