SRI Education researchers provide administrators and policymakers with the context they need to make informed decisions about special education services.
More than 7 million children in the U.S. (15 percent of all public school students) receive special education and related services.
These services can be transformative for students and families. Even so, access to and delivery of these services can vary widely across schools and districts. To continue to improve special education and better serve students, it’s critical to investigate these differences and share lessons learned across school districts.
For researchers in SRI’s Education Division, improving special education services often means working with collaborators (including families, practitioners, and policymakers) to translate and interpret data in ways that show a clear path toward positive change.
Whether working with education stakeholders to mine legacy datasets for new insights or developing next-generation AI tools to study special education services, SRI is identifying how special education can adapt and evolve.
How SRI turns data into action
When researchers want to understand broad patterns in special education — like who receives services and how outcomes unfold over time — they often turn to large longitudinal datasets, including those maintained by federal agencies like the Department of Education (ED). SRI’s recent work with the Administration for Children and Families (ACF), for example, took advantage of ED’s Early Childhood Longitudinal Study–Birth Cohort (ECLS-B) dataset. This new work identified opportunities to improve early childhood screening and referral practices in home-based and informal care settings, strengthen connections between early care providers and early intervention systems, and support families in navigating evaluation and service systems.
While SRI’s work with ACF and ECLS-B demonstrates the power of national datasets to generate relevant insights, the data systems take years to build and sustained federal investment to maintain. Researchers are increasingly exploring complementary approaches that can deliver more immediate, ground-level insights. At SRI, researchers are taking advantage of a new opportunity: using AI to build and examine data sourced directly from individual schools and districts. Individualized Education Programs (IEPs), for example — the legal agreements that outline eligible students’ special education services — provide some of the most valuable and detailed documentation of how special education services are intended to operate on the ground. But because they’re specific to individual students, they’re also closely protected — and for good reason. Exposing personally identifiable information would violate federal privacy protections and breach the trust of schools, students, and families.
But what if you could learn from these data without inviting any privacy concerns? On this front, a team led by SRI senior education researcher Adrienne Woods is making some important progress.
“Particularly if districts have a lot of messy text-based data, these tools can really help to quickly drill down to insights without spending hours and hours doing manual redaction or coding.” — Adrienne Woods
“Data collected directly from districts will sometimes provide strategies that are more immediately actionable at the local level,” Woods notes, “and schools and districts are already sitting on a wealth of data. Because we can now turn to machine learning and AI, we’re finding some incredible opportunities to uncover insights about special education at a scale that would have been impossible just a few years ago.”
Woods leads a team that is building an AI-powered platform to accelerate research about special education programming, called SEAMLESS (Special Education Applications of Machine Learning to Enhance Student Success). This platform includes tools that allow users to redact sensitive and personally identifiable information in bulk from IEPs and transform their contents into analyzable data.
“Our goal is for a school district to operate the redaction tool on their end and then just pass a batch of fully redacted documents to organizations like SRI or other technical assistance providers,” explains Woods. “And then we have a second AI-based tool that we’re using to analyze the contents of those IEPs and generate insights that can inform practice.”
Making data insights seamless for local districts
SEAMLESS is already being put to use on behalf of school districts around the country. Four districts signed data use agreements with SRI in 2025, and the research program will continue to scale this year.
“We’re just scratching the surface, but we’re already seeing how this AI-led approach can be invaluable to districts seeking to improve services across multiple schools,” says Woods.
SEAMLESS might help districts notice patterns in parent attendance at IEP meetings and adjust their outreach strategies, for example, or identify gaps in support for families who speak languages other than English. SEAMLESS can also assess the text that school staff enter into IEP forms, helping ensure that documentation is clear, consistent, accurate, and useful for monitoring student progress. Furthermore, SEAMLESS can analyze patterns in the types of supports students receive, like behavioral interventions or assistive technologies, to ensure those services are aligned with student needs and accessible to all students who need them.
“Particularly if districts have a lot of messy text-based data, these tools can really help to quickly drill down to insights without spending hours and hours doing manual redaction or coding,” Woods emphasizes. “And the tools allow you to be pretty flexible in terms of how the data are stored and maintained. We’re excited to continue applying it to questions about special education and beyond.”
Learn more about how SEAMLESS can evaluate IEP quality, predict student outcomes, illuminate differences in service delivery, improve IDEA compliance, and much more.
“From Developmental Concerns to Identification: How Early Childhood Education and Care Shapes Support” was supported by the Administration for Children and Families (ACF) of the United States (U.S.) Department of Health and Human Services (HHS) as part of a financial assistance award (Award #: 90YE0335-01-00) totaling $99,754 with 100 percent funded by ACF/HHS. The contents are those of the author(s) and do not necessarily represent the official views of, nor an endorsement, by ACF/HHS, or the U.S. Government. For more information, please visit the ACF website.
