Shari Dubos: Improving education through research, accessibility, and AI

---

• As AI makes its way into classrooms, innovation needs to be balanced with rigorous research.
• SRI Education’s Shari Dubos pursues research that addresses some of the most important questions at the intersection of technology and education.
• Here, Dubos traces her path from aspiring teacher to education researcher and shares how rigorous AI-focused research can shape the future of learning.

---

Shari Dubos is a principal education researcher at SRI whose work focuses on improving quality and access to STEM education. Her research spans digital learning tools, innovative teaching approaches, and community-building efforts that help educators and researchers learn from one another. In this conversation, Dubos reflects on the formative experiences that led her to education research, the promise of accessible technology, and why the next wave of breakthroughs in education will depend on pairing innovation with evidence. 

How did you decide that you wanted to focus your career on education research?

I knew from the time I was in first grade that I wanted to go into education. Education was always really powerful to me. I had amazing teachers who influenced both my values and my career trajectory. I worked as a camp counselor every summer that I could, and I was babysitting from the time I was 12. I loved being a student, but I also loved being around kids who were younger than me and trying to support them.

I went to college with the goal of becoming a teacher, but I became increasingly aware of persistent achievement and opportunity gaps in education. A lot of my student teaching and field hours were in urban Miami schools. I was witnessing this in real time.

That experience motivated me to do work that could address those issues at a broader level, which drew me toward educational research, evaluation, and measurement. I landed at SRI straight out of grad school. I’ve stayed because it’s a place where I can help design, develop, and rigorously evaluate innovative programs and tools aimed at improving outcomes in education.

What’s been one of your most fulfilling projects in terms of seeing how rigorous research can influence learning outcomes?

One that stands out was an Institute of Education Sciences evaluation of the Kasi Learning System, developed by Alchemie. 

We proposed a rigorous single-case design study, which is more common in medical research but made sense for this work because we were focusing on a low-incidence population: blind and low-vision high school students. If you want to do rigorous research with a low-incidence population, it can be very difficult to conduct a fully powered randomized controlled trial. So, we asked: Why not apply single-case design here?

The project examined how an accessible, multisensory chemistry tool could support students who are blind or have low vision in high school chemistry courses. We collected rich qualitative data from teachers and students, and those interviews showed increased agency, independence, and a stronger sense of belonging.

That was huge, because students who are blind or have low vision often face real barriers in STEM courses like chemistry, which are so visually based. For me, the project underscored how rigorous research can inform product development and elevate student experiences in ways that traditional metrics alone may not fully capture.

SRI Education’s work today often relates to AI: Both studying how AI impacts learning and finding innovative, responsible ways to infuse AI into education. How does that theme manifest in your research?

One great example is SRI’s Safe and Accessible Data Interactions in Education (SADIE) project.

It’s exciting because it tackles two urgent challenges in education right now: the growing data literacy gap and the rapid introduction of generative artificial intelligence into classrooms in ways that are not always truly accessible, useful, or safe.

Rather than building just another AI-powered tool, SADIE focuses on developing a middleware layer that sits between students and AI-enabled educational technology products. The idea is to help keep all AI interactions on task, accessible, and meaningful — regardless of which exact AI-driven learning platform the student is using. 

That’s important, because AI will only advance classroom learning if educators deeply trust it. We’re aiming to make AI-supported learning more trustworthy for teachers, parents, and administrators, and more effective for students, particularly those who face barriers to engaging with data and data literacy concepts in school.

Looking ahead, what breakthroughs do you think SRI Education could help drive over the next 20 years?

One of the biggest breakthroughs SRI and the broader education field could pursue is realizing the full potential of artificial intelligence to support more personalized learning at scale, while grounding that innovation in strong evidence.

We want everything we do to be evidence-based. We’re not using AI for the sake of using AI. We’re using it because we can observe how it creates real opportunities to tailor instruction to individual students’ needs, interests, and learning trajectories in ways that have historically been difficult to achieve. It can also make life much easier for teachers in areas such as lesson planning, formative assessment, and feedback.

To make AI work for schools, we have to pair AI innovation with rigorous research, accessibility, design, and thoughtful pathways for scaling evidence-based tools.

The way we think about the future is: We are never trying to replace the teacher. We want to make teachers’ jobs easier, because they have some of the hardest jobs in the world. If technology can reduce workloads while also supporting more tailored instruction and allow educators to focus more on relationship-building and high-impact teaching, that’s where AI can play an exciting role.