Research & Evidence

Built on evidence.
Still pushing.

Tyto's framework is grounded in learning science and validated through federally funded research with thousands of students. Here's what the evidence shows — and what we're investigating next.

Evidence at a glance

Some highlight findings.

01
+20%

Pre/post increase in math performance after gameplay.

Federally funded multi-site trial. Statistically significant at p = 0.01.

+20%beforePreafterPost
IES / U.S. Dept. of Education. WestEd as research partner.
02
96%

Of students wanted to learn more topics with the game.

Not enjoyment. Demand. The same students who said school is boring asked for more.

96%
IES / U.S. Dept. of Education.
03
60%

Of students became more engaged with STEM after playing.

Students reported, after playing, that the experience made them more interested in STEM. Retrospective self-report.

60% of students0%
Overdeck Family Foundation. Cambiar Foundation.
04
0.95

Correlation between game progress and learning growth.

Near-perfect linear relationship at the class level: classes that progressed further through a storyline showed proportionally higher learning growth on that storyline.

Class learning growthClass progress through storyline
IES / U.S. Dept. of Education. Multi-site trial. Aggregated by class × storyline.
collaborative talk
Coordination26%
Science talk24%
Instruction24%
Communication13%
Process7%
Off-task3%
Featured finding · 05
3%

Of student-to-student talk during gameplay was off-task.

Coded transcripts of out-loud conversation between students sitting next to each other — not self-report, not in-game chat.

NSF funded. Coded student-to-student talk during gameplay.
What we're exploring now

The questions we're chasing.

Our research doesn't stop at proving the platform works. These are the open questions we're pushing on now — what's in pilot, what we expect to publish, and where we'd welcome partners.

AI in learning
01

Can an AI conversation partner deepen student reasoning?

We've embedded a "Skeptical Friend" inside gameplay that pushes back on student thinking before group discussion. We're analyzing pilot data to see whether it produces more elaborated reasoning.

In progress
ED / IES funded
Math anxiety
02

Does the game format reduce math anxiety and open up discourse?

Exploring whether the lower-stakes game environment changes classroom culture around sharing rough-draft thinking.

In progress
ED / IES funded
How understanding forms
03

How do students' ideas evolve from rough draft to revised thinking?

Tracking the arc from individual sensemaking → AI-supported elaboration → group discourse → revised understanding. When do conceptual shifts actually happen?

In progress
ED / IES funded
How we build
04

What makes student codesign substantive — not tokenistic?

Studying what experts, teachers, and students each genuinely contribute to the design process — and where each voice changes the product.

In progress
ED / IES funded
Expanding access
05

Can immersive 3D learning work for blind and visually impaired students?

Investigating how to make 3D spaces and game-based inquiry genuinely accessible — not just compliant, but designed for equal learning from the start.

On roadmap
HHS / NIDILRR funded
AI in learning
06

Can AI evaluate the quality of inquiry-based learning content?

Moving past content generation to genuine pedagogical quality — can AI tell strong inquiry design from weak, and support authors who are building it?

On roadmap
Internal R&D
The academic foundations we build on

The principles behind Tyto's framework draw from a body of research on what makes learning games effective. A few highlights from the work that shaped our design.

Embedded Learning

Learning objectives woven into gameplay itself, not layered on as quiz breaks. Students choose to play 7× longer when learning is integrated.

Habgood & Ainsworth, 2011
Problem-Based Learning

Authentic problems activate prior knowledge, improve metacognitive awareness, and support long-term retention and transfer.

Hmelo & Evensen, 2000
Simulations

Simulation-based approaches improve achievement by 23%.

D'Angelo et al., 2014
Immersive Role-Playing

Immersive games show increased academic performance and better transfer than other game types.

Takeuchi & Vaala, 2014
Phenomenon-Based Inquiry

Students need direct experience with the phenomena they're studying to build genuine understanding.

Worth, Duque, & Saltiel, 2009
Curriculum Design

Well-designed curriculum materials make a significant difference for next-generation science learning.

Harris et al., 2014; Clark, Tanner-Smith, & Killingsworth, 2014

These are starting points — our own research program continues to build on and extend this foundation. The full report goes deeper.

Ready to build?

Tyto works with partners to build their first experience.

You don't have to figure it out alone. You're starting a collaboration. We onboard a small cohort each quarter.