What we do

Our research.

We study how young children decide who — and what — to learn from. Two threads run through everything: children's learning with social robots, across both autistic and non-autistic children, and co-design and lived experience research with autistic adults.

Thread 01

Children engaged in a learning activity with a social robot

Children & Social Robots

We've been running child–robot interaction studies around the world since 2015. Our work spans imitation, selective trust, and word learning — and increasingly, what it means for children with different learning profiles.

Our current project, funded by the Jacobs Foundation, is exploring how individual differences in children's social and emotional engagement with robot teachers shape what they learn. We follow the same children across six weekly sessions, looking for patterns in who thrives with a robot teacher and why.

We've also completed a study on autistic children's word learning and engagement in a single-session format, currently in pre-print. The next stage — a dedicated study with autistic children, with co-designed engagement measures built with the autistic community — is what we're seeking funding for now.

Autistic children are welcome in our current study

Thread 02

Adult Autistic Co-Design

Research on autism has too often been done to autistic people, not with them. That's a problem — ethically and scientifically. Research designed without the people it's about tends to ask the wrong questions.

Our co-design work centres autistic adults as experts in their own experience. In collaboration with A/Prof Jessica Paynter, we developed the FOCUS Toolkit — a free resource to make psychological support more accessible and effective for autistic adults, grounded in autistic adults' own ratings of what actually helps.

This thread shapes everything else we do. You can't build better research tools for autistic children without first listening to autistic adults.

Co-design research with autistic adults

Recent work

Just finished.

Robot overimitation

Do preschoolers copy robots more when the robot's inner workings are hidden? Across four boxes, children were more likely to copy unnecessary actions from an opaque robot than a transparent one.

Pre-print coming soon

Novel word learning from social robots by autistic and non-autistic children

Can autistic preschoolers learn new words from a social robot during a shared eBook reading session? Yes — and just as well as their non-autistic peers.

Sommer, Russell, & Paynter — pre-print available

Fear, robots, and learning

Does anxiety about robots affect what children can learn from them? We tested a brief modelling intervention designed to reduce robot-related fear before a word learning session.

Sommer, Jones, & Neumann (2025) — OSF pre-print

New papers

Just published.

2025

Violence against robots: Do humans say no?

Archer, J., Sommer, K., & Wilks, M. — Nature Scientific Reports

DOI
2025

Developmental trajectory of anthropomorphizing robotic agents relative to living and non-living kinds

Sommer, K., Sweezy, S., & Wilks, M. — International Journal of Social Robotics (in production)

DOI

Output

Selected publications.

Open-access wherever possible. Click any title for the DOI. Full list on Google Scholar.

Funding

Supported by.

Jacobs Foundation

CHF 155,000 · 2024–2027. Exploring children's individual differences in learning from, and engagement with, social robots.

Griffith University

AUD 266,928 · 2022–2025. Learning with Robots: Assessing the role of robotics for the future of childhood education.