Cognitive Load Theory
We reduce how much a child needs to hold in their head at once, showing key information clearly, breaking tasks into steps, and introducing one new idea at a time.
Grounded in Research
We don't guess at what helps children learn. Here's what shapes our approach, including where the evidence is still developing.
Design principles
How we design learning
Scaffolding
Support is there when it's needed, and steps back as your child grows more confident. Never a fixed, one-size-fits-all level of help.
Metacognition
We build in regular moments for your child to reflect: what worked, what was tricky, what they'd try differently next time. Reflection like this is consistently linked to stronger, more independent learning.
Experiential Learning
Learning sticks best when it's tied to real understanding, not just repetition. We keep curriculum work meaningful rather than mechanical, and we're building toward more real-world application in future updates.
Evidence base
What the wider research tells us
Le Cunff, Giampietro & Dommett (2024), "Neurodiversity and cognitive load in online learning: A systematic review with narrative synthesis," Educational Research Review.
Read the original research →Why it matters: The strongest bridge between online learning design and neurodivergent needs found so far. It shows most online learning research overlooks neurodivergent learners, and highlights specific findings like subtitles/redundant information helping some ADHD learners even where it doesn't help neurotypical learners.
Shahini et al. (2025), "A systematic review for artificial intelligence-driven assistive technologies to support children with neurodevelopmental disorders," Information Fusion.
Read the original research →Why it matters: Synthesises 84 studies (2018–2024) across autism, ADHD and dyslexia. Frames AI as supporting professional and parental judgement rather than replacing it, with strong emphasis on privacy and real-world deployment challenges.
Valverde Olivares et al. (2025), "XR Technologies in Inclusive Education for Neurodivergent Children: A Systematic Review 2020–2024," Children (Basel).
Read the original research →Why it matters: Not specific to Key Stage 2, but strong evidence that controlled, customisable digital environments can improve engagement and task performance for neurodivergent children.
Caveat: The studies reviewed use varied methods, and the field still needs more standardised ways of measuring outcomes.
Pi et al. (2021), "Meta-Analysis of RCTs of Technology-Assisted Parent-Mediated Interventions for Children with ASD," Journal of Autism and Developmental Disorders.
Read the original research →Why it matters: A careful, credible study. It found some tech-assisted interventions improve specific outcomes (like emotion recognition), but not broader language or social outcomes.
Caveat: We share this precisely because it shows the evidence is mixed and specific, not because we're claiming Starview replicates these results. We'd rather be upfront about what's proven versus what's promising.
Le Cunff et al. (2024), "Neurodiversity and cognitive load in online learning: A focus group study," PLOS ONE.
Read the original research →Why it matters: Useful for understanding what creates cognitive overload and what helps reduce it.
Caveat: Based on university students, not Key Stage 2 learners. We use it to inform design principles like predictability and reduced sensory load, not as direct evidence for our specific age group.
Starview is an early-stage platform. We take a cautious, evidence-informed approach rather than promising more than current research supports, and we'll keep updating this page as our own pilot data grows.