We want every product decision to have a reason behind it. That means looking at established research, learning from experienced practitioners and testing the finished experience with the people expected to use it.
This page records what informs our work. It will grow as erniee brings in relevant expertise and learns from children, families and schools.
Research on retrieval practice and spacing suggests that recalling information and revisiting it over time can support longer-term retention. The useful question for erniee is how to apply those ideas without turning learning into a repetitive test.
How this may affect the products
What remains to test
Spellify uses scheduled review and mastery states, but the educational effect has not yet been independently evaluated with children.
Personalisation should make the experience more usable and the support more relevant. It should not reduce a child to a label or isolate them from the social parts of learning.
How this may affect the products
Digital play can support creativity, autonomy, competence and relationships. The effect depends on the design. Rewards, pressure and social comparison can change how the experience feels and why a child continues.
How this may affect the products
Screen use cannot be judged by minutes alone. The child’s age, the activity, the context and what the experience displaces all matter. erniee will focus on the quality and purpose of the activity, then design a clear way to finish.
How this may affect the products
What remains to test
The best way to help children stop and move to another activity is still an open design and research question. erniee should test specific approaches rather than make a broad claim about screen time or the brain.
Technology can increase the amount and quality of practice when it is used for a clear learning purpose. It should also connect with teaching, conversation and activity away from the device.
How this may affect the products
The ICO Children’s Code says the best interests of the child should be a primary consideration in services likely to be accessed by children. It also calls for high privacy by default, data minimisation and care around design techniques that influence children’s choices.
How this may affect the products
Each record shows the question, the source, what it suggests, the product decision it informs, where that decision stands and what we still need to find out. Every public claim is checked against the source itself before it appears here.
How should a child return to words they are learning?
| Source | The science of effective learning with spacing and retrieval practiceShana K. Carpenter, Steven C. Pan and Andrew C. Butler, Nature Reviews Psychology, August 2022 |
|---|---|
| What it suggests | A review of research finding that spreading practice over time and recalling information, rather than re-reading it, tends to improve learning across ages and subjects. It is a general review of how learning works and does not evaluate any children’s app. |
| Product decision | Spellify schedules review of each word over time and asks the child to recall it in several different games. |
| Status | Implemented |
| Open question | Whether children using Spellify actually retain and use their words over time. This has not yet been evaluated with children. |
What does privacy by design mean in a children’s app?
| Source | Age appropriate design: a code of practice for online services (the Children’s Code)Information Commissioner’s Office, In force from September 2020 |
|---|---|
| What it suggests | A statutory code of 15 standards for online services likely to be used by children, including the child’s best interests, high privacy by default, data minimisation and no nudges towards weaker privacy. It covers data and design, not the quality of learning. |
| Product decision | Collect only what a defined service needs, set high privacy by default and keep commercial messages out of the child’s experience. |
| Status | Planned |
| Open question | A completed data map and DPIA for Spellify before release, and privacy explanations tested with children and adults. |
How can digital play support a child’s wellbeing?
| Source | RITEC Design ToolboxUNICEF Responsible Innovation in Technology for Children, with the LEGO Group, November 2024 |
|---|---|
| What it suggests | A framework of eight wellbeing dimensions, including autonomy, competence, emotions, relationships and creativity, developed with children in 18 countries. It is written for game designers and shows that play can support wellbeing when designed well, not that it will. It says nothing about learning outcomes. |
| Product decision | Review rewards, choices and progress mechanics against the wellbeing dimensions, and avoid lost streaks and public rankings. |
| Status | Planned |
| Open question | How children themselves experience Spellify’s rewards and choices, tested with children and relevant specialists. |
When does technology genuinely improve learning?
| Source | Using Digital Technology to Improve LearningEducation Endowment Foundation, March 2019 |
|---|---|
| What it suggests | Guidance for schools suggesting technology helps when it supports sound teaching, for example more or better practice and clearer feedback, rather than being introduced for its own sake. It does not cover screen time. |
| Product decision | Every activity has to earn its place by improving practice or feedback on the child’s own words. |
| Status | Planned |
| Open question | Whether teachers see Spellify as adding to their teaching, tested in the school pilot. |
How can an app help children understand their own learning?
| Source | Metacognition and Self-Regulated LearningEducation Endowment Foundation, Second edition, November 2025 |
|---|---|
| What it suggests | Guidance recommending that children are explicitly taught to plan, monitor and evaluate their learning, ideally within subject teaching. The evidence comes from classrooms; applying it to apps is our inference. |
| Product decision | Feedback explains what to try next, and progress is shown to the child in a way that helps them see what is secure and what needs another go. |
| Status | Planned |
| Open question | Which kinds of feedback and progress views help children reflect, and which simply become scores. |
What role should technology play in education?
| Source | Global Education Monitoring Report 2023: Technology in education: A tool on whose terms?UNESCO, July 2023 |
|---|---|
| What it suggests | A global policy review arguing that education technology should serve learners’ interests and support, not replace, teaching. It finds the evidence of impact mixed. It is not evidence for or against any particular app. |
| Product decision | Design for use alongside teachers and families, with the same words at school and at home. |
| Status | Planned |
| Open question | The simplest reliable way for a teacher and a family to share one list, defined and tested in the first programme. |
How much screen time is right, and how should an activity end?
| Source | Screen use by children aged 5 to 16: call for evidenceDepartment for Education and Department of Health and Social Care, June 2026 |
|---|---|
| What it suggests | A government call for evidence on how screen use affects children’s health, wellbeing, behaviour and learning, feeding guidance for parents expected in autumn 2026. It is a process in progress and establishes no findings yet. |
| Product decision | Give activities a recognisable end and never use loss or urgency to pull a child back. |
| Status | Planned |
| Open question | What the resulting guidance says, and which ways of finishing an activity work best for children, tested in the pilot. |
| Field | Required content |
|---|---|
| Question | The product or child question being considered |
| Source | Named paper, guidance body, date and link |
| What it suggests | A short, qualified explanation in plain language |
| Product decision | The feature or behaviour influenced by the source |
| Status | Planned, implemented, tested or independently evaluated |
| Open question | What erniee still needs to find out |
