§2.4 · Research thread · active · proprietary
BigKids
Kansas City · Independent ·
Abstract
A maths and English tutor for young children that reasons before it drills, and picks the next exercise with care. The result is public in relative terms: which of the lesson planner's mistakes fell, by about how much, and which did not. The measured figures are in the full results paper, shared on request. How a learner is modelled and how the next exercise is chosen are not published. All of it is on simulated learners; no child's learning has been measured.
1 · Article2 · PaperRequest access
Keywords: learning, children, evaluation.
1Measurements
| measure | value | conditions | source |
|---|---|---|---|
| Lessons given before the learner was ready | ≈ −75% | on simulated learners | private research log: status notes |
| Exercises repeated too soon | ≈ −25% | on simulated learners | private research log: status notes |
| Combined score | below zero → above | the planner's mistakes and aims in one number | private research log: status notes |
| Share of reasoning exercises | ≈ 10% | unchanged, and lower than the product intends | private research log: status notes |
§Where the work lives
§Related writing and lessons
- Introducing WordStash: Your Personal Vocabulary Companion (writing)
- What is Retrieval Hit-Rate? (writing)
Ask this thread