Method
The learning design
Daily Sharpen is built on well-established learning principles, applied with restraint and measured honestly.
- Retrieval with real answers: Learners type numerical or algebraic answers rather than guessing from multiple choice, strengthening recall mechanics.
- Teach before drilling: New skills begin with worked examples. Supported success is recorded as learning, but independent evidence is required for mastery.
- Correction and re-entry: A missed question requires entering the correct answer immediately, followed by a related question later to verify error repair.
- Accuracy before speed: Speed targets apply only once accuracy is established; correct-but-slow answers are never marked wrong, but they do trigger faster routing feedback.
- Deterministic question generation: Questions are generated by code that constructs the answer first, ensuring canonical correctness without runtime language models.