Active recall — what the studies actually show. The evidence list below names the studies that actually constrain closed-book prompts.
Scores here are relative ranks inside this atlas, not lab grades or official cut scores.
Evidence
- Classical source — Karpicke is the public name this atlas attaches to the origin story of testing effect.
- Modern constraint — Roediger and later work keep testing effect from being treated as magic: effort and test still matter.
- Exam listing — testing effect earns its exam-fit score when the sitting asks for a name, number or chosen method under time.
- Dunlosky utility — Dunlosky's reviews treat practice testing and distributed practice as high-utility — testing effect is one way to run them.
- Bjork difficulty — Bjork's desirable difficulty says a slightly harder retrieve of testing effect beats a fluent reread.
Mechanisms
- Retrieval as encoding — Producing testing effect from a blank page is a new encoding, not a photocopy of the reread.
- Error as signal — A miss on testing effect tells you which fact to split or rewrite. A fluent reread hides that.
- Time-box honesty — testing effect that only happens after midnight is a scheduling fail, not a character fail.
- Transfer test — If you cannot do testing effect in a different room, you have context, not skill.
Researchers
- Karpicke — Karpicke is cited as a public scientific name for testing effect, not as a private coach.
- Roediger — Roediger appears because later measurement tightened the first story of testing effect.
- John Dunlosky — Dunlosky's utility rankings are the classroom constraint: testing effect should earn its hours against cheaper rereading.
- Robert Bjork — Desirable difficulty explains why testing effect that feels slightly late is often the useful one.
Educational atlas — not an official syllabus, medical advice, or leaked item bank. Active recall: Keep the public outline in view and treat this page as method, not a leaked key.
ScienceMethodDrillsFast passPitfallsToolsStoriesLocale playbook