Mastery, spaced review, and regression

A skill is marked Mastered when your student answers correctly several times across multiple sessions. That's harder than just one good attempt - we want to see they can do it on different days, in different contexts.

Only answers they give on their own count. A right answer after a hint, or a partly-right answer, doesn't add to the count or reset it. Where a skill can be answered by typing, at least one right answer has to be typed rather than picked from a list.

Skills are taught one part at a time, and a skill being taught in parts is only marked Mastered once every part has a right answer on its own. Skipping never counts against a skill.

Once mastered, the skill goes into spaced review. We bring it back periodically to keep it sharp - first a few days later, then a week, then a fortnight. The intervals get longer as evidence stacks up.

Reviews open each session, taking up to about a quarter of the session time. When several are due, the ones overdue longest go first, then the ones with the shortest gaps so far.

When your student answers a skill right in the starting check, we don't teach it again. It shows as Already knew rather than Mastered, because one right answer on one day isn't the evidence mastery needs.

A quick review comes up a few days later. If they get it right, the skill becomes Mastered and is counted with the rest. If they get it wrong, it stays in review like any other skill and is checked again in three days. Lessons that build on it go ahead either way.

If your student gets a review question wrong, the skill stays in review rather than going back to the start. We check it again in three days and its review gaps start again from the short end, growing as they get it right.

If other work shows a mastered skill has slipped before its review is due - usually a homework answer that uses the skill comes back wrong - we flag it as a regression and add a lesson to re-teach it. You'll see it in the Attention section on your dashboard and as a notification.

Regressions aren't bad. They're how the system catches things before they become real gaps. A skill might regress because the student was tired, the question was harder, or the concept needs revisiting - we don't try to diagnose; we just bring it back.

Forgetting is normal. Spaced repetition is one of the most well-evidenced learning techniques in cognitive science. We do the bookkeeping so your student doesn't have to.