The Learning Lab for schools

At institution scale the Learning Lab stops being a per-learner profile and becomes a curriculum instrument: it tells you which of seven competences your cohort is thin on, which is a teaching decision rather than a pastoral one.

The same seven facets

Meaning, character, pinyin, tone, listening, speaking and writing — the same split a learner sees about themselves, read across a class or a year group.

The decks are deliberately the same shape for learners, tutors and institutions. There is no separate institutional metric, because a different metric for administrators is how a platform ends up reporting one thing to a school and another to the student.

Reading it at cohort scale

1. Find the facet everyone is thin on — If most of a year group is hollow on Tone, that is a curriculum gap and a scheme-of-work change — not thirty conversations.
2. Check whether it is evidence or ability — Each facet names the activity that fills it. A cohort with no Speaking data has probably never been set speaking work, which is a very different finding.
3. Compare classes, not students — Two classes at the same level with different shapes usually differ in what was set rather than in who is in them.
4. Use Errors for the shared mistake — Grouped mistakes across a cohort is the cheapest curriculum feedback available — it names the thing worth reteaching.

Two limits you will hit

Neither is a bug and both are worth understanding before a senior leader asks.

Group figures need five learners. Medians and cohort comparisons are withheld below that, because an average plus one known score identifies the rest. A class of three shows individual rows and counts, but no statistics.

Facet access follows the learner’s plan. A student on a free account exposes three facets, not seven, whatever the institution is paying. Institution access buys seats and capacity; it does not buy a deeper read of an individual’s profile than that individual has.

What it never contains

No text anyone typed. The Lab is built from outcomes — right, wrong, how long, how confident — and no endpoint behind it returns free text, for a teacher, an admin or anyone else.

And the staff view is built from the learner’s own panels, so there is nothing on an institutional screen a student cannot see about themselves. That is a sentence worth being able to say to a parent.

Getting data into it

The Lab is evidence-driven, so a newly onboarded school sees very little for the first weeks. Nothing is wrong — there is simply nothing to report yet.

The fastest way to a useful picture is variety rather than volume: set work that touches reading, listening, writing and speaking in the first fortnight. A cohort that has only done reading produces one facet and six blanks.