Every teacher checks for understanding—an exit ticket, a quick oral probe, a homework pattern. The problem is rarely collecting the signal. It is what happens to the signal afterwards: too often, nothing. The next lesson was already planned, the evidence sits in a markbook, and the moment where it could have changed something passes.

Formative assessment only earns its name when it is formative—when the result shapes the next teaching move.

Three reasons checks fail to form anything

  1. The evidence has no memory. A misconception spotted in October is invisible by December, even if it never resolved.
  2. The evidence is too coarse. A class average of 68% does not say which concept, which prerequisite, or which learners.
  3. The next step is already fixed. When tomorrow’s lesson is locked, evidence can at best trigger a vague “review”—not a targeted response.

What concept-level evidence makes possible

When each question is mapped to the concept it assesses, a five-minute check stops being a score and becomes a position on the curriculum map. Two contrasting questions can distinguish a persistent misconception from a careless slip. Patterns across weeks can show whether delayed recall is holding up or quietly decaying.

That is the difference between “the class struggled with fractions” and “equivalence is secure, but comparison with unlike denominators is uncertain for a third of the class—and it traces back to one shaky prerequisite.”

The second statement suggests a specific, cheap intervention. The first suggests worry.

Keeping the cost near zero

The common objection is time. If formative assessment means extra marking and data entry, it will not survive a busy term. The design goal we hold ourselves to in the Cuegence pilot is five minutes or less of additional administration per class per day—using evidence that normal academic activity already produces, and feeding it into the longitudinal learner model rather than a spreadsheet.

The loop, not the event

A single check is an event. Formative assessment that works is a loop: check, interpret, act, and then confirm from the next evidence whether the action helped. When the loop closes, teachers stop guessing whether last week’s intervention landed—and school leaders gain evidence-backed visibility without surveillance.

That loop is what we are building, and what pilot schools help us measure. If your school wants to evaluate it in real classrooms, join the pilot.