The timing problem

Summative assessment data arrives
when it can't change anything.

The fundamental problem with summative assessment as the primary data source for teaching improvement is timing. When a student scores 54% on the end-of-unit test, what should the teacher do? The unit is over. The class has moved on. The teacher can note it as a planning consideration for next year — but they cannot act on it this year, for this cohort.

Most curriculum assessment is designed to produce this kind of data: accurate, comparable, delayed. Useful for reporting. Useless for iteration.

Agile-compatible curriculum assessment is designed for the opposite: less accuracy, less comparability, but sufficient timeliness to produce action within the current teaching cycle. If the curriculum plan only schedules assessment at the end of units, teachers who want to iterate mid-unit are improvising assessment tools that weren't designed for the context. If the curriculum plan includes mid-unit formative checkpoints with clear action protocols, agile teaching is structurally supported.

Three levels of curriculum assessment

End-of-lesson. Mid-module.
End-of-unit.

Agile-compatible curriculum operates assessment at three timescales simultaneously. Each level serves a different purpose and produces data that feeds a different part of the agile loop.

Level
Timing · Format
Action window · Who uses it
End-of-lesson
After every lesson
Same day — next lesson · Individual teacher
Mid-module
After the first 40–50% of a module
1–3 days — within the module · Teacher, sometimes department
End-of-unit
After the full unit
Planning for next cohort, or supplementary teaching in current one · Teacher and school leadership

The end-of-lesson level is covered in C3. This article focuses on the mid-module level — the least common and most valuable for in-curriculum agile teaching, because it produces data at the moment when the teacher still has time to change what happens within the current module.

Designing the mid-module check

What it looks like and where
in the plan it sits.

1
Design 1 — The application task, not a test
A task that requires transfer, not recall

The mid-module check should not be a quiz or a test. Tests produce grades, not diagnostic signals. The mid-module check should be a task that requires students to apply the concepts covered in the first half of the module to a novel scenario — one that wasn't covered in those lessons. If students can complete it, the first half of the module has produced transferable understanding. If they cannot, specific gaps can be identified before the module's second half.

Design constraint
The task should be completable in 15–20 minutes. Longer than this and it functions as a summative assessment in disguise. Shorter and it doesn't produce enough information to be actionable.
2
Design 2 — Placed at a natural conceptual boundary
After the concept is established, before it is applied at depth

The mid-module check sits between the 'introduction and first examples' phase and the 'application and extension' phase of the module. This placement means: if the check reveals a significant gap, the teacher can address it before the application phase, rather than discovering it at the end when the gap has compounded through applied work.

Why placement matters
A mid-module check placed at 40% through the module gives the teacher 60% of the module still to act. A check placed at 80% gives only 20%. The earlier the check, the larger the action window — but the check can only be useful if students have had enough exposure to the concepts to reveal genuine gaps.
3
Design 3 — Accompanied by a decision protocol
The check produces an action, not just a score

The curriculum plan should include, alongside the mid-module check, a decision protocol: 'If fewer than 60% of students complete Q2 correctly, insert consolidation lesson X before proceeding to Lesson N. If more than 80% complete all three questions correctly, skip consolidation lesson and proceed directly to extension activity.' This protocol is the link between assessment and adaptation.

Why the protocol matters
Without a written protocol, teachers make ad-hoc decisions under pressure. With a written protocol, the decision is made at design time — when the curriculum author has the time and distance to think clearly. The teacher's job becomes data collection and protocol execution, not decision-making under time pressure.
AI-generated mid-module checks

Combining-concepts tasks
in 30 seconds.

🤖SprintUp Education's exit quiz tool for mid-module checks
SprintUp Education's exit quiz tool can generate mid-module checks as well as end-of-lesson checks — the format is identical (recall, understanding, application), but the application question uses a scenario that requires combining concepts from the first half of the module rather than a single lesson. Input the module's objectives so far — get a calibrated mid-module check in 30 seconds. Free on every school account.