Formative vs Summative Assessment: When to Use Each
Formative assessment tells you what to teach next. Summative assessment tells you what was learned. Here's how to use both — and stop blurring the line.
Most teachers can define formative vs summative assessment in their sleep — formative is for learning, summative is of learning. The harder question is when you're actually doing each, because in a busy classroom the two get blurred constantly. Here's how to tell them apart and how to design each one so it does its job.
The one-sentence test
If the result of the assessment changes how you teach tomorrow, it's formative. If it goes into the gradebook and doesn't change instruction, it's summative. A weekly "quiz" that lives in the gradebook but doesn't inform the next lesson is just a small summative — which is fine, but don't expect it to do formative work.
Formative assessment: low-stakes, high-frequency
Formative checks should be cheap to make, cheap to grade, and tied to a specific learning objective. Some examples:
- Exit tickets — 1–3 questions on today's objective
- Whiteboard checks — every student answers, you scan the room
- Targeted 5-question quizzes on a single skill
- Cold-call sequences with progressively harder questions
The point isn't the score — it's the misconception. A well-designed formative assessment has wrong answers that map cleanly to specific misunderstandings, so when you see "60% of students chose answer C," you know exactly what to reteach.
Summative assessment: stakes, breadth, finality
Summative assessments measure what students learned over a longer arc — a unit, a quarter, a year. Because the stakes are higher, they need to cover the full domain proportionally, give students enough room to demonstrate understanding (mixed item types, not just MCQ), and be calibrated against the standards.
- Unit exams — mixed MCQ, short answer, and one extended response
- End-of-term performance tasks
- State or AP exams
The common mistake: summative dressed as formative
A teacher gives a weekly chapter test, records it, moves on. The students don't get feedback in time to change anything. The data isn't disaggregated by skill. Nothing changes about instruction. That's a summative — labeling it "formative" doesn't make it one.
Fix: either (a) shorten it to 3–5 items per objective and act on the data within 48 hours, or (b) accept it as summative and design the unit exam to actually cover the full standard.
A weekly cadence that works
- Daily: 1 exit ticket per class
- Weekly: 1 targeted 5-question quiz on the week's focus skill
- Per unit (~3 weeks): 1 summative exam, mixed format
That gives you 5 formative checks per quiz, 15 per unit, and only one heavyweight assessment — enough data to teach responsively without drowning students in tests. Whatever tool you use (FoxFish or otherwise), pick one cadence and stick with it; the diagnostic power comes from comparing the same kind of data over time, not from a clever one-off.