Every teacher has heard the definitions: formative assessment checks understanding along the way, summative assessment measures it at the end. The definitions are easy. Using the distinction well in an actual classroom is where it gets harder.
The real difference isn't the format, it's the purpose
A quiz can be formative or summative depending on what you do with it. If it's used to decide who needs reteaching before moving on, it's formative, even if it's graded. If it's used as a final judgment on mastery that goes into the report card with no plan to revisit it, it's summative. The same format serving different purposes is a normal and useful thing — the mistake is treating every assessment as summative by default.
Formative assessment should change what you do next class
If a formative check doesn't change your next lesson at all — if you're planning to move forward at the same pace regardless of what it shows — it's not really functioning as formative assessment, whatever it's called. The entire value of formative assessment is the pivot it enables: reteach, regroup, move on. If there's no mechanism to act on the results, it's just an extra grade.
Low-stakes doesn't mean no feedback
A common mistake is treating "formative" as synonymous with "doesn't matter," which leads to skipping feedback on it entirely. Students still need to know what a formative check revealed about their understanding — the low stakes are about the grade, not about whether the information matters to the student.
Summative assessment should reflect the same skills you actually taught
The biggest credibility risk with summative assessment is a mismatch between what was taught and what's tested — new terminology, an unfamiliar format, a jump in difficulty nobody was warned about. If formative checks throughout the unit used a similar format to the eventual summative assessment, there are no surprises, and the summative grade reflects learning rather than test-format literacy.
A practical ratio to aim for
There's no universal rule, but a unit with frequent, low-stakes formative checks and one clear summative assessment at the end tends to produce both better learning outcomes and less anxiety than a unit with only a handful of high-stakes tests. If you're not sure whether you have enough formative checkpoints, a rough test is: could you predict, with reasonable confidence, how each student will do on the summative assessment before they take it? If not, you likely need more formative data along the way.
Used well, the two aren't competing categories — formative assessment is what makes the summative result a fair and unsurprising reflection of what was actually taught.