Formative assessment examples that show who needs help
By Volodymyr Shpodarunok · Updated 18 October 2026
The short answer
Good formative assessment is short, low-stakes and tied to one skill per question: an exit ticket of two or three questions, a hinge question mid-lesson whose wrong answers reveal a specific misconception, or a short check every few lessons. What makes any of them formative is that the next lesson changes because of the answers.
Five techniques at a glance
| Technique | When | Time | What it tells you |
|---|---|---|---|
| Exit ticket | Last 3 minutes of a lesson | 3 min | Who got today's skill, who did not |
| Hinge question | At the turning point of a lesson | 1–2 min | Whether the class can move on, and which misconception is in the room |
| Short check | Every 2–3 lessons | 10 min | What was learned and then forgotten |
| Mini whiteboards | During teaching | Seconds | A whole-class snapshot, with no record |
| Confidence rating | With any of the above | Seconds | Who is wrong and sure, the hardest to reteach |
Exit tickets
Two or three questions at the end of a lesson, each on one skill taught that day. Sort the answers into three piles as students leave: got it, partly, not yet. The not-yet pile is your first five minutes tomorrow. Keep exit tickets on today's skill only; they are not a review.
A German lesson on the accusative, for example: three sentences to complete with the right article. Students who write "der" for a masculine direct object are still using the nominative; that is one quick reteach for whoever is in the not-yet pile, not a new lesson for the whole class. Illustrative.
Hinge questions
Dylan Wiliam describes the hinge question: one multiple-choice question at the point where the lesson can go either way, where each wrong option matches a known misconception. A physics example: in a vacuum, a hammer and a feather are dropped at the same moment; which lands first? Options: the hammer, the feather, both together. Students who choose the hammer still believe heavier objects fall faster. (Astronaut David Scott ran exactly this test on the Moon in 1971.)
The answers take seconds to read. If most choose "both together", move on. If a third choose the hammer, you know exactly which idea to address before continuing.
Short checks every few lessons
Eight to twelve questions across the skills of the unit so far. They catch what was learned and then forgotten, which exit tickets miss. Retrieval also strengthens memory on its own (Roediger & Karpicke), so the check teaches while it measures. Read them by skill, not by total score.
Mini whiteboards and confidence ratings
Mini whiteboards give an instant read of the whole room but leave no record, so use them during teaching rather than as evidence. Adding a confidence rating to any question ("how sure are you, 1 to 3?") separates students who guessed from students who are confidently wrong. The second group holds a misconception and needs the most careful reteach.
Common mistakes
- Asking "any questions?" and taking silence as understanding. Ask a question every student has to answer.
- Exit tickets that test five skills at once, so the piles cannot be sorted.
- Hinge questions whose wrong options are random rather than matched to a misconception.
- Collecting answers and never looking at them before the next lesson.
- Grading every check, so students stop answering honestly and start guessing carefully.
What makes it formative
Research on formative assessment reports typical effect sizes of 0.4 to 0.7 when teachers adjust teaching to what the checks show (Black & Wiliam). A check that is marked and filed is not formative; a check that changes tomorrow's lesson is. The technique matters less than whether anyone acts on it.
Turning answers into a list
The bottleneck is reading every answer by skill, every few lessons, for every class. That is the part Gaplight does from a school's own checks: one list per teacher of who needs them and on which skill, grouped so one short lesson reaches everyone with the same gap.
Before you write a hinge question
- Name the one idea the lesson turns on.
- List the two or three wrong ways students usually think about it.
- Write one option for each wrong way, plus the right answer.
- Decide in advance what you will do if more than a quarter choose each wrong option.
Volodymyr Shpodarunok
Co-founder and CEO, Gaplight
Has spent two years building and measuring a learning engine used daily by students preparing for Ukraine’s national exam, and now works with schools on reading their own quiz results.
Sources
- Wiliam, Embedded Formative Assessment, Solution Tree (2011)
- Black & Wiliam, Inside the Black Box (1998)
- Roediger & Karpicke, Test-Enhanced Learning, Psychological Science (2006)
Examples marked as illustrative are made up to show the method; the history example comes from Gaplight’s own course data.
More guides
- How to identify learning gaps in students →
- Item analysis: what to do when a class averages 68% on a quiz →
- How to use assessment data to drive instruction →
- Reteaching strategies: whole class, small group, or one student →
- Common algebra misconceptions, and the wrong answers that reveal them →
- Gaplight and Khan Academy: what each one does, and how they fit together →