Checking for understanding — and acting on it

“Does everyone get it?” is a question the class can’t honestly answer — so ask one that shows you every student at once.

The idea

Every student holds a hidden state: got it, shaky, or lost. A good check makes that state visible — for as many students as possible, fast enough to act on today. A weak check reveals only the confident few and lets you feel informed while most of the room stays dark.

Then comes the part people skip: the check is only worth running if it changes what you do next. Same evidence, three moves — move on, reteach the whole class, or split into groups.

How it works

Rank the checks by two things: coverage (how many students you actually see) and timing (can you act on it now). Then apply a simple decision rule.

check                coverage   in-the-moment?   catch
"any questions?"     ~2 / 24    yes              only the vocal; silence != understanding
call on volunteers   ~4 / 24    yes              skews to the students who already got it
cold call (random)   ~4 / 24    yes              unbiased, but still a small sample
all-write whiteboards  24 / 24  yes              everyone answers at once  (best for a live pivot)
exit ticket            24 / 24  no (after class) everyone, but you read it too late to reteach today

decision rule on the TRUTH (not on what the weak check implied):
  ~80%+ solid, few lost .............. move on (spot-fix stragglers)
  solid majority + a real minority ... split: extend the strong, reteach the small group
  half or more not solid ............. reteach whole-class, a different way, then re-check

When to use it

Check	Best when…
All-write whiteboards / mini-slates	You need every hand’s answer right now to decide the next five minutes.
Think-pair-share	The question needs reasoning; talking first raises the quality of what you sample.
Cold call (random, warm tone)	You want an unbiased read without collecting 24 things — name after the question, not before.
Exit ticket	You’re planning tomorrow, not pivoting today — full coverage, delayed.

Watch out for

• Treating silence as a yes. “Any questions?” and thumbs-up sample the confident and the polite. The students who are lost usually don’t know what to ask.
• Volunteers as your gauge. The raised hands are the students who already understand. Reading the room off volunteers reliably overstates it.
• Checking and not acting. If the same lesson continues no matter what the whiteboards show, the check was theatre. The evidence has to be able to change your next move.
• Reteaching the same way. If a third of the class was lost, repeating the identical explanation louder rarely lands. Change the representation — an example, a model, a different entry point.
• Feedback students can’t act on. “Study harder” isn’t formative. Tell them the specific next step: “you flipped the numerator and denominator — try it again with the ratio the other way.”

Worked example

An interviewer running a teaching demo asks: “How would you know if your mini-lesson landed?” A weak answer says “I’d ask if they have questions.” A strong answer names the check and the decision rule: “I’d pose one target question and have all 24 answer on whiteboards, so I see everyone in ten seconds. If about four fifths are solid, I move on and pull the two or three stragglers aside. If it’s a split — most solid, a cluster stuck — I’d set an extension for the solid group and reteach the small group with a different example. If half the room is off, that’s on my explanation, so I reteach the whole class a new way and check again before moving forward.” That answer shows you don’t just collect data — you act on it.