Formative assessment is the kind that serves learning, not sorting. Its goal is not to put a mark at the bottom of a page, but to make visible, during the session, what has landed and what has not — for you as much as for your participants. Done well, it turns a presentation where you advance blind into a dialogue where you adjust. Let us look concretely at how to set it up, and when it becomes fair to switch to a graded assessment.
Formative or summative: two logics, two moments
The distinction is simple but structuring. Summative assessment comes at the end of the journey: it measures, it grades, it certifies. It is the exam, the final test, the mark out of 20. Formative assessment plays out during learning: it informs without sanctioning. Its verdict is not "you are worth this much", but "here is where the group stands, here is what needs revisiting".
In adult training as in higher education, the two are often confused, and that is where the mood stiffens. As soon as a participant senses that an answer is judging them, they protect themselves: they guess, they go quiet, they copy their neighbour. The strength of the formative is precisely that it removes the stakes of the grade to free up speech. People answer honestly when they know that being wrong is useful, not costly.
Formative assessment does not ask "who got it right?", but "what did not land?".
What the formative really does during a session
Its first role is to give you a real-time dashboard. You think you were clear on a concept; a quick poll shows you that 40 percent of the group is still hesitating. Without that measure, you would have moved on and left a third of the room behind. The formative is the position light that tells you whether to slow down, take another example, or on the contrary speed up because everyone is following.
Its second role is on the learner’s side. Answering a question, even ungraded, forces you to mobilise what you have just heard. It is a solid principle of active learning: you remember better what you had to recall than what you merely listened to. A question asked every twenty minutes is often worth more than a long monologue followed by a final test.
The formative toolkit: quiz, poll, word cloud
You do not need an arsenal. Three formats cover most situations. The ungraded quiz checks factual understanding: three or four multiple-choice questions, answers revealed afterwards, discussion of common mistakes. The live poll probes an opinion or a degree of confidence: "do you feel comfortable with this notion?". The word cloud brings up representations: "in one word, what does this subject evoke for you?".
- The ungraded quiz: check factual knowledge, reveal and discuss mistakes without counting them.
- The live poll: gauge a feeling, a confidence level, a choice, in ten seconds.
- The word cloud: surface the group's real representations and vocabulary.
- The discreet question: slip in an anonymous check without breaking the flow of your talk.
- The self-positioning scale: let each person place where they stand, before and after.
All these activities share the same mechanics in SnapInteract: participants join by scanning a QR code, with no app to install and no account to create, and you watch the answers project live. You will find the full catalogue on the activities page. What matters is not variety for its own sake, but choosing the format that answers your question of the moment.
The golden rule: reveal without sanctioning
The formative only holds if the climate stays safe. A few reflexes preserve it. First, keep answers anonymous while you are in "learning" mode: no one should fear being singled out. Then, reveal the right answers only after the vote, and treat a mistake as valuable information, not as a fault. A wrong answer shared by half the group is not a student problem: it is a signal about your explanation.
Finally, say it out loud. Announcing "this quiz does not count, it is here to see where we stand" changes everything: participants answer truthfully. It is the best way to get a silent room talking, the one where no one dares raise a hand. Anonymity and the absence of stakes dissolve the fear of looking foolish, and you finally harvest honest data instead of polite nods.
Closing the loop: acting on what the measure reveals
A formative assessment that changes nothing in how you run the session is useless. The real value is in the loop: I measure, I observe, I adjust, I recheck. If the poll shows the notion did not land, take it up differently — another example, an analogy, a reformulation by a participant who did understand. Then ask a short question again to confirm the gap is filled.
It is also an excellent tool for spotting disengagement before it sets in. A sudden drop in participation on an activity, increasingly random answers, a confidence scale collapsing after lunch: all of these are weak signals telling you it is time to pause, change the pace, or launch a livelier activity. The formative, well read, keeps you attentive to the room’s temperature.
When to switch to a graded assessment
The formative prepares; it does not replace. At some point you will want to certify learning: an exam mark, an end-of-module test, an assessment that counts. The right moment to switch is when learning has happened and the goal changes: you are no longer trying to correct, you are trying to measure a result. Ideally, the summative should hold no surprises: if your formative assessments have played their part, everyone knows where they stand before the test.
Technically, the switch is simple. Where the formative quiz stays anonymous and stakes-free, the graded assessment quiz assigns a mark out of 20 to each participant, with the safeguards that make the grade credible: a forced pace, delayed reveal, controlled order. And if you grade oral presentations, results relay to dedicated grading via SnapJury. The same tool supports you on both sides: the phase where you learn, then the one where you measure.
A thread running through the whole session
The best formative assessment is not an isolated moment, it is a running thread. An opening question to set the level — that is the role of diagnostic assessment —, two or three checks along the way, a synthesis point at the end. Each is brief, each informs, and together they draw the group’s real progress, not the one you imagined.
You need nothing exceptional to start: a single question asked at the right moment already shifts the dynamic. The simplest thing is to try it on your next session, with your real content and your real group, and see what the room answers when it knows it is allowed to be wrong.
Frequently asked questions
What is the difference between formative and summative assessment?
The formative serves learning: it happens during the session and informs without grading, to spot what did not land. The summative serves certification: it comes at the end and assigns a mark. One adjusts, the other measures a result.
Does a formative assessment have to be graded?
No, and grading it is even counterproductive. The formative works all the better for being stakes-free: ungraded and often anonymous, participants answer honestly instead of protecting themselves. The grade is reserved for summative assessment.
How can I run formative assessment without complicated equipment?
An ungraded quiz, a live poll or a word cloud are enough. With SnapInteract, participants join by scanning a QR code, with no app or account, and you watch the answers project in real time to adjust on the fly.
When should I move from formative to graded?
When learning has happened and the goal becomes measuring a result rather than correcting it. The summative should hold no surprises: if the formative has done its job, everyone knows where they stand before the test.
