Interact
You have prepared a solid module. But prepared for whom, exactly? A group discovering the topic and a group that has already practised it do not need the same course. Diagnostic assessment is the act of measuring your participants' knowledge and mental models before you dive in. Five well-placed minutes that spare you from re-explaining what the room already knows, or from skimming over what is sorely missing.
Diagnostic, formative, summative: don't mix them up
Diagnostic assessment sits right at the start, before learning begins. Its purpose is not to grade but to map: what do your participants already know, what do they imagine, where are their blind spots? It differs from formative assessment, which supports learning along the way, and from summative assessment, which validates achievement at the end.
The classic trap is skipping this step for lack of time. In fact it is often the opposite: ten minutes of diagnosis save you an hour of miscalibrated teaching. You arrive with a plan; the diagnosis tells you which parts to speed up and which to deepen.
A good diagnosis does not judge the group. It tunes the course to the group.
The diagnostic quiz, ungraded and low-stakes
The most direct format is the opening quiz. A handful of multiple-choice questions on the module’s key ideas, projected on screen, that everyone answers from their phone by scanning a QR code. The golden rule: state clearly that it is not graded. You are not trying to trap anyone, you are trying to see. With nothing at stake, participants answer honestly, and you get a truthful snapshot of the level.
In practice, show the distribution of answers after each question. If 80% of the room got it right, you know you can move fast. If answers scatter, you have found your point of focus. To build these questions, our article on the assessment quiz explains how to phrase items that truly reveal what people understand, rather than what they guess.
- Three to five questions at most: the diagnosis must stay light.
- Questions that target prerequisites and misconceptions, not trivia.
- A projection of aggregated results, never an individual score on display.
- A word to reassure: 'This isn't a grade, it's my compass.'
The cloud of initial mental models
Even before testing precise knowledge, it is valuable to surface mental models. Ask an open question such as ‘When I say artificial intelligence, what word comes to mind?’ and let everyone answer in a word or two. The answers form a cloud projected in real time, where the most-cited terms grow larger.
This cloud shows you at a glance the group’s vocabulary, its fixations, its misunderstandings. It also creates a shared starting point: you can build on the words that appear to introduce your content. We explain how to make the most of this format in our guide to the word cloud, including how to draw out a quiet group at the start of a session.
A group's first few words say more than ten minutes of presentation.
The confidence scale to gauge how people feel
A participant can know a concept without feeling able to use it. The self-positioning scale measures exactly that feeling: ‘On a scale of 1 to 5, how comfortable do you feel with this subject?’ Everyone places themselves, and the spread tells you whether you are facing intimidated beginners, false experts, or a mixed group.
This format is especially useful in adult training, where the gap between the real level and the felt level is common. Crossed with a quiz, the scale reveals things: a group that answers correctly but reports low confidence mostly needs reassurance and hands-on practice. Our article on the self-positioning scale covers the wordings that work.
Adapting your course in real time, without redoing everything
A diagnosis is useless if you don’t act on it. The idea is not to rewrite your module live, but to pull a few simple levers: skip a section already mastered, slow down on a sticking point, swap an example for one closer to the group. With SnapInteract, your whole session lives in a single window — slides, the 14 interactive activities, projected timer and participant tracking — which makes these adjustments natural rather than acrobatic.
You drive all of this from your Mac, optionally with the iPhone or Watch remote, while participants follow on their own screen. And because your data stays on your Mac — we keep nothing on our servers — a slightly personal diagnosis never drifts into a third-party cloud. If you want to see how these building blocks fit together across a whole session, take a look at our account of a lively session end to end.
- Spot the two or three sections to speed up or deepen.
- Keep the diagnosis in mind: it justifies your choices to the group.
- Run a mini check at the end to measure the distance travelled.
- Reuse the same quiz at the start and the end to make progress tangible.
Three formats, one intention
Ungraded quiz, cloud of mental models, confidence scale: these three formats serve the same intention, which is to start where the group truly is, not where your plan imagines it. You can chain them in ten minutes, or keep just one depending on the time available. The essential thing is to get into the habit of looking before teaching.
It is also a way to open with participation rather than a monologue: the group engages from the very first minute, and you finally know who you are talking to. None of this requires any install on the participant side, nor an account to create — a QR code is enough. The best time to try this approach is your next module.
Frequently asked questions
Frequently asked questions
Should a diagnostic assessment be graded?
No, definitely not. Its purpose is to measure a starting point, not to sanction. A grade would discourage honesty and skew your reading. State clearly that it doesn't count: you'll get far more truthful answers.
How long should it take at the start of a module?
Five to ten minutes is plenty. Three to five quiz questions, or a word cloud, or a confidence scale: the goal is a quick snapshot, not an exam. That time is quickly repaid by a better-calibrated course.
How can a group take part with nothing installed?
With SnapInteract, participants join by scanning a QR code from their phone, with no app and no account. You project the question, they answer, results appear live. Your data stays on your Mac and we keep nothing on our servers.
Can you reuse the same diagnosis at the end of a module?
Yes, and it's an excellent idea. Asking the same questions again at the end makes progress visible, for you and for the group. It's a simple, concrete way to show the distance travelled during the session.
