Most digital citizenship teaching is a list: do not share your password, think before you post, tell an adult. Students learn the list, repeat it back accurately, and then meet a situation the list does not cover — which is nearly all of them — and have nothing to reason with. Scenarios work better, but only if they are written carefully, and the most common way they are written quietly turns them into something else.
Why Rules Transfer Badly
A rule is a compressed decision. Somebody thought a situation through and stored the conclusion. That is efficient right up until the situation differs from the one it was compressed from, at which point the student has the conclusion and none of the reasoning that produced it.
The visible symptom is a class that can recite the principle and cannot apply it. They are not being careless. They were given an answer and never shown the working.
Two Question Shapes, Two Different Skills
There is an important distinction that most scenario materials collapse. "What would you do?" and "what is actually wrong here?" measure different things, and one of them is far more commonly asked.
The situation is described, the problem is clear, and the student picks a course of action. This measures conduct selection, and it is the right shape for questions about how to treat people.
The situation is described with nothing flagged, and the question is what the real problem is, with four candidate problems offered. This measures noticing — and it is the harder skill, because in life nothing arrives labelled.
The reason it matters is that a response item usually contains its own answer. If your stem says the app "wants far more than it needs", you have identified the issue inside the question and are now only testing whether the student can pick the sensible reply.
Writing a Scenario That Cannot Be Gamed
The failure mode is uniform keyed answers. If the right answer is always the cautious one, students find that within about six questions, and from then on the assessment measures pattern-matching.
- Vary what the keyed answer looks like — sometimes act, sometimes wait, sometimes name the problem, sometimes hand it on.
- Make the distractors real issues that are wrong for this case, not obviously silly options.
- Keep option lengths comparable, because "pick the longest one" is a strategy students discover without being taught it.
- Do not let one register dominate: if telling an adult is right every time, it is being rewarded as a pattern rather than as a judgement.
That last point is worth checking rather than assuming. Count how often each answer shape is the keyed one across the whole set — if any shape is right much more often than chance would give it, the set has a shortcut in it.
Reading a Result Without Turning It Into a Grade
The output worth having is not a number. It is the list of which situations were read differently and why, because that list is a lesson plan somebody else already wrote for you.
It also matters that the bands are set against the content rather than against other students. A four-option instrument gives about a quarter of the marks to anyone clicking blindly, so any band structure where a chance score comes back sounding acceptable is telling students something untrue about themselves.
What a Score Cannot Tell You
A digital citizenship result describes twenty-odd hypothetical situations. It does not describe a student's life, it cannot see what happens in a group chat you are not in, and it should never be read as an indication that a particular child is at risk or is a risk.
The Digital Ethics & Citizenship test is built on the two-shape approach above — ten identification items, fourteen response items, three separately scored areas, and bands written against the content — and its result is deliberately shaped as a list of situations to discuss rather than a verdict about a person.