The case in three lines. The Green Gazette runs an online poll on closing the skate park at eight, and prints THE GREEN SAYS YES: 115 of 140 votes, 82%. The website's records show that 96 of the votes came through a link shared in one residents' group set up to complain about the noise, and among everyone else the vote was about half and half. The skaters, almost all under sixteen, were never sent the poll, and Tara has to write the corrections box.

One question to ask. "What should the box say?" Then: "Is the neighbour wrong about the noise?" Let them argue both sides.

One thing to do at the table. Think of a poll, a review score or a petition you have seen online. Ask who was likely to see it, who would bother to answer, and who would never have known it existed.

The grown-up layer. The case is about self-selection: when people choose whether to answer, the answers show who cared enough to answer, not what everyone thinks. It gives the young reporter her strongest point, that every vote was real and anyone could have voted. It gives the neighbour his: the noise late at night is real, and he has a right to sleep. The habits it models are finding out where the answers came from, counting the groups apart, and asking who never had the chance.

The case leaves the line blank on purpose. Ask for it before you offer yours.