Questions about the instrument.

A poll asks cold. We show the costs of both answers first, then measure again, and publish the movement between the two. The gap is the finding.

Both costs get identical weight, identical length, and no colour. Their order alternates between sessions, so neither side owns the first slot.

In the strict sampling sense, often it is not, and we say so beside the number. Our claim is never that humanity said something. It is that agreement held everywhere we looked, with the sample stated.

Referring source is recorded on every response and published with the results. A surge from one source is visible in the published cut rather than hidden inside the headline figure.

Because seeing it would change what you say, and then the number would measure our interface instead of you.

On the facts, we publish what the panel established with sources. On what should be done, no. That is the part we hand to everyone.

Yes. No account, no cookies in the widget, no precise location, no fingerprinting. Country is self-declared and optional.