That’s how random samples work. A small number like that is enough to get a decent estimate of the entire population. Give ar take a percentage point if three. That is, if the sample is actually random. That’s actually near impossible to guarantee, as people who pick up their phones (or answer their door, or walk in the street and allow folks to ask them a question, or answer a letter) are different from those who don’t.

The content is still visible though (at least on my end)