[user104658]

To be fair I can appreciate that open surveys can be skewed by responder bias that's a pretty basic principle of social research, but that's why you conduct RANDOM opinion surveys in the first place, so that you get a broad spectrum of general viewpoints and not one group organising to affect the poll. There are clear and established ways to minimise the effect of responder bias. So if they're choosing not to, that's when you need to ask questions.