[ericpauley]

> so according to you paying to pretty much-untargeted people to answer questions is better than getting some of the relevant people to fill out the survey? Interesting.

Could you address this feedback directly instead of making a straw man? How do you prevent non-response bias?

[BartBoch]

By targeting responders as precise as possible so they answer because they want to, not because they need to.

[ericpauley]

That doesn't answer the question. Responders who see the idea as valuable are more likely to answer, which could skew your data. (this is more or less the definition of non-response bias) What is your solution to this? Original response proposed a pre-selected population who have incentive to respond either way.

[BartBoch]

By going this logic, there is not a thing called neutral or unbiased? Don't you agree?