There’s a Yakov Smirnoff joke in there somewhere.
I think the conspiracy theorist nuts would have a ball with these “secret audio signals”. Hope they don’t pick up on it.
Yes, my OP question was based on my local public Classical music station, in their recent pledge drive, claiming that X% of their listening public becoming members.
Wtf? 
I just got my $3 diary a few months ago!
Was it Nielsen for sure or some other group like Arbitron?
Arbitron has been owned by Nielsen for some years now (it’s now “Nielsen Audio”), but it seems that there are some smaller research companies doing radio ratings, as well.
Yes, it was Nielsen.
Yeah, but a self-selecting and paid sample isn’t very representative of the larger population, is it? I know if someone asked me to do that, I’d just be like no thanks, I’m not going to waste my time with your survey. So even if the sample demographically similar, those folks either care more about radio or at last need money more, no?
It’s not a self-selected sample. A self-selected sample is one where the research company simply takes all interested respondents who volunteer, and does nothing to screen them, or balance them for demographics and other characteristics. As I already noted, you cannot actively volunteer to be on the Nielsen panel; they identify a household which they believe meets the characteristics that they need, and recruit you.
But, yes, many people simply don’t want to participate in a market research study. It’s an eternal problem in all market research, and you are correct, we cannot say with absolute certainty that the people who participate are attitudinally similar to those who refuse to, even if they are demographically similar. Yes, they pay respondents to participate, but the fact that they don’t pay very much means that very few respondents are simply doing it for the money.
Market research will never be able to generate a perfect sample of a population or audience, because of “non-response bias” (i.e., not capturing the attitudes or behaviors of the sorts of people who refuse to participate). But, this is a known fact in the industry, and in most cases, it’s not generally believed to be a major problem with the data generated.