Lensman
Well-Known Member
Is TouringPlans generating population-level estimates by weighting responses to correct for differences between their sample and population? Or are they just publishing the raw results from their self-selected survey respondents?Small, sure. The trick is for the sample to be representative. I can't imagine that TP has that -- too much self-selection.
ETA: Just saw another poster said the same. While TP's sample might be better than most companies', I still doubt that you could accurately extrapolate.
Is TP calculating survey error at all?
Besides any more subtle sampling errors caused by self-selection, the big self-selection bias is respondents will tend to be people who have strong opinions about the matter being surveyed - so you'll tend to be surveying the people who either strongly liked or strongly disliked something while finding out little about the people in the middle.