4 ms·
And <any kind of science or pseudoscience> != <scientific objectivity>. The idea of scientific objectivity is "forming conclusions based only on independently t
by jrajav 14y ago
And <any kind of science or pseudoscience> != <scientific objectivity>. The idea of scientific objectivity is "forming conclusions based only on independently testable measurements." That seems applicable here.
- 001sky 14y agopolitical polling data is not independently testable
- spamizbad 14y agoThere's actually a fairly major independent test that will be concluded on Tuesday, November 6th. ... but you're technically correct :)
- Evbn 14y agoWhat are elections?
- arrrg 14y agoHuh? If you know the methodology it’s very simple to independently test polling data.
- pjscott 14y agoIf political polling data gives you information about the outcome of an election -- if it can make your probability estimate more accurate than that of someone who hasn't seen any such data -- then I don't see what's so all-important about independent testability. Evidence is evidence; the possibility that a poll is skewed merely weakens its evidentiary value, by increasing the probability that the poll would have those results if they're not true.