Weird. You ran a test, got an unexpected result and decided that the sample most be wrong? I mean, you *could* be right but you don't *know* you are and have determined the data is flawed based upon your prediction rather than vice versa.
i mean you know i ask a ton of polls right? my followers are generally white 20-35 year old men in STEM, trending more libertarian and pro sex work. I know this and I know it would be surprising if this *didn't* have an impact on my poll results.
Self selecting.
Try pulling out the hashtag #gold: cuts across much; used in many context, cultures, finance to slum.
Though none found will give a damn.
But how? These questions would startle and trigger any representative sample of the population. Any answer is somehow already contained within the question and in the medium though which it’s asked.
Great questions! They do be making me think.