Getting a representative cross sample of Likely Republican Caucus Goers cannot be an easy thing to do.
A purely random survey would have to cover at least half the precincts and start with a call list of AT LEAST 5,000.
The question that needs to be answered is...where did you get the list of 574?
Repeat from another thread:
Iowa will probably have 150,000 - 200,000 GOP caucus-goers. More than half, up to 70%, are regulars who have attended a number of caucuses over the years, and the party knows them by name and address.
‘Random’ polls rely on voluntary answers from a random population, not from known names and addresses. The problem with these polls should be obvious.
Margin of errors are at least twice what pollsters will admit. And that’s for honest polls, not push polls, which, of course, show up on demand to fit a MSM meme.
They said their responses were all landline or cell phone, iirc.
They weighted this poll more heavily toward known Carson supporters than they did the last poll. One can’t help but be suspicious of that.