Small sample size this time (504 with a 4.4% MOE). But that’s still large enough to paint a pretty good picture.
Yea I think the sample size was smaller because it was not just registered Republicans— but rather it was people who watched the debate and are gonna vote in the Republican primaries..
True. :-)
Trunp is between 32 and 40. Carson is between 8 and 16. Fiorina is between 6 and 14. I think Trump’s lead is outside the MOE
All the way back to Monday the media was excited about Carson's closing the gap on Trump. Funny, that poll's 6% MOE didn't concern them.
Since Trump garners a lot of his support from self-identified indy Republican leaners who can vote in many states' open primaries, watch for the GOP-e pollsters to sample only registered Republicans during Carly-flavor-of-the-week-Fiorina's rise.
Not to bore everyone, but the media is intent on concealing that no point in a confidence interval is any more likely than any other point. So, 20% support with a MOE of +/-4% means it is just as likley 16% - or 24% - as it is the 20% midpoint. The midpoint is stated merely for convenience.
The media's illiteracy is revealed when they say, "Candidate X rose from 20% last week to 22% this week."
Uhhh, no they didn't. Or, if they did, statistical methods cannot make such a determination (short of an impractically huge sample size).
Watch for the sample population trick I warned about above.