.......Which brings us to Donald Trump. There’s no question he’s polling better than any other GOP candidate, and by a mile.
Now, the very fact that all the polls agree on this is probably enough to make it clear he has the lead at the moment. As it happens, that’s not worth very much — at the end of August 2011, Rick Perry had about the same numbers Trump does now, and in 2007 Rudy Giuliani was doing better than Trump is today.
But what if all the polls are junk? What if, in the desperate act of just trying to get a sample of any significant size, pollsters throw out elementary logic in reporting the results?
What if what they’re polling does not actually reflect the true nature of those who will vote in Republican primaries and caucuses?
Forget the rhetorical questions. These polls are junk, and they are not reflective. To take the most recent example, CNN trumpeted a poll showing Trump with 24 percent of GOP voters nationwide. The problem is that 52 percent of those surveyed by CNN said they’re going to vote in the GOP primaries.
That’s insane. The overall electorate is about 225 million. In 2012, 20 million voted in the GOP primaries. That’s 10 percent, not 52 percent.
A few days earlier, CNN did a poll of Iowa voters showing Trump in the lead — and even more outrageously, the poll’s findings suggested that 62 percent of Iowa’s population would attend the January caucuses. Oh? In 2012, a whopping 2.5 percent of Iowa’s adult population turned out. CNN’s poll is literally off by a factor of 25.
CNN isn’t the only offender by a long stretch. An Aug. 2 NBC/Wall Street Journal poll featured 1,000 adults, of whom 252 were said to be likely Republican primary voters. That’s 25 percent of all Americans — 2 ¹/₂ times the size of the 2012 GOP electorate.
The same poll sought to break down the preferences of those 252 voters and report them out as though they were statistically reliable as relates to the Republican Party — when a 250-person nationwide sample is far too tiny to be representative. As a DC expert said to me yesterday, “These polls are either way too big or way too small.”
We’re flying blind here, people. Nobody knows what’s really going on.
http://nypost.com/2015/08/18/trumps-in-the-lead-but-these-polls-are-junk/
Now, the very fact that all the polls agree on this is probably enough to make it clear he has the lead at the moment. As it happens, that’s not worth very much — at the end of August 2011, Rick Perry had about the same numbers Trump does now, and in 2007 Rudy Giuliani was doing better than Trump is today.
But what if all the polls are junk? What if, in the desperate act of just trying to get a sample of any significant size, pollsters throw out elementary logic in reporting the results?
What if what they’re polling does not actually reflect the true nature of those who will vote in Republican primaries and caucuses?
Forget the rhetorical questions. These polls are junk, and they are not reflective. To take the most recent example, CNN trumpeted a poll showing Trump with 24 percent of GOP voters nationwide. The problem is that 52 percent of those surveyed by CNN said they’re going to vote in the GOP primaries.
That’s insane. The overall electorate is about 225 million. In 2012, 20 million voted in the GOP primaries. That’s 10 percent, not 52 percent.
A few days earlier, CNN did a poll of Iowa voters showing Trump in the lead — and even more outrageously, the poll’s findings suggested that 62 percent of Iowa’s population would attend the January caucuses. Oh? In 2012, a whopping 2.5 percent of Iowa’s adult population turned out. CNN’s poll is literally off by a factor of 25.
CNN isn’t the only offender by a long stretch. An Aug. 2 NBC/Wall Street Journal poll featured 1,000 adults, of whom 252 were said to be likely Republican primary voters. That’s 25 percent of all Americans — 2 ¹/₂ times the size of the 2012 GOP electorate.
The same poll sought to break down the preferences of those 252 voters and report them out as though they were statistically reliable as relates to the Republican Party — when a 250-person nationwide sample is far too tiny to be representative. As a DC expert said to me yesterday, “These polls are either way too big or way too small.”
We’re flying blind here, people. Nobody knows what’s really going on.
http://nypost.com/2015/08/18/trumps-in-the-lead-but-these-polls-are-junk/
