Posted on 05/20/2016 8:36:12 AM PDT by Sean_Anthony
. . . Trump is gaining, and 52 percent of Democrats are dissatisfied with the choices
If you were excited about yesterdays Fox poll showing Trump up 45-42, and thus crestfallen to see this one, dont feel that way at all. For one thing, Fox and Rasmussen polls always show the Republican running stronger than polls put out by MSM sources. You want to believe a liberal outlet like CBS is tilting the scales by oversampling Democrats, and I absolutely believed they were doing that in 2012 to make Obama look stronger than he really was.
On Election Night, I kicked myself for believing that and starting wrestling with the possibility that maybe we had become a center-left nation after all.
Internals?
There are other ways to over-sample in a particular candidate’s favor. If a candidate is strongly represented by females, then over-sampling females makes a difference.
We think that only repubs and democrats make a difference. Not so. Over sampling minorities would make a difference. It doesn’t have to be much if a little bit takes place in a number of areas.
Almost all polls are guilty of landline/cell phone fixation, with ignoring non-response rates, and with repeat calling.
Here we go again. Where is the skewedpolls.com idiot? Probably hanging out with Presidents McCain and Romney.
and of course there’s no way to poll people who are so pissed off that they won’t respond to pollster solicitation
This poll is skewed. In the middle of the data there is a question asking people to describe themselves as dem. or rep. The poll is skewed 8% toward dem. in this “self reporting” question. Only half of those polled were registered dems or reps, and those numbers were fairly even. But in the other half, they obviously oversampled dems greatly, probably by up to 16 to 18% in order to end up with a TOTAL sample that described itself dem by an 8% margin. Here is the link to the data if anyone wants to look to over.
https://www.scribd.com/doc/313214968/New-York-Times-CBS-Poll-May-19th-National
Millenials and left-leaning “moderates” are outright pissed off about Hillary’s theft of the nomination. My bet is that if there’s isn’t a riot at the Democrat convention, Trump will absolutely dominate the election in November. Trump vs. Hillary would go 44-6, IMO.
A month from now at least some polls will have Trump up 5 OUTSIDE of the MOE. Her fall will be THAT quick and dramatic.
Trump will Kick her ASS!
At the rate it is going she will be down double digits to Trump by the Dem convention this summer.
I want the Dems to think she has a chance to win this fall.
“I absolutely believed they were doing that in 2012 to make Obama look stronger than he really wa
Obama did win.
The day after Trump won big in Indiana CNN came out with a poll showing Hillary Clinton by 12. Now we have this poll following two poles 1 by Rasmussen and one by Fox News Channel showing Trump leading Clinton in both by five and three points respectively.
The CNN poll was obviously an outlier and I believe the reason for this Poll as well as the CNN poll was to have an effect on the RCP average of polls.
Psyops
I see a lot of articles referring to oversampling of Democrats. Rush even talked about it a lot during the last 2 presidential elections. But if there are more Democrat voters in the country than Republicans, wouldn’t it make sense to poll more Dems than Repubs?
Hillary is going to collapse big time!
Exactly. If they don’t respond, then you’re getting nothing from entire parts of society.
See my tag.
refusing to respond to a pollster is not the same thing as refusing to vote
Yes and no. Doing a poll of all people living in Berkely, CA would show an overwhelming majority for HRC, while polling only those in Hereford, Texas would show just the opposite. The trick with a poll like this is to try to achieve a sample that is reflective of the actual likely electorate in Nov—those who will go to the polls, with a correct balance of R, D, and I, i.e.,reflective of the percentages of those in the population, while attempting to determine what the preferences are within those groups, and how likely they are to vote and vote those preferences in November.
So, not every poll showing a dem winning are necessary flawed—could be accurate. But the poll in question came to the implicit conclusion that those listed as neither D nor R were in fact 18% more likely to “identify” as dem when pushed—that’s the only way to account for the fact that overall the poll showed an 8% dem. identification preference notwithstanding that of registered voters, the poll sampled an almost equal number of dems and reps. This finding seems highly suspect.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.