2016 Postmortem
Related: About this forumMonmouth University Poll: Clinton 45/Sanders 43 in New Hampshire
http://www.monmouth.edu/assets/0/32212254770/32212254991/32212254992/32212254994/32212254995/30064771087/ab385337-2d62-493e-b695-f4350bd8f4d1.pdfYes it's within the margin of error, but look at the trendline. From Aug to Oct, all polls had Sanders ahead; since then Clinton has been ahead in 4 of 7 and 2 of last 3.
sufrommich
(22,871 posts)workinclasszero
(28,270 posts)Bernie is in a fight for his political life in....New Hampshire!!!!
NCTraveler
(30,481 posts)It's a must for Sanders as he has decided to run a smaller campaign. Money isn't needed as bad in such a small state and Sanders is on more equal ground there considering the restrictions he has placed on his campaign. The party will unify around her, not that they haven't already, if she takes NH.
Cosmic Kitten
(3,498 posts)was conducted by telephone from
October 29 to November 1, 2015 with
403 New Hampshire voters likely to vote in
the Democratic presidential primary.
This sample has a margin of error of +4.9 percent.
The poll was conducted by the Monmouth University
Polling Institute in West Long Branch, NJ.
The poll has Hillary gaining with women!
Suprise, NOT.
sample included 173 males, 230 females
LOL, what a pitiful poll.
403 New Hampshire voters likely to vote
What a pitiful sample size!
LULZ
workinclasszero
(28,270 posts)in a scientific poll for it to be valid?
onenote
(42,704 posts)Earlier today you posted to a PPP poll that showed Carson tied with Clinton nationally. The sample size of that poll (representing over 118 million voters, based on 2012 turnout) was around 1300. The NH poll surveyed 403 voters (representing around 403,000 voters, based on 2012 NH Democratic Primary turnout).
I haven't posted the results of any polls because at this stage I don't think they're as meaningful as some would have it. Whether this is a year in which the polls capture where the electorate really is at probably won't be known until the first couple of primaries. But if you're going to tout a poll, don't attack other polls on grounds that the polls you rely on can be attacked as well.
Cosmic Kitten
(3,498 posts)Political polling only serves
to influence perceptions,
not predict outcomes.
Hence, this poll which over samples
females, (about 30% more F than M)
shows Hillary closing the gap with Bernie.
The overall small sample size,
disproportionate sample of women,
and the large margin of error create
an illusion that Hillary is gaining.
It's actually pollsters putting their
thumb on the scale to create a false perception
which the Hillary *fanwagon* touts as a win.
It's transparent manipulation of statistics
under the guise of a *science* used to
distort of reality.
Clinton, on the other hand, has reversed deficits she held among women now leading Sanders 56% - 37% compared to trailing 42% - 47% in September and among voters age 50 and older now leading 56% - 38% compared to trailing 42% - 47% two months ago.
This disparity in results can be
attributed to over sampling women...
Any wonder the poll found she
"reversed deficits she held among women"
As to the change in the "Over 50" group...
There are no crosstabs on age and gender
but it's safe to say the over 50 are
predominately female because the shift
tracks with gender.
The ONLY thing disingenuous in pretending
that Hillary is actually gaining on Bernie.
onenote
(42,704 posts)The exit polls from the 2008 New Hampshire Democratic primary indicate that women voted in disproportionate numbers to men: 57 percent to 43 percent -- which is probably why those percentages were used in the Monmouth poll. http://www.cnn.com/ELECTION/2008/primaries/results/epolls/index.html#NHDEM
Are those numbers going to be correct in 2016 just because that's what they were in 2008? Who knows? But that's my point -- until we actually see some actual voting I don't think we can be certain that past years are accurate predictors of this year -- in ANY poll.
My other point stands: its kinda silly for you to mock a New Hampshire-specific poll for having only 403 participants (representing less than 2 tenths of one percent of the actual voting base) but to post a national poll that had only 1300 participants (representing around 7 one hundredths of one percent of the actual voting base).
mythology
(9,527 posts)A complete and utter lack of knowledge on a subject. That is a completely valid sample size. It makes you look painfully ignorant to dismiss it.
It's your business of course, but nobody remembers the unskewthepolls.com guy as anything but a fool.
JaneyVee
(19,877 posts)MineralMan
(146,317 posts)It's still not a clear and commanding lead, as it is now in Iowa, but it's significant in terms of change.
sufrommich
(22,871 posts)MineralMan
(146,317 posts)Both appear to be about the Monmouth poll, though. I haven't gone to sources yet.
DemocratSinceBirth
(99,710 posts)And if there is a hotly contested GOP race some indys will vote in that race.
All is good.
NurseJackie
(42,862 posts)Cosmic Kitten
(3,498 posts)Clinton, on the other hand, has reversed deficits she held among women now leading Sanders 56% - 37% compared to trailing 42% - 47% in September and among voters age 50 and older now leading 56% - 38% compared to trailing 42% - 47% two months ago.
This disparity in results can be
attributed to over sampling women...
Any wonder the poll found she
"reversed deficits she held among women"
As to the change in the "Over 50" group...
There are no crosstabs on age and gender
but it's safe to say the "over 50" are
predominately female because the shift
tracks with percentage change within gender.
30% more females polled correlates with the
shift in percentages.
Hillary is NOT gaining.
brooklynite
(94,581 posts)A poll is a set of questions asked of a profiled population set. The larger the set (say "women" the more accurate the response will be. However, the broad strokes of the survey will be accurate regardless of the sample size.
When you complain that "of course her share of women went up" because of a larger sample size, you're suggesting that all the additional women survey (presumably intentionally?) supported Clinton. If Sanders was ahead with women in a small sample, he should be ahead with women in a large one.
Cosmic Kitten
(3,498 posts)If women give Hillary a slight advantage
decreasing men and adding women
will NOT result in Bernie remaining ahead
Its blatantly weighting the sample to favor Hillary.
FFS, how about some intellectual honesty.
onenote
(42,704 posts)As I said upthread, I don't know whether or not the 2016 turnout will breakdown along the same lines as the 2008 turnout and that, not some theory that this poll is part of an anti-Bernie "manipulation of statistics" is why I discount it.
It's also why I discount every other poll.