Discussion:
Amazing, they conduct polls using a pool of 1000-3000 voters
(too old to reply)
PeteWasLucky
2024-07-28 05:32:49 UTC
Permalink
https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/
--
----Android NewsGroup Reader----
https://piaohong.s3-us-west-2.amazonaws.com/usenet/index.html
jdeluise
2024-07-28 06:41:45 UTC
Permalink
Post by PeteWasLucky
https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/
Is that any different from the norm? I'd say trust no polls.
Scall5
2024-07-28 07:16:08 UTC
Permalink
Post by PeteWasLucky
https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/
Is that any different from the norm?  I'd say trust no polls.
Polls are even less trust-worthy these days as most people don't take a
call from an unknown phone number...
--
---------------
Scall5
Pelle Svanslös
2024-07-28 07:22:53 UTC
Permalink
Post by Scall5
Post by PeteWasLucky
https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/
Is that any different from the norm?  I'd say trust no polls.
Polls are even less trust-worthy these days as most people don't take a
call from an unknown phone number...
Not to mention the dilemma of the deplorable. "Can't come out of the
closet!"
--
"And off they went, from here to there,
The bear, the bear, and the maiden fair"
-- Traditional
The Iceberg
2024-07-28 08:59:52 UTC
Permalink
Post by Pelle Svanslös
Post by Scall5
Post by PeteWasLucky
https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/
Is that any different from the norm?  I'd say trust no polls.
Polls are even less trust-worthy these days as most people don't take
a call from an unknown phone number...
Not to mention the dilemma of the deplorable. "Can't come out of the
closet!"
well you leftists are so violent and over-react on absolutely anything
that disagrees with your wrong opinion.
Pelle Svanslös
2024-07-28 07:00:19 UTC
Permalink
Post by PeteWasLucky
https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/
What exactly is amazing? Polls have always been conducted by randomly
sampling a subset of the population. Nothing new here. A sample size of
a 1000 gives roughly a 3% margin for error. It is in fact amazing how
fast these things converge with so little. A 10k sample size would give
you approximately 1% margin for error. But that would cost 10x more, and
there are other sources of uncertainty.

Polls are polls.
--
"And off they went, from here to there,
The bear, the bear, and the maiden fair"
-- Traditional
The Iceberg
2024-07-28 09:01:06 UTC
Permalink
Post by Pelle Svanslös
Post by PeteWasLucky
https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/
What exactly is amazing? Polls have always been conducted by randomly
sampling a subset of the population. Nothing new here. A sample size of
a 1000 gives roughly a 3% margin for error. It is in fact amazing how
fast these things converge with so little. A 10k sample size would give
you approximately 1% margin for error. But that would cost 10x more, and
there are other sources of uncertainty.
Polls are polls.
except the poll's audience were previous Democrat voters by a big
majority, dimvit.
PeteWasLucky
2024-07-28 16:40:44 UTC
Permalink
On 28.7.2024 8.32, PeteWasLucky wrote:> https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/What exactly is amazing? Polls have always been conducted by randomly sampling a subset of the population. Nothing new here. A sample size of a 1000 gives roughly a 3% margin for error. It is in fact amazing how fast these things converge with so little. A 10k sample size would give you approximately 1% margin for error. But that would cost 10x more, and there are other sources of uncertainty.Polls are polls.-- "And off they went, from here to there,The bear, the bear, and the maiden fair"-- Traditional
Polls can be much trustworthy if they care to have them more reflective to an actual election day.

"Caldwell said there are also behavioral differences between Democrats and Republicans. Democrats are more likely to participate in a poll but less likely to vote. Republicans are less likely to respond to a survey and more likely to vote. Distinctions like these add to the complexity of nailing down accuracy. "
--
----Android NewsGroup Reader----
https://piaohong.s3-us-west-2.amazonaws.com/usenet/index.html
Sawfish
2024-07-28 16:55:30 UTC
Permalink
Post by PeteWasLucky
On 28.7.2024 8.32, PeteWasLucky wrote:>https://www.forbes.com/sites/saradorn/2024/07/27/trump-vs-harris-2024-polls-trump-holds-slight-lead-over-harris-in-harrisxforbes-poll/What exactly is amazing? Polls have always been conducted by randomly sampling a subset of the population. Nothing new here. A sample size of a 1000 gives roughly a 3% margin for error. It is in fact amazing how fast these things converge with so little. A 10k sample size would give you approximately 1% margin for error. But that would cost 10x more, and there are other sources of uncertainty.Polls are polls.-- "And off they went, from here to there,The bear, the bear, and the maiden fair"-- Traditional
Polls can be much trustworthy if they care to have them more reflective to an actual election day.
"Caldwell said there are also behavioral differences between Democrats and Republicans. Democrats are more likely to participate in a poll but less likely to vote. Republicans are less likely to respond to a survey and more likely to vote. Distinctions like these add to the complexity of nailing down accuracy."
Add to this is that there is a giant demand for polls, so that the
results can be manipulated and/or presented, and of course this means
there's money to be made.

The devil is in the methodology, and few, very few, are capable of
unwinding these methodologies so as to really and truly *understand* the
basis on which the polls are conducted and evaluated.

In this regard they are a sort of specialized study. Both polls and
studies are there to be purchased by interested parties who want to add
credibility to their claims. No less a public figure than Winston
Churchill was claimed to have said: "I trust only statistics that I have
falsified, myself."

It's up to us to figure out, which, or even *if*, studies and polls are
to be trusted, and to what degree.

See why one is best off relying on personal experience?  :^)
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"If we use Occam's Razor, whose razor will *he* use?" --Sawfish
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Pelle Svanslös
2024-07-28 17:06:31 UTC
Permalink
"Distinctions like these add to the complexity of nailing down accuracy."
Like I said, there's other sources of uncertainty besides sampling
errors. There's nothing amazing about sample sizes of 1k, however.

Hope this will be another election with a good turnout. And hope Twump
will again be left scratching his head. "How can I lose when I get more
votes than when I won?! Stop the steaw!"
--
"And off they went, from here to there,
The bear, the bear, and the maiden fair"
-- Traditional
Loading...