CDZ Comparing the Polls; Why Are Monmouth and Rasmussen so far Apart?

[320 Years of History]

the "Unskewed Polls" guy from 2012, who tweaked all polls by creating a pretend world where the # of Republican likely voters is the same # as the # of Democratic likely voters

...And insofar as the # of Republican likely voters is not the same # as the # of Democratic likely voters, that exercise was a mildly interesting academic pursuit, but that's all it was.

Hypothesis contrary to fact

 

[JimBowie1958]

The alienation phenomenon is not unique to individuals who at the polling moment are inclined to prefer Trump. It is something that has been found with regard to anyone whose preferred candidate is, in their mind, disliked by the majority of or many other voters. There is no particular party or candidate that is associated with that respondent behavior.

Yeah, normally it is factor out on both sides, but when you have one candidate that is vilified to a great extent the way Trump is or the Brexit supporters, then it is overwhelming in its unbalance and tilts things from the shamed candidate and their supporters. Else how do you explain the consistent 5% under estimation of Trumps vote through the primaries?

Off Topic:
Frankly, I think the phenomenon itself, and what motivates individuals to behave as it indicates, is more interesting than is what it portends for how one should construe poll results.

In what way is it more interesting?

Red:
That stance suggests that the quantity of remarks and discussion is, to poll respondents, is more relevant than the central points of what is so frequently said. Now, that may be so for some folks, but I had rather believe that hearing something more often does not in one's mind make it any more so or make one any more despondent over hearing it or with regard to the object of the remarks/poll inquiry.

Purple:
I don't know there is a 5% underestimation of Trump's support among the electorate. I will know after votes have been cast and tallied.

Blue:
It is more interesting in that inquiry into and discovery of human behavior and impetus is more interesting than is how those behaviors and inspirations affect a set of specific political outcomes at a specific point in time. I feel that way because I think humanity itself, not its impacts, is the most interesting thing in our world.

The "some folks" are about 5 to 10% of the voting population, which is my point.

The primaries had about a 5% underestimation of Trumps final vote in most states, particularly the open primary states. The "Brexit" factor might have around an 8% hidden vote for Trump, but it may be redundant with what is already seen in the primaries.

And yes, I agree that people are interesting, particularly with my Aspergers, but I still like numbers better.

 

[JimBowie1958]

the "Unskewed Polls" guy from 2012, who tweaked all polls by creating a pretend world where the # of Republican likely voters is the same # as the # of Democratic likely voters

...And insofar as the # of Republican likely voters is not the same # as the # of Democratic likely voters, that exercise was a mildly interesting academic pursuit, but that's all it was.

Hypothesis contrary to fact

According to Pew, we have 46% of registered voters are Democrats and 40% are Republicans.

The polls generally run much higher representation of Democrats to Republicans based on the 2012 Romney performance.

But Donald Trump is not Mittens Romney.

 

[GaryDog]

the "Unskewed Polls" guy from 2012, who tweaked all polls by creating a pretend world where the # of Republican likely voters is the same # as the # of Democratic likely voters

...And insofar as the # of Republican likely voters is not the same # as the # of Democratic likely voters, that exercise was a mildly interesting academic pursuit, but that's all it was.

Hypothesis contrary to fact

According to Pew, we have 46% of registered voters are Democrats and 40% are Republicans.

The polls generally run much higher representation of Democrats to Republicans based on the 2012 Romney performance.

But Donald Trump is not Mittens Romney.

Indeed, he's FAR more unpopular with Republicans. He may well lose Georgia.

 

[320 Years of History]

The alienation phenomenon is not unique to individuals who at the polling moment are inclined to prefer Trump. It is something that has been found with regard to anyone whose preferred candidate is, in their mind, disliked by the majority of or many other voters. There is no particular party or candidate that is associated with that respondent behavior.

Yeah, normally it is factor out on both sides, but when you have one candidate that is vilified to a great extent the way Trump is or the Brexit supporters, then it is overwhelming in its unbalance and tilts things from the shamed candidate and their supporters. Else how do you explain the consistent 5% under estimation of Trumps vote through the primaries?

Off Topic:
Frankly, I think the phenomenon itself, and what motivates individuals to behave as it indicates, is more interesting than is what it portends for how one should construe poll results.

In what way is it more interesting?

Red:
That stance suggests that the quantity of remarks and discussion is, to poll respondents, is more relevant than the central points of what is so frequently said. Now, that may be so for some folks, but I had rather believe that hearing something more often does not in one's mind make it any more so or make one any more despondent over hearing it or with regard to the object of the remarks/poll inquiry.

Purple:
I don't know there is a 5% underestimation of Trump's support among the electorate. I will know after votes have been cast and tallied.

Blue:
It is more interesting in that inquiry into and discovery of human behavior and impetus is more interesting than is how those behaviors and inspirations affect a set of specific political outcomes at a specific point in time. I feel that way because I think humanity itself, not its impacts, is the most interesting thing in our world.

The "some folks" are about 5 to 10% of the voting population, which is my point.

The primaries had about a 5% underestimation of Trumps final vote in most states, particularly the open primary states. The "Brexit" factor might have around an 8% hidden vote for Trump, but it may be redundant with what is already seen in the primaries.

And yes, I agree that people are interesting, particularly with my Aspergers, but I still like numbers better.

Pink:
And my point is that until votes are cast and tallied, the accuracy of your claim about the extent to which "some folks" constitute 5% to 10% of voters or poll respondents is indeterminate. You, anyone even, can certainly speculate and argue that to be so, as can others, other who are particularly credible thus making the POV at least worth considering. If in thinking as you have noted in "pink" assuages concerns you have about the election's outcome, fine; do so if it allows you sleep well or something.

For my part, I can wait 21 days (perhaps longer if election results are contested, as I suspect they will be all over the nation) to find out whether in hindsight it is worth examining the measures of error given in the polls and by doubters of their predicted outcome(s). Quite simply, I don't now care much what polls indicate for I have no wagers on the outcome.

Brown:
The primary votes were those cast by the most fervent of Republicans, and in the few open primaries, any voter in a given state. We are now considering the general election, not the primaries.

 

[320 Years of History]

the "Unskewed Polls" guy from 2012, who tweaked all polls by creating a pretend world where the # of Republican likely voters is the same # as the # of Democratic likely voters

...And insofar as the # of Republican likely voters is not the same # as the # of Democratic likely voters, that exercise was a mildly interesting academic pursuit, but that's all it was.

Hypothesis contrary to fact

According to Pew, we have 46% of registered voters are Democrats and 40% are Republicans.

The polls generally run much higher representation of Democrats to Republicans based on the 2012 Romney performance.

But Donald Trump is not Mittens Romney.

Indeed, he's FAR more unpopular with Republicans. He may well lose Georgia.

A Texas conservative with whom I was speaking today mentioned that he fears Trump may in fact lose Texas. That strikes me as unlikely, but, hey, stranger things -- things such as grown men unbidden groping women by their genitals and then bragging about having done so -- have happened.

 

[Toro]

Who considers that to be the case?

538 has done a statistical analysis of thousands of polls and came to that conclusion.

 

[Picaro]

It was interesting that the Brexit polls showed Brexit losing by 3-4%. Funny.....but it won by 3-4%. It is very difficult for pollsters to measure voter intensity.....(i.e. Who's voters will actually show up).

I think Hillary is probably up right now...but generally I am distrustful of the polls and think the race is much closer than the polls are indicating.

A better indicator is who will be able to get their voters to the polls. Trump rallies have people lined up for blocks to get in, while Hillary has to bribe schools to bus a hundred kids in for an audience. Democrats have turned off a big chunk of their base with their blatant dishonesty, so while polling may show close numbers, the ballots cast will definitely show a very different story, and that is what's different about this election. Most of the 'experts' have been wrong for many months about Trump's chances, and they still are, otherwise these sociopaths and gimps wouldn't be screaming and wetting themselves so much over it, and this with the spending so lop-sided in favor of Hillary to boot. The stench of corruption surrounding the Democrats is too strong to overcome with sniveling and lying this time around.

And again they can only record answers from people willing to answer them or give them the time of day.

 

[Picaro]

the "Unskewed Polls" guy from 2012, who tweaked all polls by creating a pretend world where the # of Republican likely voters is the same # as the # of Democratic likely voters

...And insofar as the # of Republican likely voters is not the same # as the # of Democratic likely voters, that exercise was a mildly interesting academic pursuit, but that's all it was.

Hypothesis contrary to fact

According to Pew, we have 46% of registered voters are Democrats and 40% are Republicans.

The polls generally run much higher representation of Democrats to Republicans based on the 2012 Romney performance.

But Donald Trump is not Mittens Romney.

Indeed, he's FAR more unpopular with Republicans. He may well lose Georgia.

A Texas conservative with whom I was speaking today mentioned that he fears Trump may in fact lose Texas. That strikes me as unlikely, but, hey, stranger things -- things such as grown men unbidden groping women by their genitals and then bragging about having done so -- have happened.

Yes, another rumor sent out to mislead and confuse. lol it's ridiculous.

 

[JimBowie1958]

Pink:
And my point is that until votes are cast and tallied, the accuracy of your claim about the extent to which "some folks" constitute 5% to 10% of voters or poll respondents is indeterminate.

Lol, thank you Captain Obvious, lol.

Of course it is speculation, what else would it be BEFORE THE ELECTION?

 

[JimBowie1958]

Trump Leads Clinton By 1 Point Going Into Debate — IBD/TIPP Poll

After more than a week of blistering attacks from Democrats, celebrities and the press, Donald Trump has managed to pull ahead of Hillary Clinton by a 1.3 percentage point margin — 41.3% to 40% — in a four-way matchup, according to the new IBD/TIPP poll released today.

And this poll was the most accurate of the 2012 election.

http://fivethirtyeight.blogs.nytime...and-worst-in-the-2012-presidential-race/?_r=0

It is interesting that the phone polls were generally more accurate than internet polls in that election. It seems to not be the case based on the primary results this year.

Trump is changing so many things about our election system , it is amazing.

 

[JimBowie1958]

A better indicator is who will be able to get their voters to the polls. Trump rallies have people lined up for blocks to get in, while Hillary has to bribe schools to bus a hundred kids in for an audience. Democrats have turned off a big chunk of their base with their blatant dishonesty, so while polling may show close numbers, the ballots cast will definitely show a very different story, and that is what's different about this election. Most of the 'experts' have been wrong for many months about Trump's chances, and they still are, otherwise these sociopaths and gimps wouldn't be screaming and wetting themselves so much over it, and this with the spending so lop-sided in favor of Hillary to boot. The stench of corruption surrounding the Democrats is too strong to overcome with sniveling and lying this time around.

And again they can only record answers from people willing to answer them or give them the time of day.

Excellent points, but I guess we will have to see November 9th.

 

[JimBowie1958]

Who considers that to be the case?

538 has done a statistical analysis of thousands of polls and came to that conclusion.

 

[Picaro]

Who considers that to be the case?

538 has done a statistical analysis of thousands of polls and came to that conclusion.

I linked to it, I think, at least somewhere on the board, anyway. On edit: post # 8; it should link to his pollster ratings on that page as well.

 

[320 Years of History]

Pink:
And my point is that until votes are cast and tallied, the accuracy of your claim about the extent to which "some folks" constitute 5% to 10% of voters or poll respondents is indeterminate.

Lol, thank you Captain Obvious, lol.

View attachment 94472

Of course it is speculation, what else would it be BEFORE THE ELECTION?

Well, what else is there to say that is worth saying this close to the election? The line of discussion you've presented with your pic and related comment reminds of thoughts I had about my students when I was a teaching/graduate assistant in college. In the period a week to week and half before I gave an exam -- bearing in mind I gave an exam every four weeks -- sure, I thought about what concepts and techniques the students were grasping and with which of them they struggled, and I structured my lectures to help reinforce and focus in their minds those things which I could tell gave them fits.

At some point, however, whether they got "this or that" concept ceased to be a concern that I could take an extra ten minutes here or five minutes there and expound upon, for there was other material that I had to address prior to the test and whatever multiply-covered material they'd not grasped, well, they were on their own to figure it out somehow before sitting for the test. Do that as they might, I was beyond the point of there being anything I could do about it, other than for individual students who came to my office hours for specific and individualized guidance.

FWIW, on for my final exams, I, of my own volition, offered exam reviews that I held outside of class periods. How many people roughly on average showed up? Ten, out of a class of ~150 students. And how many students earned "As" in the whole class? For the class in question, eight. (And what a PITA that was because I had to present my lecture outlines to the faculty coordinator to justify why I had given fewer than 15 "As." Had I known what that process entailed, I quite likely then -- not now -- would have found a way to bump five to eight of my "eighty-nine point 'whatever'" students to a 90.)

The point of the preceding anecdote is that at some point, there's no point in speculating about how an event will turn out. Things at that point are what they are and will be as they will be. There's no point in "wishing on stars" and looking for, believing in or acting upon that which one perceives may alter events or one's predicted outcomes about them.

As go the polls about the 2016 Presidential election, we have reached that point, regardless of what the polls predict, and we are at that point because there is no foreseen catastrophe coming down the pike and because the characterizations of themselves the two major candidates have given the American people are not going to in 20 days dramatically change, which is what they would need to do...In short, Trump and Mrs. Clinton would need in 20 days to convince the American electorate that they either are well on the way toward having or indeed have transformed themselves into the 21st century equivalents of George Washington, FDR (1936), Ronald Reagan (1984), or James Monroe (1820). Well, that's just not going to happen.

 

[SingleVoyce]

Trump Leads Clinton By 1 Point Going Into Debate — IBD/TIPP Poll

After more than a week of blistering attacks from Democrats, celebrities and the press, Donald Trump has managed to pull ahead of Hillary Clinton by a 1.3 percentage point margin — 41.3% to 40% — in a four-way matchup, according to the new IBD/TIPP poll released today.

And this poll was the most accurate of the 2012 election.

http://fivethirtyeight.blogs.nytime...and-worst-in-the-2012-presidential-race/?_r=0

It is interesting that the phone polls were generally more accurate than internet polls in that election. It seems to not be the case based on the primary results this year.

Trump is changing so many things about our election system , it is amazing.

Two polls released today show radically different results.

This poll:

Clinton Vaults to a Double-Digit Lead

shows Clinton with a 12 point lead and claims to have an MOE of 3.5.

The latest IDP poll:

Clinton Vs. Trump: IBD/TIPP Presidential Election Tracking Poll

shows Trump with a two point lead and claims to have an MOE of 3.6.

How is that possible? Are the MOE numbers completely meaningless?

There is a significant difference among women voters. IDP shows Clinton with an 8 point lead (46 - 38) and the ABC poll shows Clinton with a 20 point lead (55 - 35) but the ABC polls also shows a virtual tie among male voters while the IDP poll has a significant advantage for Trump among that group.

While it's true that the ABC poll was less accurate in 2012 than the IDP poll, it was, in fact, biased toward the Republican candidate.

I can't see any significant difference between the polling methodologies of the two polling organizations so I am having a very difficult time making any sense of this.

 

[JimBowie1958]

Trump Leads Clinton By 1 Point Going Into Debate — IBD/TIPP Poll

After more than a week of blistering attacks from Democrats, celebrities and the press, Donald Trump has managed to pull ahead of Hillary Clinton by a 1.3 percentage point margin — 41.3% to 40% — in a four-way matchup, according to the new IBD/TIPP poll released today.

And this poll was the most accurate of the 2012 election.

http://fivethirtyeight.blogs.nytime...and-worst-in-the-2012-presidential-race/?_r=0

It is interesting that the phone polls were generally more accurate than internet polls in that election. It seems to not be the case based on the primary results this year.

Trump is changing so many things about our election system , it is amazing.

Two polls released today show radically different results.

This poll:

Clinton Vaults to a Double-Digit Lead

shows Clinton with a 12 point lead and claims to have an MOE of 3.5.

The latest IDP poll:

Clinton Vs. Trump: IBD/TIPP Presidential Election Tracking Poll

shows Trump with a two point lead and claims to have an MOE of 3.6.

How is that possible? Are the MOE numbers completely meaningless?

There is a significant difference among women voters. IDP shows Clinton with an 8 point lead (46 - 38) and the ABC poll shows Clinton with a 20 point lead (55 - 35) but the ABC polls also shows a virtual tie among male voters while the IDP poll has a significant advantage for Trump among that group.

While it's true that the ABC poll was less accurate in 2012 than the IDP poll, it was, in fact, biased toward the Republican candidate.

I can't see any significant difference between the polling methodologies of the two polling organizations so I am having a very difficult time making any sense of this.

Yes, at first I couldnt see any significant differences either, back in December of last year, and it started bugging me. Theoretically these were samples taken from the same population and yet they showed very different results even nearly a year ago, so I dont think that it is a change in the election population that is driving this, nor a difference in anything other than the ones that gave Trump a smaller share tended to be of only registered voters.

This difference in voters based on registered vrs those polls that also include the voters who have commitment to vote this year the first time because of Trump but are not yet registered.

But I was thinking the variations should narrow as more first time voters register, but then again, many voters dont make their minds up until the last two weeks before the election.

I dont know exactly what is causing the dissonance in the polls, but we see it every year. Even in calmer elections we have a wider variation in the poll results than the margin of error allows for. So they MU?ST be polling slightly different segments of the population, but which ones and how?

This year the variation is much more pronounced, but we are only feeling the matter in the dark with eyes wide open, but there is simply not enough data, and we will likely only know what the root problem is until a couple of moths after the election, lol.

 

[mamooth]

At this point in 2012, IBD was calling Romney 5 points ahead. They're a joke pollster, along with Rasmussen. They get paid to push polls that are deliberately heavily Republican-biased, to help build enthusiasm in the Republican base. Then, in their final poll, they'll lose that bias, so that they can point to that one decent poll and pretend that they weren't shills the whole time.

Fools the rubes, as this thread shows.

As far as turnout goes, it looks like the Democrats will outperform polls. Trump has the stink of loser about him, and nobody wants to turn out for a loser. Clinton voters, OTOH, are thrilled to have one of the best candidates ever. Trump losers confuse "Noise and rage" with "enthusiasm". And the Clinton ground game is stellar, while Trump doesn't have a ground game.

 

[Bush92]

I think that a comparison of how Rasmussen and Monmouth conduct their polling is illustrative of the complexities of polling, when done honestly, and how expected results can alter the end result.

First, Monmouth has Hillary up by 12, and they have tended to trend Hillary much higher than Trump all summer.

Monmouth University

The Monmouth University Poll was sponsored and conducted by the Monmouth University Polling Institute from October 14 to 16, 2016 with a national random sample of 805 registered voters. Interviews were conducted by a live caller in English, including 400 drawn from a list of registered voters (200 landline / 200 cell phone) and 405 using random digit dial (201 landline / 204 cell phone).

Then we have the Rasmussen methodology

Methodology - Rasmussen Reports™

Data for Rasmussen Reports survey research is collected using an automated polling methodology.

Generally speaking, the automated survey process is identical to that of traditional, operator-assisted research firms such as Gallup, Harris, and Roper. However, automated polling systems use a single, digitally-recorded, voice to conduct the interview while traditional firms rely on phone banks, boiler rooms, and operator-assisted technology.

Here is a huge difference right there. While Monmouth uses a real person to ask the questions that might shame some into not saying they are voting for Trump, more people have an easier time telling an automated voice what they plan to do.

For tracking surveys such as the Rasmussen Reports daily Presidential Tracking Poll, the automated technology ensures that every respondent hears exactly the same question, from the exact same voice, asked with the exact same inflection every single time.

Another huge difference here; how much does the polling caller influence the response when there is a "racist" politically unacceptable candidate involed with one of their supporters? Quite a bit of influence I would think.

All Rasmussen Reports' survey questions are digitally recorded and fed to a calling program that determines question order, branching options, and other factors. Calls are placed to randomly-selected phone numbers through a process that ensures appropriate geographic representation. Typically, calls are placed from 5 pm to 9 pm local time during the week. Saturday calls are made from 11 am to 6 pm local time and Sunday calls from 1 pm to 9 pm local time.

To reach those who have abandoned traditional landline telephones, Rasmussen Reports uses an online survey tool to interview randomly selected participants from a demographically diverse panel.

Another huge difference as not everyone has a landline any more, and using a survey tool like this can mix the numbers of voters without land lines in more evenly.

Rasmussen is also not calling only people who are registered. The Trump supporters who have not yet registered are not on the voting lists likely printed up and distributed from earlier this summer, and are not in Monmouth's lists as a result.

After the surveys are completed, the raw data is processed through a weighting program to ensure that the sample reflects the overall population in terms of age, race, gender, political party, and other factors. The processing step is required because different segments of the population answer the phone in different ways. For example, women answer the phone more than men, older people are home more and answer more than younger people, and rural residents typically answer the phone more frequently than urban residents.

For surveys of all adults, the population targets are determined by census bureau data.

While Monmouth does this as well, they dont give a lot of details about how they do it. "Trust in the Force Luke," right? With Rasmussen they go into the details and they do this job methodically and with a computer process designed to prevent influence by human factors like flawed expectations.

For political surveys, census bureau data provides a starting point and a series of screening questions are used to determine likely voters. The questions involve voting history, interest in the current campaign, and likely voting intentions.

Monmouth uses only registered voters as their criteria for "likely voters", which again ignores all the new voters that Trump is bringing into the system.

In addition to that, Trump voters are less likely to answer polls as they are alienated from the media in general as even FOX News has been going after their guy.

I think Rasmussens numbers are more accurate in this election and they still underestimate Trumps support by a good 5% to 10% due to politically correct shame, irritation.with the media, hiding their voting intentions from those who may over hear them responding to a poll.

If you look at the primaries and how inaccurate the polls were then, you notice that in all but two cases of the ones I gave here, they under estimated Trumps numbers by about 5% to 15% and in one case off by 33%!

Take one example, the NBC/Marist poll that claimed Trump had dropped to a 5% lead in South Carolina. Trump actually came away with a 10% majority. But Marist only uses registered voters in its sampling, so that might well explain the incongruity and their missing trumps margin of victory by 100%.

Clinton with Wide Lead over Trump Nationally | Home of the Marist Poll: Pebbles and Pundits

In a national survey conducted from Monday through Wednesday this week[†], Democrats Hillary Clinton and Tim Kaine, 48%, lead Republicans Donald Trump and Mike Pence, 33%, by 15 points among registered voters in the race for the White House.

This has been a regularly recurring phenomenon all year as polls of registered voters under estimate trumps vote by at least 5%.

I think that the majority of the differences we see in these polls show Trump apparently losing support when he is actually only seeing more of his voters shamed into not responding Truthfully.

So when you see these polls, if you want a more accurate reflection of the likely results, take Trumps percentage and add 5%.

Podesa email talks of putting out skewed polls to show fake momentum for Hillary.

 

[Bush92]

I think that a comparison of how Rasmussen and Monmouth conduct their polling is illustrative of the complexities of polling, when done honestly, and how expected results can alter the end result.

First, Monmouth has Hillary up by 12, and they have tended to trend Hillary much higher than Trump all summer.

Monmouth University

The Monmouth University Poll was sponsored and conducted by the Monmouth University Polling Institute from October 14 to 16, 2016 with a national random sample of 805 registered voters. Interviews were conducted by a live caller in English, including 400 drawn from a list of registered voters (200 landline / 200 cell phone) and 405 using random digit dial (201 landline / 204 cell phone).

Then we have the Rasmussen methodology

Methodology - Rasmussen Reports™

Data for Rasmussen Reports survey research is collected using an automated polling methodology.

Generally speaking, the automated survey process is identical to that of traditional, operator-assisted research firms such as Gallup, Harris, and Roper. However, automated polling systems use a single, digitally-recorded, voice to conduct the interview while traditional firms rely on phone banks, boiler rooms, and operator-assisted technology.

Here is a huge difference right there. While Monmouth uses a real person to ask the questions that might shame some into not saying they are voting for Trump, more people have an easier time telling an automated voice what they plan to do.

For tracking surveys such as the Rasmussen Reports daily Presidential Tracking Poll, the automated technology ensures that every respondent hears exactly the same question, from the exact same voice, asked with the exact same inflection every single time.

Another huge difference here; how much does the polling caller influence the response when there is a "racist" politically unacceptable candidate involed with one of their supporters? Quite a bit of influence I would think.

All Rasmussen Reports' survey questions are digitally recorded and fed to a calling program that determines question order, branching options, and other factors. Calls are placed to randomly-selected phone numbers through a process that ensures appropriate geographic representation. Typically, calls are placed from 5 pm to 9 pm local time during the week. Saturday calls are made from 11 am to 6 pm local time and Sunday calls from 1 pm to 9 pm local time.

To reach those who have abandoned traditional landline telephones, Rasmussen Reports uses an online survey tool to interview randomly selected participants from a demographically diverse panel.

Another huge difference as not everyone has a landline any more, and using a survey tool like this can mix the numbers of voters without land lines in more evenly.

Rasmussen is also not calling only people who are registered. The Trump supporters who have not yet registered are not on the voting lists likely printed up and distributed from earlier this summer, and are not in Monmouth's lists as a result.

After the surveys are completed, the raw data is processed through a weighting program to ensure that the sample reflects the overall population in terms of age, race, gender, political party, and other factors. The processing step is required because different segments of the population answer the phone in different ways. For example, women answer the phone more than men, older people are home more and answer more than younger people, and rural residents typically answer the phone more frequently than urban residents.

For surveys of all adults, the population targets are determined by census bureau data.

While Monmouth does this as well, they dont give a lot of details about how they do it. "Trust in the Force Luke," right? With Rasmussen they go into the details and they do this job methodically and with a computer process designed to prevent influence by human factors like flawed expectations.

For political surveys, census bureau data provides a starting point and a series of screening questions are used to determine likely voters. The questions involve voting history, interest in the current campaign, and likely voting intentions.

Monmouth uses only registered voters as their criteria for "likely voters", which again ignores all the new voters that Trump is bringing into the system.

In addition to that, Trump voters are less likely to answer polls as they are alienated from the media in general as even FOX News has been going after their guy.

I think Rasmussens numbers are more accurate in this election and they still underestimate Trumps support by a good 5% to 10% due to politically correct shame, irritation.with the media, hiding their voting intentions from those who may over hear them responding to a poll.

If you look at the primaries and how inaccurate the polls were then, you notice that in all but two cases of the ones I gave here, they under estimated Trumps numbers by about 5% to 15% and in one case off by 33%!

Take one example, the NBC/Marist poll that claimed Trump had dropped to a 5% lead in South Carolina. Trump actually came away with a 10% majority. But Marist only uses registered voters in its sampling, so that might well explain the incongruity and their missing trumps margin of victory by 100%.

Clinton with Wide Lead over Trump Nationally | Home of the Marist Poll: Pebbles and Pundits

In a national survey conducted from Monday through Wednesday this week[†], Democrats Hillary Clinton and Tim Kaine, 48%, lead Republicans Donald Trump and Mike Pence, 33%, by 15 points among registered voters in the race for the White House.

This has been a regularly recurring phenomenon all year as polls of registered voters under estimate trumps vote by at least 5%.

I think that the majority of the differences we see in these polls show Trump apparently losing support when he is actually only seeing more of his voters shamed into not responding Truthfully.

So when you see these polls, if you want a more accurate reflection of the likely results, take Trumps percentage and add 5%.

look at the Hassan Ayotte race. UNH poll had a skewed methodology. Hassan up by 9 all of the sudden? They polled 9% more Democrats.