Can someone offer an explanation for the differences in polls?

[Anonymous]

I see polling numbers at both national and state level that are all over the place - and I am not talking minor differences.

It makes me wonder if these polls are rigged to some degree by those with an agenda representing their own biases. If this is the case, I suspect in the final week or two before the election we will likely see these differences evaporate because ultimately the reputation of the pollsters will be affected if they are wrong by a wide margin.

Can anyone offer a rational explanation for these differences between reputable pollsters?

[Anonymous]

Anonymous wrote:I see polling numbers at both national and state level that are all over the place - and I am not talking minor differences.

It makes me wonder if these polls are rigged to some degree by those with an agenda representing their own biases. If this is the case, I suspect in the final week or two before the election we will likely see these differences evaporate because ultimately the reputation of the pollsters will be affected if they are wrong by a wide margin.

Can anyone offer a rational explanation for these differences between reputable pollsters?

Here is some place you can start - https://www.historians.org/about-aha-and-membership/aha-history-and-archives/gi-roundtable-series/pamphlets/are-opinion-polls-useful/why-do-polls-get-different-results

[Anonymous]

They are meaningless op.

[Anonymous]

Different samples, different respondent weightings, different ways of phrasing the same question/topic can lead to divergent results.

The reason Nate Silver is so successful is because he aggregates poll results and averages out the statistical noise/biases that are inherent in individual polls, also factoring in historical reliability of individual pollsters (compared to the actual final results). Using such a massive set of data gets you closer to the real result.

[Anonymous]

Anonymous wrote:I see polling numbers at both national and state level that are all over the place - and I am not talking minor differences.

It makes me wonder if these polls are rigged to some degree by those with an agenda representing their own biases. If this is the case, I suspect in the final week or two before the election we will likely see these differences evaporate because ultimately the reputation of the pollsters will be affected if they are wrong by a wide margin.

Can anyone offer a rational explanation for these differences between reputable pollsters?

Here is some more information. You can see that the polls are not wildly divergent within the margin of error.

http://projects.fivethirtyeight.com/2016-election-forecast/national-polls/

[Anonymous]

Anonymous wrote:They are meaningless op.

they aren't meaningless or campaigns would not use them.

But you can't just take them at face value. You need to understand the methodology.