Pity partyfor Mittens ..
- Thread starter retired-libera
- Start date
JoeRider
Federalist Live Forever
Huffy Post is on spot as always.
Wisconsin Recall Election: Disaster Looms for GOP
Craig Holman: Wisconsin Recall Election: Disaster Looms for GOP
I tend to stick with the most reliable pollsters. Oddly enough, Gallup is the most widely known, and recently, the absolute worst in reliability and consistency.
But something has come up in the news, lately - many of the major polling outlets are using a model based on the 2008 election. This is a mistake, and certain to make outcomes look favorable for Obama. I think this is a deliberate ploy to mask support for Romney in an attempt to weaken support and depress the vote.
I don't work with opinion polls, but I have been working on surveys most of the last 22 years. Something, some of you know - polls are only as good as the model it is based on - the "methodology". If they only ever polled campus students or Democrats, they wouldn't need answers. Before you conduct your sampling - your calls - you need to have an idea of the universe you are drawing from.
I wish I could tell how many times I've heard the stupid retort of - "they can't ask everyone, so how can they know?". Well, that is precisely the entire point of a survey - correctly ascertaining the whole by looking at a part.
If you are trying to learn how people will vote, you have to determine what group you are collecting data from. For example, there's really no point in asking someone who can't vote - such as someone under 18 or someone who is not a citizen. There's also no reason to ask someone who is not registered to vote - not if you are trying to learn the outcome of an election as opposed to how popular a candidate is. But you also have to know the makeup of your sample - how many do you expect will be women, what is their party affiliation, what is their age. You have to add in the likelihood of getting answers - for example, some polling agencies have determined that Thursdays are the night best for telephone calls, because if you call on Friday or Saturday night, you have very little chance of finding young singles at home. If you call in the middle of the day, you're just going to find stay-at-home moms, retired people and unemployed people.
And so on.
When you're done, you weight the results according to the data you have. If out of a large group you get say, only a few Republicans, but you know there should be 35% of your group - you give them more weight. If you have far more women or youth, you give them less weight.
HERE is the problem:
What if your data says - there's only 29% Republicans? And your random sample shows more? You diminish their contribution.
The problem with these polls is, many of them assume the turnout this November will be JUST LIKE four years ago. And that year was very different from 2004, 2006, and 2010. ANY sample based on that assumption will absolutely show a MUCH greater result for Obama.
-
-
So, for now, ignore polls. Vote as you want. Try to find out more about your candidate rather than how he's doing on the back stretch.
But something has come up in the news, lately - many of the major polling outlets are using a model based on the 2008 election. This is a mistake, and certain to make outcomes look favorable for Obama. I think this is a deliberate ploy to mask support for Romney in an attempt to weaken support and depress the vote.
I don't work with opinion polls, but I have been working on surveys most of the last 22 years. Something, some of you know - polls are only as good as the model it is based on - the "methodology". If they only ever polled campus students or Democrats, they wouldn't need answers. Before you conduct your sampling - your calls - you need to have an idea of the universe you are drawing from.
I wish I could tell how many times I've heard the stupid retort of - "they can't ask everyone, so how can they know?". Well, that is precisely the entire point of a survey - correctly ascertaining the whole by looking at a part.
If you are trying to learn how people will vote, you have to determine what group you are collecting data from. For example, there's really no point in asking someone who can't vote - such as someone under 18 or someone who is not a citizen. There's also no reason to ask someone who is not registered to vote - not if you are trying to learn the outcome of an election as opposed to how popular a candidate is. But you also have to know the makeup of your sample - how many do you expect will be women, what is their party affiliation, what is their age. You have to add in the likelihood of getting answers - for example, some polling agencies have determined that Thursdays are the night best for telephone calls, because if you call on Friday or Saturday night, you have very little chance of finding young singles at home. If you call in the middle of the day, you're just going to find stay-at-home moms, retired people and unemployed people.
And so on.
When you're done, you weight the results according to the data you have. If out of a large group you get say, only a few Republicans, but you know there should be 35% of your group - you give them more weight. If you have far more women or youth, you give them less weight.
HERE is the problem:
What if your data says - there's only 29% Republicans? And your random sample shows more? You diminish their contribution.
The problem with these polls is, many of them assume the turnout this November will be JUST LIKE four years ago. And that year was very different from 2004, 2006, and 2010. ANY sample based on that assumption will absolutely show a MUCH greater result for Obama.
-
-
So, for now, ignore polls. Vote as you want. Try to find out more about your candidate rather than how he's doing on the back stretch.
MarieB
New Member
Hmm, according to Ramussen, Romney is up in Colorado and Iowa
2012 Electoral College Scoreboard - Rasmussen Reports™
Daily Tracking here
Daily Presidential Tracking Poll - Rasmussen Reports™
2012 Electoral College Scoreboard - Rasmussen Reports™
Daily Tracking here
Daily Presidential Tracking Poll - Rasmussen Reports™
Looks like we have another "nhboy" on our forum.
Dontel58
New Member
JoeRider
Federalist Live Forever
No significant otherIt is His father
Hessian
Well-Known Member
Nc...& fl...?
I will paraphrase: The obama camp said that the NC vote might be the crucial to tipping them to victory (hence the convention/Muslim get-together was scheduled there)...yet recently all the eyes shift back to Wisconsin & Ohio...
hey dems, why have you given up on NC?...internal polling tell you that you already lost NC?
When was obama in Florida last....have they bailed there too?
With 40+days left: obama's travel schedule may indicate that they are trying to hold on to previously supportive states--that: is a sign of weakness.
I will paraphrase: The obama camp said that the NC vote might be the crucial to tipping them to victory (hence the convention/Muslim get-together was scheduled there)...yet recently all the eyes shift back to Wisconsin & Ohio...
hey dems, why have you given up on NC?...internal polling tell you that you already lost NC?
When was obama in Florida last....have they bailed there too?
With 40+days left: obama's travel schedule may indicate that they are trying to hold on to previously supportive states--that: is a sign of weakness.
JoeRider
Federalist Live Forever
Ssssss........don't give them any hints about the fact that they might be losing ground. Next thing you know they will be pushing third party candidates to try and take votes away
retired-libera
New Member
Google that hotbed of Obama support Fox News. They show similar results.IMHO I think the reason is because the Repugnicant party is controlled by right wing fundamentalist religious fanatics. They are so misogynistic that nobody but old white men will vote for them . I am an old white man but I ain;t that old .
Last edited:
-SS-
DHS Extremist
Best post of an otherwise retarded thread.
I was hoping he would answer before anyone provided information like this.
The OP has a propensity for not understanding many things, including citizenship.