news polls, believe or not

[g+16]

In all my years I have lived all across the United States, and I have never been called for a poll for anything, not on policticals, drugs use, guns, nothing. Then you looks at the numbers given for the poll and they say they polled 1,000 people, who in the hel@ would really believe that a thousand people can speak for the entire U.S.!! I've always said if you really want to know what most people think put in question on the election ballet, then you would at least have the say of anyone who bothered to vote. I've always known that anyone that wants to can play with numbers to get them to come out anyway THEY want, so why waste their time and money???

 

[badwolf.usmc]

Small samples, and the margin of error « What’s new

Here is a good article on the math behind polls. It answers your question about how 1,000 people can "speak" for the entire US.


A poll can mean anything you want it to mean, and there are many issues with polls in general that math can't fix. Like anything that deals with people's opinions, just take them with a grain of salt.

 

[JOLLYKART]

After seeing your icon, it made me think about the time I got sampled on a Orville Redenbacher taste test in a mall, I pick the other guys! LOL

 

[NYFelon]

I've been phone-polled before. Most times the questions are worded in a way that they elicit the response the pollster is looking for.

 

[kwatters]

I've been phone-polled before. Most times the questions are worded in a way that they elicit the response the pollster is looking for.

That has been my experience.

 

[melensdad]

There are several VERY ACCURATE national polling companies that do a very good job of showing a snapshot of public opinion. To simply dismiss all polls as biased is nothing short of absurd.

Probably the most accurate of the polling companies are:

• Rasmussen Reports™: The Most Comprehensive Public Opinion Data Anywhere
• Public Policy Polling
• Quinnipiac University Polling Institute - Quinnipiac University – Hamden, Connecticut
• Home of the Marist Poll: Pebbles and Pundits
• Pew Research Center

There are other reputable polling companies like Gallup and Harris. Each of the above, has been proven, over long time periods, and through many elections, to be good predictors of the national mood and even elections. With all polls it is important to understand that they are snapshots in time. Their use as a predictor of the future is dubious over long term but they can be used to fairly and accurately predict short term public moods.

I believe one of the best predictors is REAL CLEAR POLITICS as they take the results of SEVERAL DIFFERENT POLLS and they average them out so if one poll has a conservative bias and another has a liberal bias the R.C.P. data averages those biases.

As for one single poll that I generally trust, it would be Rasmussen. Pew Research is also high on my list of trusted pollsters.

 

[Plinker]

Only those with home telephones are polled (AFAIK). And that pool is getting smaller every year.

 

[dross]

Statistical sampling can be very accurate.

Assuming your test (in the case of a poll, that test is questions and answers) is valid, and truly random, the sample size determines the accuracy of the poll plus or minus a certain percentage.

The problem with polls of people are many. The questions can be a problem, interpretation of the answers can be a problem, and it's hard to get a truly random sampling.

This can lead to manipulation as well as honest error.

The techniques to ensure all of the above are well established over time, however. There are certain questions and certain answer choices that have been proven to be valid.

Polls run by the media are useless. The internal polling done by the political parties to give themselves a true picture of the political situation are about as accurate as the process can be.

Melensdad pointed out above some good ones that anyone can access.

It's just as silly to dismiss all statistics as it is to think they are always accurate.

 

[melensdad]

Only those with home telephones are polled (AFAIK). And that pool is getting smaller every year.

Where are your facts to back up this statement

I know it is NOT TRUE.

Statistical sampling can be very accurate.

. . .

It's just as silly to dismiss all statistics as it is to think they are always accurate.

Exactly.

You need to study polls and polling companies just like you study anything else. There are some that are excellent predictors and some that are there for the simple purpose of pushing a point.

Companies like Harris and Gallup can be VERY GOOD or they can be VERY BAD. Typically in national and state elections they are good (not the best) but good. On the other hand they are also so commercialized that they often will work for advocacy groups and structure an poll that pushes an agenda.

Don't buy into any poll's data. Study it. Study the source. Read the questions. Be a skeptic, but also be willing to believe.

 

[dross]

Be a skeptic, but also be willing to believe.

'Zackly.

Skeptical ignorance is, well, it's just plain ol' ignorance, same as naive ignorance.

 

[88GT]

Funny this should come up. I was called Monday night. I believe it was one of the campaigns for Indy mayor; it was definitely limited to mayoral candidates, just not sure who commissioned the thing. Anywho, I answered correctly regarding all the 'issue' questions, but they think I'm a 41-year-old Latina woman who is separated from her husband.

People put too much stock in polls. They are temperature testers of the crudest kind. Like Mom putting the back of her hand on your forehead or cheek to see if you have a fever. All she's gonna be able to tell is whether you have one or not. Actual body temp requires a better tool. Polls are like that. They'll be able to tell the general mood of people towards this or that, but specificity beyond that is outside the scope of a poll to determine.

 

[Sylvain]

Funny this should come up. I was called Monday night. I believe it was one of the campaigns for Indy mayor; it was definitely limited to mayoral candidates, just not sure who commissioned the thing. Anywho, I answered correctly regarding all the 'issue' questions, but they think I'm a 41-year-old Latina woman who is separated from her husband.

People put too much stock in polls. They are temperature testers of the crudest kind. Like Mom putting the back of her hand on your forehead or cheek to see if you have a fever. All she's gonna be able to tell is whether you have one or not. Actual body temp requires a better tool. Polls are like that. They'll be able to tell the general mood of people towards this or that, but specificity beyond that is outside the scope of a poll to determine.

And you mean they were wrong because you are still with your husband?

 

[melensdad]

They are temperature testers of the crudest kind. Like Mom putting the back of her hand on your forehead or cheek to see if you have a fever. All she's gonna be able to tell is whether you have one or not. Actual body temp requires a better tool. Polls are like that. They'll be able to tell the general mood of people towards this or that, but specificity beyond that is outside the scope of a poll to determine.

If what you say is true then how is it that, within reasonable limits, the good polls can predict not simply an election, but actual election RESULTS within a very close margin of error?

 

[Blackhawk2001]

If what you say is true then how is it that, within reasonable limits, the good polls can predict not simply an election, but actual election RESULTS within a very close margin of error?

Luck? Sometimes polls are very wrong, too.

 

[Bunnykid68]

Funny this should come up. I was called Monday night. I believe it was one of the campaigns for Indy mayor; it was definitely limited to mayoral candidates, just not sure who commissioned the thing. Anywho, I answered correctly regarding all the 'issue' questions, but they think I'm a 41-year-old Latina woman who is separated from her husband.

People put too much stock in polls. They are temperature testers of the crudest kind. Like Mom putting the back of her hand on your forehead or cheek to see if you have a fever. All she's gonna be able to tell is whether you have one or not. Actual body temp requires a better tool. Polls are like that. They'll be able to tell the general mood of people towards this or that, but specificity beyond that is outside the scope of a poll to determine.

Your moms hand was probably more accurate than the polls

 

[sgreen3]

You can make a poll come out to however you want it to, thats why I dont pay any attention to news polls. They are as bias as the news channel reporting them.

 

[jbombelli]

DEWEY DEFEATS TRUMAN!

'Nuff said.

 

[Scutter01]

I get called for polls several times per week. I invariably interrupt their opening spiel to tell them my rates for taking polls. It usually goes something like this:

Annoying pollster: "Hi, I'm calling from (some poll). This is not a sales call...."
Me: "Oh, great! Let me tell you my rates. I normally charge $50/minute to take phone polls, but I'm having a 10% off special this week if you buy at least a five-minute block. It's like getting thirty seconds FREE! How will you be paying today?"
Pollster: "What?"
Me: (repeating request for payment)
Pollster: "I'm sorry, we don't pay for polls."
Me: "So, your time is worth more than mine?"
Pollster: "What?" (I get this a lot. My comments completely derail them. It's awesome.)
Me: "Well, they're paying you to administer the poll, right? I expect to be paid to take the poll. Why should your time be more valuable than my time? So, how will you be paying me for my time today? I can take most major credit cards."
Pollster: "But...you have the chance to make your opinion count!" (or some variation on the feel-good payment response)
Me: "I'm sorry, I've already stated my rates twice now."
Pollster: "Uhh....let me get my supervisor..."

The call usually ends 15 seconds later.

 

[sgreen3]

Scutter that is Awesome!

 

[dsol]

I was polled before the 2008 election. I answered neither when they asked who I supported. I told them I would vote for a bag of leaves over either of the asshats running.