Robo-poll actual survey question...

[Taco John]

Give it up for "scientific" polling!

There will be a number of people running for President in 2008 as Republicans. Some of the most talked about are Rudy Giuliani, John McCain, Mitt Romney and Fred Thompson. If the Republican primary were today which of these men would you vote for, or would you vote for somebody else?

Giuliani................... .19% McCain................... . 8% Romney ................. .10% Thompson.............. .24% Other...................... .14% Undecided.............. .25%


http://www.publicpolicypolling.com/pdf/surveys/PPP_Release_110707.pdf

[Taco John]

Hahah! I called the guy who ran the poll if he thought that his question was biased and predisposed the listener to his pre-determined candidate. He said that if we wanted to pay $2500, he'd be happy to phrase the question differentley or some such nonsense.

[Adept Havelock]

Hahah! I called the guy who ran the poll if he thought that his question was biased and predisposed the listener to his pre-determined candidate. He said that if we wanted to pay $2500, he'd be happy to phrase the question differentley or some such nonsense.

"Nonsense?" A libertarian condemning the free market, and the right of the pollster to perform a poll as the buyer requested? Whodathunkit?

Sounds like he offered you the same opportunity as whoever purchased this poll. What's the problem?

Maybe we need a government department regulating evil pollsters, to make sure the polls are done "properly".

[Cochise]

I wouldn't have included the "some of the most talked about" line but I don't think it's that horrendous. If you remove it I don't think you'd like the poll either.

[recxjake]

They should just say all 8 names. I wouldn't include Keyes or Cox.

[SNR]

They should just say all 8 names. I wouldn't include Keyes or Cox.Cox dropped out a couple weeks ago

[recxjake]

Cox dropped out a couple weeks ago

hmmm.. missed that... oh well.

[banyon]

"Nonsense?" A libertarian condeming the free market, and the right of the pollster to perform a poll as the buyer requested? Whodathunkit?

Sounds like he offered you the same opportunity as whoever purchased this poll. What's the problem?

Maybe we need a government department regulating evil pollsters, to make sure the polls are done "properly".

Burn. :Bartee:

[Adam]

"Nonsense?" A libertarian condemning the free market, and the right of the pollster to perform a poll as the buyer requested? Whodathunkit?

Sounds like he offered you the same opportunity as whoever purchased this poll. What's the problem?

Maybe we need a government department regulating evil pollsters, to make sure the polls are done "properly".

I consider myself a Libertarian, and I condemn poor sectors of the free market all the time. If I think American Idol is completely stupid, and I mock people who watch it, does that mean I am not a Libertarian?

[Nightfyre]

"Nonsense?" A libertarian condemning the free market, and the right of the pollster to perform a poll as the buyer requested? Whodathunkit?

Sounds like he offered you the same opportunity as whoever purchased this poll. What's the problem?

Maybe we need a government department regulating evil pollsters, to make sure the polls are done "properly".
Libertarians maintain the right to criticize poor work and discourage people from allowing said work to sell, just not through the government. So by calling your attention to the subject, we, as consumers, can discourage you from believing the polls, reducing their effectiveness, ultimately leading to the demise of a bad company.

[Adept Havelock]

I consider myself a Libertarian, and I condemn poor sectors of the free market all the time. If I think American Idol is completely stupid, and I mock people who watch it, does that mean I am not a Libertarian?

Libertarians maintain the right to criticize poor work and discourage people from allowing said work to sell, just not through the government. So by calling your attention to the subject, we, as consumers, can discourage you from believing the polls, reducing their effectiveness, ultimately leading to the demise of a bad company.

Can either of you explain how it can be considered poor business practice or poor work to give the client what they asked for?

The pollsters response suggests the person who commissioned the report wanted it phrased in exactly this manner.

He offered TJ the very same opportunity.

He provided the customer with the service he requested. Isn't that the open market at work? What's the problem?

If there is a problem, I'd suggest it lies with the person who comissioned the poll, not the pollster. :shrug:

[Nightfyre]

Can either of you explain how it can be considered poor business practice or poor work to give the client what they requested?

The pollsters response suggests the person who commissioned the report wanted it phrased in exactly this manner.

He offered TJ the same opportunity.

He provided the customer with the service he requested. Isn't that the open market at work? What's the problem?

If there is a problem, I'd suggest it lies with the person who comissioned the poll, not the pollster.
It's poor practice to call it scientific and pretend it accurately portrays the opinions of the American public while not disclosing either question format or sampling method.

[Adept Havelock]

It's poor practice to call it scientific and pretend it accurately portrays the opinions of the American public while not disclosing either question format or sampling method.

Wow. :shake:

Eerily reminiscent of recxjakes early days, before he gave up on putting his foot in his mouth and stuck to cutting and pasting bolierplate.

Perhaps you should have read page 9 of the .PDF TJ linked, as it explicitly shows the question format.

The poll also clearly states:

PPP surveyed 679 likely Democratic primary voters and 767 likely Republican primary voters on November 5. The surveys have margins of error of ± 3.8% and ± 3.5% respectively. Other factors, such as refusal to be interviewed and weighting, may introduce additional error that is more difficult to quantify. Complete results are attached and can be found at www.publicpolicypolling.com. These include issue priorities for both parties, and crosstabs breaking down the results by gender, race, age and region.

So there's the sampling method. :shrug:

Again, it seems the client provided the customer with the service requested, and apparently practiced full disclosure regarding questions, sampling method, and margin for error.

I can see no Ideological reason a Libertarian would have a problem with that.

Well, no reason other than whoever purchased this poll left their pet candidate RP out in the cold. ;) :Poke:

[Nightfyre]

Wow. :shake:

Eerily reminiscent of recxjakes early days, before he gave up on putting his foot in his mouth and stuck to cutting and pasting bolierplate.

Perhaps you should have read page 9 of the .PDF TJ linked, as it explicitly shows the question format.

The poll also clearly states:

PPP surveyed 679 likely Democratic primary voters and 767 likely Republican primary voters on November 5. The surveys have margins of error of ± 3.8% and ± 3.5% respectively. Other factors, such as refusal to be interviewed and weighting, may introduce additional error that is more difficult to quantify. Complete results are attached and can be found at www.publicpolicypolling.com. These include issue priorities for both parties, and crosstabs breaking down the results by gender, race, age and region.

So there's the sampling method. :shrug:

Again, it seems the client provided the customer with the service requested, and apparently practiced full disclosure regarding questions, sampling method, and margin for error.

I can see no Ideological reason a Libertarian would have a problem with that.

Well, no reason other than whoever purchased this poll left their pet candidate RP out in the cold. ;) :Poke:
I have no problem with the client paying them to generate a poll. But I do have rights as the intended audience to boycott polls on the premise that they are only scientific to the extent that the client who purchases the poll requires them to be. Its really very simple. This is how free markets work, I'm surprised your having trouble with that concept. If I make the politicians tool ineffective, the politician will stop commissioning polls. :shrug: Free market at work.

[Adept Havelock]

I have no problem with the client paying them to generate a poll. But I do have rights as the intended audience to boycott polls on the premise that they are only scientific to the extent that the client who purchases the poll requires them to be. Its really very simple. This is how free markets work, I'm surprised your having trouble with that concept. If I make the politicians tool ineffective, the politician will stop commissioning polls. :shrug: Free market at work.

Wait a second. I thought the problem was that the pollster did not "disclose either question format or sampling method."

That's not your problem with it? Then why did you claim it was? :hmmm:

Thank you for admitting (albeit in a roundabout manner) the only actual problem with the poll is you don't like it.

[Taco John]

The pollsters response suggests the person who commissioned the report wanted it phrased in exactly this manner.


It wasn't a commissioned report. It was his own poll. He commissoned it himself.

I don't have a problem with him doing a poll in how he wants to do it. I have a problem with him passing it off as a legitimate, schientific endeavor.

[Adept Havelock]

It wasn't a commissioned report. It was his own poll. He commissoned it himself.

I don't have a problem with him doing a poll in how he wants to do it. I have a problem with him passing it off as a legitimate, schientific endeavor.

So? If it wasn't comissioned, I would think a libertarian would be the last person to tell another what they should or shouldn't do with their own labor or company. Perhaps I am mistaken, or is this some form of "neo-libertarianism"? :p

He stated everything about questions and method sampling. Full disclosure, right there in the PDF you linked to. That should settle the "legitimate or scientific" aspects of your kvetching.

Your only actual problem with the poll is the same as Nightfyres.

You don't like the fact he left your pet candidates name out of it. :shrug:

[Nightfyre]

Wait a second. I thought the problem was that the pollster did not "disclose either question format or sampling method."

That's not your problem with it? Then why did you claim it was?

Thank you for admitting (albeit in a roundabout manner) the only actual problem with the poll is you don't like it.
What do I have to qualify poor work for you or something? Thats basically what I said. Polling is an art, not a science, and art is subjective, therefore, poor polling work to me is a lot like rap music: A bunch of words strung together that don't mean anything or reflect on the realities of the world.

Libertarians maintain the right to criticize poor work and discourage people from allowing said work to sell, just not through the government. So by calling your attention to the subject, we, as consumers, can discourage you from believing the polls, reducing their effectiveness, ultimately leading to the demise of a bad company.

[Adept Havelock]

What do I have to qualify poor work for you or something? Thats basically what I said. Polling is an art, not a science, and art is subjective, therefore, poor polling work to me is a lot like rap music: A bunch of words strung together that don't mean anything or reflect on the realities of the world.

Wait, it's an art? Then why in this post are you so concerned with the Scientific aspects of it?

It's poor practice to call it scientific and pretend it accurately portrays the opinions of the American public while not disclosing either question format or sampling method.

Oh wait, you weren't actually concerned with those points after I pointed out he'd disclosed both of those.

Now where do you want to move the goalposts to? ROFL

Have a nice night, Paulites.

[Taco John]

What do I have to qualify poor work for you or something? Thats basically what I said. Polling is an art, not a science, and art is subjective, therefore, poor polling work to me is a lot like rap music: A bunch of words strung together that don't mean anything or reflect on the realities of the world.

Don't bother... He's just being a contrarian for the sake of being a contrarian.

It's easy to be a contrarian when you don't stand for anything.

[Nightfyre]

Wait, it's an art? Then why in this post are you so concerned with the Scientific aspects of it?



Oh wait, you weren't actually concerned with those points after I pointed out he'd disclosed both of those.

Now where do you want to move the goalposts to? ROFL

Have a nice night, Paulites.
What's inconsistent about my position? I said it wasn't scientific, but was trying to be passed off as such. :shrug:

[patteeu]

The term "scientific" when used in polling usually just means that the sample was selected by the pollster to ensure a degree of randomness so that the results can be generalized to the total population (whatever that population happens to be, e.g. teenage boys, likely voters, parents, etc.). It doesn't ensure that the questions will be appreciated by all nor does it ensure that the results will answer all questions.

There is nothing inherently wrong with the question that this guy asked. Assuming he did a good job of selecting his sample, the results should be accurate for the population he was targeting. But they are accurate only with respect to the actual question not to some question you wish they would have asked. This poll tells us very little about how Ron Paul would do head to head against Hillary because it doesn't ask about that. What it does ask about is, given a handful of specifically named candidates, which one are you most likely to vote for (or something to that effect).

IOW, I agree with Adept that there isn't anything wrong with this poll except that you guys don't like the question. Rather than complaining about the question though, you should simply recognize that it isn't supposed to test Ron Paul's strength against these named candidates. To the extent that the pollster tries to pass it off as if it does, you'd be justified in criticizing him. To the extent that the pollster was testing something else, your frustration is misplaced.

[Taco John]

The term "scientific" when used in polling usually just means that the sample was selected by the pollster to ensure a degree of randomness so that the results can be generalized to the total population (whatever that population happens to be, e.g. teenage boys, likely voters, parents, etc.). It doesn't ensure that the questions will be appreciated by all nor does it ensure that the results will answer all questions.

--or be free from bias.



There is nothing inherently wrong with the question that this guy asked.


The hell there isn't... It's a question loaded with bias. This is basic 101 stuff...

[patteeu]

--or be free from bias.






The hell there isn't... It's a question loaded with bias. This is basic 101 stuff...

As long as the question accurately captures what the pollster wants to capture, bias is irrelevant. In fact, every poll is biased by it's questions. You don't like it because the results won't show how Ron Paul is doing. I presume the pollster is aware of this and isn't going to rely on the poll to understand the level of support for Ron Paul.

IOW, the question is worthless for the purpose you want to use the poll for, but it may well get at exactly what the pollster wants to measure.

[Taco John]

As long as the question accurately captures what the pollster wants to capture, bias is irrelevant.


Which is the whole point. These polls are used to engineer public opinion, not measure it.

[patteeu]

Which is the whole point. These polls are used to engineer public opinion, not measure it.

I don't know about that. You could be right, but it's not necessarily the case. Who was this poll produced for? If it's someone who believes the race will be narrowed to the 4 mentioned candidates by the time it gets to North Carolina, it may be a sincere effort to evaluate the race they envision. By including all the minor candidates, they wouldn't get as good a picture of how people will vote when the minor candidates drop out.

As long as the questions are available, you have all the information you need to discount the results and to alert all interested parties that Ron Paul's support isn't being tested in this particular poll.

Every poll is a product of it's questions and every poll result can be misused by those who don't pay attention to what the question is really asking. For example, back in September, you came to the following conclusion (http://www.chiefsplanet.com/BB/showthread.php?p=4204909#post4204909) based on a poll:

The American people want out of Iraq, regardless of the military situation there.

A closer look at the poll question (http://www.usatoday.com/news/polls/tables/live/2007-09-09-iraq.htm) reveals a different story though:

If you had to choose, which do you think is better for the U.S. -- [ROTATED: to keep a significant number of troops in Iraq until the situation there gets better, even if that takes many years, (or) to set a time-table for removing troops from Iraq and to stick to that timetable regardless of what is going on in Iraq at the time]?

The poll doesn't offer respondents the opportunity to choose among all possible options. For example, there's no option for setting a timetable and adjusting it based on the circumstances.

Every poll is biased by it's questions. Every poll can be misused in one way or another as you did when you tried to pass this particular result off as a clear indication of support for the retreat-at-all-costs option in Iraq.

[Adept Havelock]

Don't bother... He's just being a contrarian for the sake of being a contrarian.

It's easy to be a contrarian when you don't stand for anything.

ROFL

[Cochise]

WTF didn't he just put up a billboard asking for text messages? What a dumbass. This poll is worthless.

[Taco John]

Hmmm... That's interesting...

A huge billboard that says only:

Text in a vote for your favorite candidate:

Rudy Giuliani (45848)
Mike Huckabee (58394)
Ron Paul (48392)
Mitt Romney (87489)
John McCain (43092)
Fred Thompson (48594)


One of my chief problems with phone polling is that they have to contact those being polled in order to get the results. This creates at least two problems:

1) People like me who primarily use a cell phone and don't answer numbers that I don't recognize on the home phone caller ID never get their opinions registered.

2) People don't wait for a phone call in order to vote. They have to actually take action in order to vote. Primary voters take action based on motivation, whether that motivation be a sense of duty, or a sense of enthusiasm for their candidate. Opinion polling neglects this key aspect.

[patteeu]

Hmmm... That's interesting...

A huge billboard that says only:

Text in a vote for your favorite candidate:

Rudy Giuliani (45848)
Mike Huckabee (58394)
Ron Paul (48392)
Mitt Romney (87489)
John McCain (43092)
Fred Thompson (48594)


One of my chief problems with phone polling is that they have to contact those being polled in order to get the results. This creates at least two problems:

1) People like me who primarily use a cell phone and don't answer numbers that I don't recognize on the home phone caller ID never get their opinions registered.

2) People don't wait for a phone call in order to vote. They have to actually take action in order to vote. Primary voters take action based on motivation, whether that motivation be a sense of duty, or a sense of enthusiasm for their candidate. Opinion polling neglects this key aspect.

The billboard idea won't really work because it relies on people to self-select.

Here's an alternative that takes into account the things you're concerned with. Knock on doors for your random sample instead of making phone calls. At each address, explain that your poll is going to evaluate real support for the various candidates and to register their support they will need to write a one page essay about why they like candidate X and return it using a prepaid envelope through the USPS by a certain date. That would measure the intensity of the support (by requiring a substantial effort to get their data point counted) and it wouldn't suffer from the self-selection problem of your billboard concept.

[Taco John]

The billboard idea won't really work because it relies on people to self-select.

Isn't that how people vote? A real election day poll is not a scientific cross selection of the country. It's a self selection.

[patteeu]

Isn't that how people vote? A real election day poll is not a scientific cross selection of the country. It's a self selection.

The problem with your plan is that it won't be the same people self-selecting. It may be that the methodologies that pollsters use to identify "likely voters" are flawed, but letting people self-select in the way you describe practically guarantees a flawed result. Any kind of accurate result would be a coincidence.

[Cochise]

The problem with your plan is that it won't be the same people self-selecting. It may be that the methodologies that pollsters use to identify "likely voters" are flawed, but letting people self-select in the way you describe practically guarantees a flawed result. Any kind of accurate result would be a coincidence.

People shouldn't be able to insert themselves into a sample. By that I mean, you shouldn't ask to be selected. It should be random, such as - your phone rings, someone asks if if you'd like to participate in a short survey about race X.

It shouldn't be something I can vote in 50 times, then go to my candidate's website and whip up a frenzy of others to all vote 50 times, or go to a website that will send X number of automated messages at the poll as well. Participation should be closed to people who are approached at random.