I Hate Polls!

[longknife]

Don't you?



For goodness sake, the first actually voting of any kind doesn't have until early NEXT YEAR! To me, this seems to be nothing but media hype, especially so their Talking Heads can give all of us their “expert opinion,” which usually turns out to be so far off base, they're out in left field.

 

[Redfish]

The polls are designed to influence public opinion, not to report on it.

It is statistically ridiculous to claim that a poll of 1000 people accurately reflects the opinions of 330,000,000 people.

its a game. the only winners are the pollsters who are paid to "do a poll for me that proves X".

 

[DGS49]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

 

[Redfish]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

 

[longknife]

I don't care what they say about the "statistical reliability" of polls. I don't care who they are run by, they are all governed by at least two things that keep them from truly being accurate.

How many people bother to pick up the phone and then allow themselves to be bothered by someone asking them questions about things they may not be all that interested in?

On the other side, perhaps it is ONLY the politically active who respond.

And how many answer the questions OPPOSITE of what they think just to screw up the polls?

 

[Misty]

The first poll I ever took was online about 2 weeks ago.

I didn't win a prize. Lol

 

[paddymurphy]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

 

[Redfish]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

You mean like the ones that said Romney would win?

 

[paddymurphy]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

You mean like the ones that said Romney would win?

No, dumbass. They were few. The study of polling from 5

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

You mean like the ones that said Romney would win?

The vast majority of polls were within the margin of error.

 

[jwoodie]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

Statistical sampling is only valid when applied to a static population. Public opinion polls are even less reliable because of the human interaction involved in obtaining the sample. Political preference polls are worse yet, since they tend to be self selective and are greatly affected by name recognition at a particular point in time.

 

[Toro]

The polls are designed to influence public opinion, not to report on it.

It is statistically ridiculous to claim that a poll of 1000 people accurately reflects the opinions of 330,000,000 people.

its a game. the only winners are the pollsters who are paid to "do a poll for me that proves X".

Didn't they teach you statistics at Harvard? Or were you a poet?

 

[Toro]

I don't care what they say about the "statistical reliability" of polls. I don't care who they are run by, they are all governed by at least two things that keep them from truly being accurate.

How many people bother to pick up the phone and then allow themselves to be bothered by someone asking them questions about things they may not be all that interested in?

On the other side, perhaps it is ONLY the politically active who respond.

And how many answer the questions OPPOSITE of what they think just to screw up the polls?

Do you have any empirical evidence to back this up?

It would be an odd assertion, given that not only political parties but also corporations spend mountains of money on statistical sampling to understand how people think.

 

[Redfish]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

You mean like the ones that said Romney would win?

No, dumbass. They were few. The study of polling from 5

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

You mean like the ones that said Romney would win?

The vast majority of polls were within the margin of error.

an educated guess would be within the margin of error.

Here's mine. the republican candidate will win in 2016 by +- 5% of the popular vote. You don't need a pollster to tell us what we already know.

 

[PaintMyHouse]

Don't you?

I like polls but as with all numbers you have to be careful how you use them...

 

[Redfish]

The polls are designed to influence public opinion, not to report on it.

It is statistically ridiculous to claim that a poll of 1000 people accurately reflects the opinions of 330,000,000 people.

its a game. the only winners are the pollsters who are paid to "do a poll for me that proves X".

Didn't they teach you statistics at Harvard? Or were you a poet?

Yes, they did and what I said is statistically correct. Figures don't lie but liars figure.

a 5% random sample is needed for a poll to be statistically meaningful. The pollsters try to get around that by saying their sample is not random but proportionally represents every faction in the total population. For example if 1% of the population is gay males who support the right to life then they must have 1 of them in every hundred sampled, and so forth for every possible faction in the country. Its foolishness, but you are free to buy into it if you want to.

 

[bendog]

Don't you?



For goodness sake, the first actually voting of any kind doesn't have until early NEXT YEAR! To me, this seems to be nothing but media hype, especially so their Talking Heads can give all of us their “expert opinion,” which usually turns out to be so far off base, they're out in left field.

The Polls are annoying as hell, but I reserve my real distain for Russians. (-:

 

[Old Rocks]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

Man, I have taken more than Stat 101, and you are completely wrong. Speak on things you actually know something about, if there is such a thing, because you obviously know nothing of statistics.

 

[Old Rocks]

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

You mean like the ones that said Romney would win?

No, dumbass. They were few. The study of polling from 5

"It is statistically ridiculous..."

If you really believe this you are a statistical ignoramous. A properly taken, un-biased poll of 100 people could EASILY come within a percent or two on a one-of-two question.

You can argue about the integrity of the pollsters, the difficulty of getting a true random sampling, or whether the questions asked are relevant, but random sampling is incredibly accurate, as anyone who has taken even a single stats course will attest.

I suggest that you go to your local library and check out a stat 101 text. Then read it. You will learn that for a sample to be truly representative of the total population it must be truly random and at least 5%. I understand how the pollsters claim that they overcome that by selecting their tiny samples to represent all factions, but its bullshit. A sample of 1000 out of 330,000,000 is STATISTICALLY RIDICULOUS. There is simply no way that they can proportionately represent all of the many factions in the US in a sample of 1000. Its a joke, but if you want to believe it, go right ahead. The pollsters are getting rich because of ignorance like yours.

I suppose, then, you can explain how close the polls have been to the actual outcome in most recent elections.

You mean like the ones that said Romney would win?

The vast majority of polls were within the margin of error.

an educated guess would be within the margin of error.

Here's mine. the republican candidate will win in 2016 by +- 5% of the popular vote. You don't need a pollster to tell us what we already know.

LOL. And what did you already know prior to 6Nov12. LOL

 

[g5000]

You mean like the ones that said Romney would win?

Only the voices in the heads of retards said that. I can show you some of the tards around here who predicted a landslide, if you like. Weren't you one of them?

2012 - General Election: Romney vs. Obama | RealClearPolitics

 

[jwoodie]

Stratified sampling is not useful when the population cannot be exhaustively partitioned into disjoint subgroups (i.e., no common elements). It is a misapplication of the technique to make subgroups' sample sizes proportional to the amount of data available from the subgroups, rather than scaling sample sizes to subgroup sizes.