NationStates Jolt Archive

(What Irony) A Poll regarding the accuracy of polls. (Poll coming soon)

Do you feel that Polls are accurate or inaccurate? What are your opinions of polls? Do they really work and project the majority? Please explain.

example poll: (Sorry, I couldn't find the source)
Poll by Zogby Intn'l:
result: 72% of U.S. soldiers want to "Leave Iraq".
We polled 944 U.S. soldiers +/-3% across Iraq.
---------------------------------------------------------------
Now, by using their numbers, more results with 100% math accuracy. My results are........that only 0.67% of U.S. soldiers were polled. How did I get that number? Well, I took the number of U.S. soldiers currently serving in Iraq, which is 141,000 times the proper percentage of 0.67% (0.67%=944.7 U.S. soldiers polled).

Now, that 72%, is of the 944 soldiers polled equals 679.68 U.S. soldiers. So, how can this be accurate when there is less than 1% surveyed? I believe that polls can be very misleading. I feel that they don't represent the truth.

Maybe, for this reason, President Bush says he doesn't pay to much attention to poll statistics due to misleading/inaccurate results.

I don't feel that the results of the above mentioned poll fairly represents the majority of the U.S. troops currently serving in Iraq. I also believe that some of these type polls are used to undermine the U.S. Army and our Government. They, seemingly, are jumping into the mad-media and is being used as some sort of ammunition to influence their hatefulness.

Specificly, that poll is nothing more than dog dirt.

God Bless America!!!

are you still on about this`? :rolleyes:

I don't have much faith in Polls or statistics...

There was a picture in my psychology textbook that I think relates to this.

Woman: Hey, Bob! According to this new poll, 93% of the American people believe that polls have no statistical validity whatsoever!

Man: Well, gosh! If that's what everyone thinks, then it must be true!

It was apparently from This Modern World by Tom Tomorrow in 1991.

are you still on about this`? :rolleyes:

I was inspired to make a poll. It nearly hijacked the earlier thread and that was not right. So, Yes, I suppose I am on about this.

It needed it's own thread, so I created it's own thread.

Is there a problem with a poll regarding poll accuracy? I think not and if you don't like it then why did you waste your time commenting on it? If you don't wish to participate, then don't. Nobody is forcing your fingers. With all due respect, just ignore this thread instead of bickering about. Thank you.

Polls are really meant as simply a cross-section of the population's opinion, which is why every poll has a certain percentage reserved for "margin of error".

Polls are really meant as simply a cross-section of the population's opinion, which is why every poll has a certain percentage reserved for "margin of error".

I'm not disagreeing with you and I know the "margin of error", however, a lot of the media uses these polls to mislead the public for their biased agenda.

So, I was just curious, whether or not, NSG felt they are accurate or not.

Thanks for your post.

I'm not disagreeing with you and I know the "margin of error", however, a lot of the media uses these polls to mislead the public for their biased agenda.

So, I was just curious, whether or not, NSG felt they are accurate or not.

Thanks for your post.
WHen doesn't the media mislead the public when its to their benefit?

WHen doesn't the media mislead the public when its to their benefit?

That's a very good point. The media enjoys stirring the pot. They use these polls for ammunition for their biased agenda.

That's like asking how long a piece of string is.

The accuracy depends entirely on the poll itself. A non partisan, scientific poll of a representative sample is pretty accurate. A poll with an agenda that only asks people of the same opinion is not.

I will repeat what I said in the other thread:

Let me explain to your where your logic train ran off track.

Let us say you have a giant bag filled with poker chips. You can't see the poker chips in the bag, but you know two things:

1) you know there are 10,000 chips in the bag

2) each chip is either blue or yellow

You do not know the proper ratio, this is roughly synonomous with the poll here, we know the number of soldiers (number of chips in the bag) and we know the two options are either stay, or go (blue, or yellow) we just don't know how many of each.

So let's say you dip your hand into the bag and pull out 100 of those chips. 90 are blue, 10 are yellow. Again this is similar to the poll, we take a part of the whole, and ask what they think.

Now the reasonably intelligent person (so, not you, but let me explain what an actual thinking human being would do) would go "hmm, I know there are 10,000, I know I pulled 100, and of those 100 90 are blue 10 are yellow, so if I look at this as a representative sample, it is pretty safe to assume that of this 10,000, about 9,000 are blue and about 1,000 are yellow. It may not be exact, but this is what the sample seems to suggest".

Again back to the poll, a reasonable sample, reasonably diversified can create a decent sample of the whole. Not 100% representative, but a general idea.

You on the other hand seem to think "all the chips left in the bag are yellow, since we pulled out all the blue"

You kinda see the difference?

In my experience, only the 100% accurate polls are 100% accurate.

Do you feel that Polls are accurate or inaccurate? What are your opinions of polls? Do they really work and project the majority? Please explain.Keyword is project. It's like the weather report: a prediction. Depending on who does the poll (due to their methodology) indicates how accurate it is.

example poll: (Sorry, I couldn't find the source)
Poll by Zogby Intn'l:
result: 72% of U.S. soldiers want to "Leave Iraq".
We polled 944 U.S. soldiers +/-3% across Iraq. Wow. What a coincidence.

Anyway, what is the question? Do 72% of US soldiers want to leave Iraq or do they want the US to leave Iraq?

Now, that 72%, is of the 944 soldiers polled equals 679.68 U.S. soldiers. So, how can this be accurate when there is less than 1% surveyed? I believe that polls can be very misleading. I feel that they don't represent the truth.It's simple really. If you can manage to ensure that everyone has a chance for being randomly selected, you don't sample from one specific area, and take other precautions regarding statistics, you get a representative result. How can it be accurate if less than 1% are surveyed? If the ratio of opinions is the same in the selected group as in the whole group, then the result is accurate. You can never know this for sure, but you can take steps to try and remove as much bias as possible from the result.

Your feelings may be right, I have no clue about the methodology of this poll. But I doubt you have that much more information than I do to come to a conclusive opinion that this poll is biased.

Maybe, for this reason, President Bush says he doesn't pay to much attention to poll statistics due to misleading/inaccurate results.Ah, so there is an agenda behind this simple debate on statistics and sampling. Good to know.

I don't feel that the results of the above mentioned poll fairly represents the majority of the U.S. troops currently serving in Iraq. I also believe that some of these type polls are used to undermine the U.S. Army and our Government. They, seemingly, are jumping into the mad-media and is being used as some sort of ammunition to influence their hatefulness.

Specificly, that poll is nothing more than dog dirt.You need a bit more than just a feeling to debunk statistics.

I will repeat what I said in the other thread:

Let me explain to your where your logic train ran off track.

Let us say you have a giant bag filled with poker chips. You can't see the poker chips in the bag, but you know two things:

1) you know there are 10,000 chips in the bag

2) each chip is either blue or yellow

You do not know the proper ratio, this is roughly synonomous with the poll here, we know the number of soldiers (number of chips in the bag) and we know the two options are either stay, or go (blue, or yellow) we just don't know how many of each.

So let's say you dip your hand into the bag and pull out 100 of those chips. 90 are blue, 10 are yellow. Again this is similar to the poll, we take a part of the whole, and ask what they think.

Now the reasonably intelligent person (so, not you, but let me explain what an actual thinking human being would do) would go "hmm, I know there are 10,000, I know I pulled 100, and of those 100 90 are blue 10 are yellow, so if I look at this as a representative sample, it is pretty safe to assume that of this 10,000, about 9,000 are blue and about 1,000 are yellow. It may not be exact, but this is what the sample seems to suggest".

Again back to the poll, a reasonable sample, reasonably diversified can create a decent sample of the whole. Not 100% representative, but a general idea.

You on the other hand seem to think "all the chips left in the bag are yellow, since we pulled out all the blue"

You kinda see the difference?

I sure do.

I'm not disagreeing with you and I know the "margin of error", however, a lot of the media uses these polls to mislead the public for their biased agenda.

So, I was just curious, whether or not, NSG felt they are accurate or not.

Thanks for your post.

Because obviously, the pollsters did their work with a disporportionately cowardly bunch - what, who? The Catering Corps? Traitors to a man...

*Mr. Bodacious, please come to the Customer Service desk - your Tinfoil Hat is ready*

Okay with that concept, I can use NY City or California as an example (in regards to CA, They do have a republican governor due to Davis's complete screw up but CA is by majority what you would call a Blue state and also NY, Blue state) Now, with that being said, a pollster goes to LA, NY City, Boston, Chicago, Seattle (all Blue, mind you) but they also go to per se Miami (blue city in a red state) but so not to look biased but more fair they go to Dallas, New Orleans, Charlotte, and Nashville.

Five cities in Blue States and five cities in Red States. Look at the populations in those cities and please try and use some common sense. As you said, it's the Luck of the draw and as far as I am concerned, Luck can NOT be in the place of accuracy.

Keyword is project. It's like the weather report: a prediction. Depending on who does the poll (due to their methodology) indicates how accurate it is.

Wow. What a coincidence.

Anyway, what is the question? Do 72% of US soldiers want to leave Iraq or do they want the US to leave Iraq?

It's simple really. If you can manage to ensure that everyone has a chance for being randomly selected, you don't sample from one specific area, and take other precautions regarding statistics, you get a representative result. How can it be accurate if less than 1% are surveyed? If the ratio of opinions is the same in the selected group as in the whole group, then the result is accurate. You can never know this for sure, but you can take steps to try and remove as much bias as possible from the result.

Your feelings may be right, I have no clue about the methodology of this poll. But I doubt you have that much more information than I do to come to a conclusive opinion that this poll is biased.
Ah, so there is an agenda behind this simple debate on statistics and sampling. Good to know.

You need a bit more than just a feeling to debunk statistics.

Okay, well, common sense tells me that a 0.67% does not fairly represent the entire 141,000 troops.

Okay with that concept, I can use NY City or California as an example (in regards to CA, They do have a republican governor due to Davis's complete screw up but CA is by majority what you would call a Blue state and also NY, Blue state) Now, with that being said, a pollster goes to LA, NY City, Boston, Chicago, Seattle (all Blue, mind you) but they also go to per se Miami (blue city in a red state) but so not to look biased but more fair they go to Dallas, New Orleans, Charlotte, and Nashville.

Five cities in Blue States and five cities in Red States. Look at the populations in those cities and please try and use some common sense. As you said, it's the Luck of the draw and as far as I am concerned, Luck can NOT be in the place of accuracy.Yes. Luck is all about accuracy. Random selection is what it's about. If you have questions about it, ask Arthais. He seems to know more about statistics than the rest of us. Just don't come to premature conclusions based on your ignorance of statistics.

Okay, well, common sense tells me that a 0.67% does not fairly represent the entire 141,000 troops.You are wrong. 0.67% can very well represent the entire 141,000 troops if picked under the right conditions. It's how they've been picking them that we need to know in order to consider this poll inaccurate.

Yes. Luck is all about accuracy. Random selection is what it's about. If you have questions about it, ask Arthais. He seems to know more about statistics than the rest of us. Just don't come to premature conclusions based on your ignorance of statistics.

I don't feel it's my "ignorance of statistics". I like to call it common sense of the statistics. As for the debate, nobody has shown any proof of them being accurate. The numbers do not add up to accuracy.

King Bodacious, if it is possible, I would suggest taking a course or reading a book on statistics. Before last year, I had a similar view of polls as you do, and after taking a statistics course, I just how accurate many polls can be, even with a small sample size. That being said, I don't know the details of the poll conducted in your example, so I can't say if it is accurate or not.

You are wrong. 0.67% can very well represent the entire 141,000 troops if picked under the right conditions. It's how they've been picking them that we need to know in order to consider this poll inaccurate.

How am I wrong? The poll specifically said 944 U.S. soldiers were surveyed/polled. Well, by doing a little math, that 944 soldiers represent 0.67% of the current 141,000 troops. So, tell me again, how is that wrong?

you can make statistics read anyway you want.

just have to ask the right people and inculde only the right information.

I don't feel it's my "ignorance of statistics". I like to call it common sense of the statistics. As for the debate, nobody has shown any proof of them being accurate. The numbers do not add up to accuracy.

Out of curiosity, do you really expect the companies who do these survey to survey everybody in the demographuc being surveyed? Because if you do, you should know that that would be prohibitively expensive in most cases.

I don't feel it's my "ignorance of statistics". I like to call it common sense of the statistics. Have you actually had any in depth exposure to the field of statistics before?

How am I wrong? The poll specifically said 944 U.S. soldiers were surveyed/polled. Well, by doing a little math, that 944 soldiers represent 0.67% of the current 141,000 troops. So, tell me again, how is that wrong?You're assuming that that is too small a number to give an accurate picture. It isn't necessarily.

How am I wrong? The poll specifically said 944 U.S. soldiers were surveyed/polled. Well, by doing a little math, that 944 soldiers represent 0.67% of the current 141,000 troops. So, tell me again, how is that wrong?

It's pretty obvious that you just don't get it. At some later date, when you have a better understanding of how this all works, I think you're going to be feeling a little sheepish about this thread, Mr. Bodacious.

How am I wrong? The poll specifically said 944 U.S. soldiers were surveyed/polled. Well, by doing a little math, that 944 soldiers represent 0.67% of the current 141,000 troops. So, tell me again, how is that wrong?

because a small sample, if picked to be an accurately representative sample can serve as a...well...accurately representative sample.

That is the fundamental principle behind statistical analysis, a principle that has been tested and upheld for as long as there has been statistical mathematics, and a principle you seem blatantly and willfully ignorant of.

because a small sample, if picked to be an accurately representative sample can serve as a...well...accurately representative sample.

That is the fundamental principle behind statistical analysis, a principle that has been tested and upheld for as long as there has been statistical mathematics, and a principle you seem blatantly and willfully ignorant of.Put it better than I could :p

I need mouthwash.

Okay, well, common sense tells me that a 0.67% does not fairly represent the entire 141,000 troops.

hundreds of years of tried and tested statistical mathematic prinicples have proved you wrong.

because a small sample, if picked to be an accurately representative sample can serve as a...well...accurately representative sample.

That is the fundamental principle behind statistical analysis, a principle that has been tested and upheld for as long as there has been statistical mathematics, and a principle you seem blatantly and willfully ignorant of.

and a small fact that you seem very blatantly and willfully ignorant of is that you can make a statistic read anyway you want it to read, and im sure people whos job it is to take statistics know exactly how to do so.

and a small fact that you seem very blatantly and willfully ignorant of is that you can make a statistic read anyway you want it to read, and im sure people whos job it is to take statistics know exactly how to do so.

a statistic does not "read" in any way. I statistic simply is. Statistics does not seek to interpret data, it simply seeks to present that data and lead the interpretation of it up to people in that field.

Now if you're saying that statisticians can INTENTIONALLY ask questions/chose samples to skew the data with bias, well of COURSE it can, however please see my point where I said "accurately representative sample". If you skew the sample, it's no longer accureately representative.

Once again, statistics dont "read" anything, statistical analysis is a mere presentation of data, nothing more, it does not seek to explain, elaborate upon, or analyse WHAT that data means in the slightest.

People who think it does don't understand what statistics are.

and a small fact that you seem very blatantly and willfully ignorant of is that you can make a statistic read anyway you want it to read, and im sure people whos job it is to take statistics know exactly how to do so.You can, but that doesn't debunk statistics as a whole. Population ecology would be utterly useless if we didn't have statistics, and we'd have to wait until we have those nifty Star Trek scanners before you could do any work on that.

It's pretty obvious that you just don't get it. At some later date, when you have a better understanding of how this all works, I think you're going to be feeling a little sheepish about this thread, Mr. Bodacious.

What's with your squacking? The point of these threads in NSG is to debate. or I thought. I didn't realize these forums were meant for people like you to say, "You're wrong, I'm Right, Go home now. Debate over." Either debate this issue or don't post. Don't just tell me I'm wrong with no facts backing it up or give me opinionated, fictious thinking. That doesn't win "debates".

Nobody, has given me any factual details of why I am wrong in my common sense approach to the regards of "poll accuracy".

How am I wrong? The poll specifically said 944 U.S. soldiers were surveyed/polled. Well, by doing a little math, that 944 soldiers represent 0.67% of the current 141,000 troops. So, tell me again, how is that wrong?

There are people who do research to find out the exact number of people you need to poll to get an accurate sample of the true population with a certain margin of error. For America, the number of people often polled is around 1,000, which may not seem to be a high number, but it is a balance of accuracy and practicality. It is not 100% accurate, and no one should claim that it is 100 % accurate, but I think, that the way the 1,000 number was determined is a formula that proved that 95% of the time a poll of American citizens is conducted with 1000(ish) people, the true mean will fall in the margin of error that they provide, or something like that. I could look up the actual formula, but that would take to long. However, if they applied that formula to the number of troops, and they came up with 944, then I would guess it is accurate. Honestly, the number of people they polled seems slightly high, if I remembered the 1000 people mark accurately (I'd say I'm 95% sure its correct:p ).

What's with your squacking? The point of these threads in NSG is to debate. or I thought. I didn't realize these forums were meant for people like you to say, "You're wrong, I'm Right, Go home now. Debate over." Either debate this issue or don't post. Don't just tell me I'm wrong with no facts backing it up or give me opinionated, fictious thinking. That doesn't win "debates".

Nobody, has given me any factual details of why I am wrong in my common sense approach to the regards of "poll accuracy".

funny how you have yet to reply to this:

Let us say you have a giant bag filled with poker chips. You can't see the poker chips in the bag, but you know two things:

1) you know there are 10,000 chips in the bag

2) each chip is either blue or yellow

You do not know the proper ratio, this is roughly synonomous with the poll here, we know the number of soldiers (number of chips in the bag) and we know the two options are either stay, or go (blue, or yellow) we just don't know how many of each.

So let's say you dip your hand into the bag and pull out 100 of those chips. 90 are blue, 10 are yellow. Again this is similar to the poll, we take a part of the whole, and ask what they think.

Now the reasonably intelligent person (so, not you, but let me explain what an actual thinking human being would do) would go "hmm, I know there are 10,000, I know I pulled 100, and of those 100 90 are blue 10 are yellow, so if I look at this as a representative sample, it is pretty safe to assume that of this 10,000, about 9,000 are blue and about 1,000 are yellow. It may not be exact, but this is what the sample seems to suggest".

Again back to the poll, a reasonable sample, reasonably diversified can create a decent sample of the whole. Not 100% representative, but a general idea.

You on the other hand seem to think "all the chips left in the bag are yellow, since we pulled out all the blue"

You kinda see the difference?

Okay with that concept, I can use NY City or California as an example (in regards to CA, They do have a republican governor due to Davis's complete screw up but CA is by majority what you would call a Blue state and also NY, Blue state) Now, with that being said, a pollster goes to LA, NY City, Boston, Chicago, Seattle (all Blue, mind you) but they also go to per se Miami (blue city in a red state) but so not to look biased but more fair they go to Dallas, New Orleans, Charlotte, and Nashville.

Five cities in Blue States and five cities in Red States. Look at the populations in those cities and please try and use some common sense. As you said, it's the Luck of the draw and as far as I am concerned, Luck can NOT be in the place of accuracy.

I did respond to your bogus "luck" thoughts compared to poll accuracy. With my own "bogus" rebuttal. Pay attention.

I also started this thread to see if people would be able to educate me on the reasons "why poll's are accurate" but instead of giving facts and to educate, you have decided to slander as "ignorance of statistics" and yet you refuse to give me any factual input. You just give your own "bogus" Luck comparisons. Also, your style of debate is to slander.

Continue on......

Also, if you have forgotten, let's go back to the poll. Answers were are polls 100% accurate, a bunch of BS, or other.

As for your slander towards my "ignorance of statistics" or my "willingful ignorance", I am left to believe you voted for the 100% accuracy. Which I do not feel that polls can fairly represent a 100% accuracy.

I did respond to your bogus "luck" thoughts compared to poll accuracy. With my own "bogus" rebuttal. Pay attention.

I also started this thread to see if people would be able to educate me on the reasons "why poll's are accurate" but instead of giving facts and to educate, you have decided to slander as "ignorance of statistics" and yet
you refuse to give me any factual input.

Fine, here you go, canadian ministry of agriculture giving details on how to take representative samples of crops in order to get an accurate picture of crop yields:

http://www.grainscanada.gc.ca/pubs/FactsFarm/factsfarmers13-e.htm

Here's a pdf article explaining the use of representative samples in statistics, entitled: Organizational research: Determining appropriate sample size for survey research.

http://www.osra.org/itlpj/bartlettkotrlikhiggins.pdf

Here's a wiki link on the subject:

http://en.wikipedia.org/wiki/Sampling_(statistics)

Here's some formula on probability statistics, and a bio of a statistician who developed the formulas:

http://en.wikipedia.org/wiki/Laplace



You just give your own "bogus" Luck comparisons. Also, your style of debate is to slander.

In text it's libel. It's also not libel if it's true. Therefore calling you an idiot isn't libel.

Continue on......

OK, since you asked. You're an idiot.

Also, if you have forgotten, let's go back to the poll. Answers were are polls 100% accurate, a bunch of BS, or other.

As for your slander towards my "ignorance of statistics" or my "willingful ignorance"

I voted "other", as no poll, unless it contained a full 100% sample size, can be 100% accurate.

I am left to believe you voted for the 100% accuracy. Which I do not feel that polls can fairly represent a 100% accuracy.

I am not an idiot, unlike you, and am not likely to fall into stupid word tricks. Polls are not 100% accurate and I never claimed otherwise. They are accurate based on a statistically calculated margin of error. The fact that you would actually think I would fall into such a stupid word trap only further proves the obvious gap between us.

I did respond to your bogus "luck" thoughts compared to poll accuracy. With my own "bogus" rebuttal. Pay attention.

I also started this thread to see if people would be able to educate me on the reasons "why poll's are accurate" but instead of giving facts and to educate, you have decided to slander as "ignorance of statistics" and yet you refuse to give me any factual input. You just give your own "bogus" Luck comparisons. Also, your style of debate is to slander.

Continue on......

Gee, the opportunity to tutor a sulking ideologue with a political axe to grind. Be still my heart. And just how do you plan to compensate the bastard miserable enough to see to your education - 'cause cookies wouldn't begin to cover 0.67 % of it...

I also started this thread to see if people would be able to educate me on the reasons "why poll's are accurate"

Most people on NS aren't teachers. If you want to be enducated about statistics then perhaps you should find a forum on statistical mathematics.

Also, if you have forgotten, let's go back to the poll. Answers were are polls 100% accurate, a bunch of BS, or other.

As for your slander towards my "ignorance of statistics" or my "willingful ignorance", I am left to believe you voted for the 100% accuracy. Which I do not feel that polls can fairly represent a 100% accuracy.

I do not think you know much about statistics, but I had to vote other, cause it's not 100% accurate, polls, by statistical law, are not 100% accurate. I would guess it's a bunch of BS either, unless they had horrible questions, or made up results, or did not pick an accurate sample. I had to pick other cause that was the most resonable option.

edit: Yay for Arthais101!

Also, if you have forgotten, let's go back to the poll. Answers were are polls 100% accurate, a bunch of BS, or other.

As for your slander towards my "ignorance of statistics" or my "willingful ignorance", I am left to believe you voted for the 100% accuracy. Which I do not feel that polls can fairly represent a 100% accuracy.KB, you are ignorant of how stastics work; you have shown that and continue to do so. We've been patiently telling you that you are wrong and tried to explain statistics to you to help you rid yourself of your misconceptions.

Instead of debating, shall we resort to name calling, how old are we? A few of you prefer to resort to immature behavior instead of doing what is supposed to be done here, to debate.

So, basicly, as I understand it, since I disagree or I do not understand the reasoning behind your views, I am automaticly declared an idiot.

Now, that is some rational thoughts and declarations. You have convinced me of the proper debate tactics.

KB, you are ignorant of how stastics work; you have shown that and continue to do so. We've been patiently telling you that you are wrong and tried to explain statistics to you to help you rid yourself of your misconceptions.

which is why at this point I've given up and started calling him an idiot.

Instead of debating, shall we resort to name calling, how old are we? A few of you prefer to resort to immature behavior instead of doing what is supposed to be done here, to debate.

So, basicly, as I understand it, since I disagree or I do not understand the reasoning behind your views, I am automaticly declared an idiot.

Now, that is some rational thoughts and declarations. You have convinced me of the proper debate tactics.1 + 1 = 2 isn't something you should debate. There isn't really another side to the argument. We're not debating with you, we're trying to educate you.

Instead of debating, shall we resort to name calling, how old are we? A few of you prefer to resort to immature behavior instead of doing what is supposed to be done here, to debate.

So, basicly, as I understand it, since I disagree or I do not understand the reasoning behind your views, I am automaticly declared an idiot.

Now, that is some rational thoughts and declarations. You have convinced me of the proper debate tactics.

a debate, by definition, is two sides arguing a point they believe in.

By your own admission you sought to be "educated" on the subject, ergo you don't know about it to form an opinion.

Either you are asking for information, or debating a point. Don't try to say you're doing both, it's a logical inconsistancy. A debate is to present one view point and argue for it, which you can't do if you're asking to be taught the very subject you're arguing about.

Now, we have given you examples, which you have rejected. And since you do not understand the subject, and are seemingly incapable of understanding the subject, a subject that is in fact actually pretty simple in basic idea, I am left with one conclusion.

You are an idiot.

And that's not how you debate, but by your own words you aren't debating anything. That's what you say when you conclude the person you are talking with is mentally lacking.

And having concluded that you are mentally lacking, I call you an idiot, yet again.

Instead of debating, shall we resort to name calling, how old are we? A few of you prefer to resort to immature behavior instead of doing what is supposed to be done here, to debate.

So, basicly, as I understand it, since I disagree or I do not understand the reasoning behind your views, I am automaticly declared an idiot.

Now, that is some rational thoughts and declarations. You have convinced me of the proper debate tactics.

You said yourself, you started this thread to see if someone could educate you about statistics, not to debate.

Actually you a declared an idiot because it has been shown that you are ignorant of statistical mathematics(as am I btw) but refuse to accept this and simply claim that your math was 100% accurate.

Most people on NS aren't teachers. If you want to be enducated about statistics then perhaps you should find a forum on statistical mathematics.

To set the Record Straight:

I am aware that "Most people on NS aren't teachers", though, possibly they could be, however, I do believe that a "Good Debate" can be very educational. When debates turn to slander, they do nothing more than cause tensions to rise, on NSG it could be classified as "Flame-baiting".

http://en.wikipedia.org/wiki/Sampling_(statistics)


You missed the parenthesis.

http://en.wikipedia.org/wiki/Sampling_(statistics)

KB, click on all of the other links in that post. Please

Every poll is one hundred percent accurate all of the time.

Now, don't get me wrong. I like forum polls, cause they're fun.
I don't much care for the news polls though because any one of them can be designed to decieve with selected pollees that you are unaware of.

To set the Record Straight:

I am aware that "Most people on NS aren't teachers", though, possibly they could be, however, I do believe that a "Good Debate" can be very educational. When debates turn to slander, they do nothing more than cause tensions to rise, on NSG it could be classified as "Flame-baiting".

Then why did you start a thread expecting to be taught when there's only a possibility of anyone being able to do so? Seems to me that the obvious thing to do would be to find a forum on statistical mathematics and pose your questions there.

And this isn't really a debate, you're asking us to show you how you are mistaken about the accuracy of that poll about how many US soldiers want to leave Iraq(or something)

You missed the parenthesis.

http://en.wikipedia.org/wiki/Sampling_(statistics)

KB, click on all of the other links in that post. Please

Good catch, fixed.

Good catch, fixed.

Post of DEATH!

Good catch, fixed.
No problem, good links.

Post of DEATH!

Death to what?

No problem, good links.



Death to what?

The thread. Though now it seems to be struggling into the spam stage of life.

In my OP, I thought it was clearly explained that I, myself, am debating "100% accuracy of polls". In earlier posts people were claiming that they were completely accurate, I disagree....

While browsing the internet I found another reason...... http://www.spacetownusa.com/2004/09/death-of-polls.html

Call me an idiot if you will. I am going to continue my stand of rebute when it amounts to 100% poll accuracy.

In my OP, I thought it was clearly explained that I, myself, am debating "100% accuracy of polls". In earlier posts people were claiming that they were completely accurate, I disagree....

While browsing the internet I found another reason...... http://www.spacetownusa.com/2004/09/death-of-polls.html

Call me an idiot if you will. I am going to continue my stand of rebute when it amounts to 100% poll accuracy.

Like I said on the first page (http://forums2.jolt.co.uk/showpost.php?p=11811562&postcount=12), only the polls that are 100% accurate are 100% accurate. Like this poll for example.

I, also, came about this....... http://www.freerepublic.com/focus/f-news/1004883/posts

In my OP, I thought it was clearly explained that I, myself, am debating "100% accuracy of polls". In earlier posts people were claiming that they were completely accurate, I disagree....

While browsing the internet I found another reason...... http://www.spacetownusa.com/2004/09/death-of-polls.html

Call me an idiot if you will. I am going to continue my stand of rebute when it amounts to 100% poll accuracy.

WE (AT LEAST I) ARE NOT CLAIMING THEY ARE 100% ACCURATE. THEY ARE NOT EVER GOING TO BE 100% ACCURATE.


And to that guy, they don't only use lists from phone books or whatever. Good polls will use random digits to create phone number that may not be listed based with random area codes. Of course there is going to be slight bias in who actually answers the polls, but there is very little that can be done about that, as those people are part of the overall population as well.

In my OP, I thought it was clearly explained that I, myself, am debating "100% accuracy of polls". In earlier posts people were claiming that they were completely accurate, I disagree....Point them out to me. No one that knows statistics would ever say that. "Accurate" and "completely accurate" are two separate cookie jars.
Call me an idiot if you will. I am going to continue my stand of rebute when it amounts to 100% poll accuracy.Strawman. No one has been arguing 100% accuracy of polls.

I, also, came about this....... http://www.freerepublic.com/focus/f-news/1004883/postsSo because one Republican fondles underage pages, they all do?

I, also, came about this....... http://www.freerepublic.com/focus/f-news/1004883/posts

As far as I can tell, the writer is saying that the polls that are taken closer to election time show what will happen in the election better. If that is what the author is saying, and that is all he is saying, then I say duh. If the author is saying more, and I'm very wrong, it's because I didn't read all of the article cause I don't have time.

Okay with that concept, I can use NY City or California as an example (in regards to CA, They do have a republican governor due to Davis's complete screw up but CA is by majority what you would call a Blue state and also NY, Blue state) Now, with that being said, a pollster goes to LA, NY City, Boston, Chicago, Seattle (all Blue, mind you) but they also go to per se Miami (blue city in a red state) but so not to look biased but more fair they go to Dallas, New Orleans, Charlotte, and Nashville.

Five cities in Blue States and five cities in Red States. Look at the populations in those cities and please try and use some common sense. As you said, it's the Luck of the draw and as far as I am concerned, Luck can NOT be in the place of accuracy.

Well of course, you're deliberately setting up a biased poll. You have no evidence that this is what others do.

Well of course, you're deliberately setting up a biased poll. You have no evidence that this is what others do.

Exactly, I was just, simply, comparing it to the earlier example of 2 colored poker chips to show how equally dumb it was.

My stance still continues because I don't see how any one can legitimately claim it to be fair. Fact: less than 1% does NOT equal 100%. 1% does NOT equal 50%. 1% does NOT equal 5%.

In earlier posts people were claiming that they were completely accurate

show me one.

Show me one person who in this thread, or the other, who claimed that polls are 100% accurate. Show me one time when someone said that.

My stance still continues because I don't see how any one can legitimately claim it to be fair. Fact: less than 1% does NOT equal 100%. 1% does NOT equal 50%. 1% does NOT equal 5%.

congratulations, you passed first grade math.

Unfortunatly you never managed to get to the grade where they teach statistics.

My stance still continues because I don't see how any one can legitimately claim it to be fair. Fact: less than 1% does NOT equal 100%. 1% does NOT equal 50%. 1% does NOT equal 5%.

I think you and "Dr." Gene Ray (http://www.timecube.com) have some common ground, y'know.

My stance still continues because I don't see how any one can legitimately claim it to be fair. Fact: less than 1% does NOT equal 100%. 1% does NOT equal 50%. 1% does NOT equal 5%.
More Power to you, then. It seems you need it.

Statistics is quite a large and involved area of mathematics. If you wish to understand it properly, you're going to have to go and learn about it. "Common sense" really doesn't cut it. Once you understand it you'll be able to decide whether or not a given survey is reasonable, without just either globally accepting all polls, or rejecting them as nonsense.

Before we can make any claim as to whether the survey in the OP is accurate or at all misleading we need to know a lot more information about it, including:

1) How they selected the sample to survey ( A proper, representative random sample across the entire force and country? Just the catering corps? Just those in Baghdad?). 944 people may well be a large enough sample of the population, but if you're only sampling, say, cooks in Baghdad then you probably can't extrapolate the data to include air-service personnel in Fallujah.

2) What was the exact question asked? For instance, was it "Would you like to leave Iraq?", "Do you think most US soldiers would like to leave Iraq?", "Should the USA remove its military forces from Iraq some day?", or perhaps "Should the USA remove its military forces from Iraq as soon as possible?"? I'm sure you can see the different responses you might get from those different questions.

"For every complex problem, there is a solution that is simple, neat, and wrong." - H.L. Mencken (1880-1956)

Exactly, I was just, simply, comparing it to the earlier example of 2 colored poker chips to show how equally dumb it was.

No, not really.
Take a course in statistics and probability in order to understand what others are telling you here.

If you flip a coin 100 times, pure math tells us that you will get 50 heads and 50 tails.
Sometimes you get 48 heads and 52 tails.
Sometimes you get 60 heads and 40 tails.
Sometimes (even more rarely) the coin will land edge on and be neither head nor tails.

Just because you say, well, the math says that it will come up heads half the time doesn't mean that's ACTUALLY what happens when you sit there and flip.

It's an estimate. It's always an estimate. And there is always a margin of error. It does not mean that there is anything nefarious indicated.

Let's look at your poll. Your poll is intentionally biased, because as we all know, all polls have a margin of error. Your choices are 100% accurate, Bullshit, or Other. Well, we know no poll is 100% accurate -- that's common sense. We know polls are not complete bullshit. Therefore, you've skewed this poll intentionally towrd other.

You set out to make a biased poll to make a point. You are assuming that everyone sets out to make a point and therefore sets up a biased poll. That is merely your feeling, and you've not proven (nor can you) that x number of polls are biased.

Is it really necessary for the negativity when it comes to one disagreeing with your own views?
Is it possible to conduct ourselves in a rational way?
Is it possible to refrain from the childish name calling acts?
Is it possible to except the fact, that there are people in this world who disagree with your views?

Yes, I am weighing common sense vs scientific polls. I do NOT feel that they "fairly represent" the whole. I consider it to be a false generalization.

I suppose the exit polls to the last presidential election were pretty accurate in showing Kerry was going to win the election.

My stance still continues because I don't see how any one can legitimately claim it to be fair. Fact: less than 1% does NOT equal 100%. 1% does NOT equal 50%. 1% does NOT equal 5%.And? 1/100 DOES equal 10/1000.

Polls depend on the accuracy of three things

A) sample and procedure (how large is your population, what do you do with the data after achieved)
B) Questioning (are the questions clear and not misleading)
C) Interpretation (Does the data say what it really means to say)


The poll the OP was referencing was really a failure to C … it was not (if I remember right) a failure of the process or the actual sample. Though really it might have been B too but there is not enough information to say there was a failure in A

Anyways polls can be highly accurate … exceptionally accurate but the trick … like with cars or any complex thing is to learn enough about it to interpret its accuracy yourself. Learn some basics about statistics. It will serve you well

My stance still continues because I don't see how any one can legitimately claim it to be fair. Fact: less than 1% does NOT equal 100%. 1% does NOT equal 50%. 1% does NOT equal 5%.

You have not gotten as far as statistics have you? We are talking about polls not surveys here.

A poll can be highly accurate … a sampling of 2000 people can EASILY represent the views of an entire nation to well under a 5% error rate.

Is it really necessary for the negativity when it comes to one disagreeing with your own views?Except that your views on how statistics work are plain false.
Is it possible to conduct ourselves in a rational way?Start being rational. If you want to learn about statistics, don't act as though you knew how they worked.
Is it possible to refrain from the childish name calling acts?Is it possible to stop the childish "I don't feel that statistics work this way, and I'm right in being wrong" acts?
Is it possible to except the fact, that there are people in this world who disagree with your views?If I found someone that disagreed with me on the outcome of 1 + 1, I doubt I'd react much differently.

Yes, I am weighing common sense vs scientific polls.No you're not.
I do NOT feel that they "fairly represent" the whole. That's possible, but only when considering the sampling methods. Other than that, it's only important that they accurately represent the whole.
I consider it to be a false generalization.Oh, the irony:
I suppose the exit polls to the last presidential election were pretty accurate in showing Kerry was going to win the election.

No, not really.
Take a course in statistics and probability in order to understand what others are telling you here.

If you flip a coin 100 times, pure math tells us that you will get 50 heads and 50 tails.
Sometimes you get 48 heads and 52 tails.
Sometimes you get 60 heads and 40 tails.
Sometimes (even more rarely) the coin will land edge on and be neither head nor tails.

Just because you say, well, the math says that it will come up heads half the time doesn't mean that's ACTUALLY what happens when you sit there and flip.

It's an estimate. It's always an estimate. And there is always a margin of error. It does not mean that there is anything nefarious indicated.

Let's look at your poll. Your poll is intentionally biased, because as we all know, all polls have a margin of error. Your choices are 100% accurate, Bullshit, or Other. Well, we know no poll is 100% accurate -- that's common sense. We know polls are not complete bullshit. Therefore, you've skewed this poll intentionally towrd other.

You set out to make a biased poll to make a point. You are assuming that everyone sets out to make a point and therefore sets up a biased poll. That is merely your feeling, and you've not proven (nor can you) that x number of polls are biased.
Well said simplification

(What Irony) A Poll regarding the accuracy of polls. (Poll coming soon)

Do you feel that Polls are accurate or inaccurate? What are your opinions of polls? Do they really work and project the majority? Please explain.

example poll: (Sorry, I couldn't find the source)
Poll by Zogby Intn'l:
result: 72% of U.S. soldiers want to "Leave Iraq".
We polled 944 U.S. soldiers +/-3% across Iraq.
---------------------------------------------------------------
Now, by using their numbers, more results with 100% math accuracy. My results are........that only 0.67% of U.S. soldiers were polled. How did I get that number? Well, I took the number of U.S. soldiers currently serving in Iraq, which is 141,000 times the proper percentage of 0.67% (0.67%=944.7 U.S. soldiers polled).

Now, that 72%, is of the 944 soldiers polled equals 679.68 U.S. soldiers. So, how can this be accurate when there is less than 1% surveyed? I believe that polls can be very misleading. I feel that they don't represent the truth.

Maybe, for this reason, President Bush says he doesn't pay to much attention to poll statistics due to misleading/inaccurate results.

I don't feel that the results of the above mentioned poll fairly represents the majority of the U.S. troops currently serving in Iraq. I also believe that some of these type polls are used to undermine the U.S. Army and our Government. They, seemingly, are jumping into the mad-media and is being used as some sort of ammunition to influence their hatefulness.

Specificly, that poll is nothing more than dog dirt.

God Bless America!!!
Would you even be on about this if the poll verified your feelings? I doubt it. My guess is that you're casting about for an excuse to deny this poll's accuracy because you don't like the outcome.

Would you even be on about this if the poll verified your feelings? I doubt it. My guess is that you're casting about for an excuse to deny this poll's accuracy because you don't like the outcome.

Hear, hear.

Hear, hear.

Hey, I didn't like it back in 2002 when Bush was at 60%+ in the polls, but I had to deal. King Bodacious needs to learn to deal as well.

Is it really necessary for the negativity when it comes to one disagreeing with your own views?
Is it possible to conduct ourselves in a rational way?
Is it possible to refrain from the childish name calling acts?
Is it possible to except the fact, that there are people in this world who disagree with your views?

Yes, I am weighing common sense vs scientific polls. I do NOT feel that they "fairly represent" the whole. I consider it to be a false generalization.

You know, the vast majority of people here did not begin to get frustrated with you until you start confusing "explaining" with "negativity", calling their posts "irrational", and laughably asking us to "except" the fact that there are people who don't agree when you seem to have difficulty ACCEPTING that others can disagree with you without it being a personal attack.

You know, the vast majority of people here did not begin to get frustrated with you until you start confusing "explaining" with "negativity", calling their posts "irrational", and laughably asking us to "except" the fact that there are people who don't agree when you seem to have difficulty ACCEPTING that others can disagree with you without it being a personal attack.

With all due respect, I think you need to reread page one to see who casted the first stone. Also, I didn't feel I used any attack tactics. I may have used a couple of defense tactics but that would be all.
I do understand the right to disagree. I was simply attacked for disagreeing.
If you were to read the first page you would see who casted the first insult, it was not I.

This is my last reply on this topic

I think it's ashame that you believe it is I who enflamed this topic.
Since, I must have offended everybody, I apologize for offended anyone.

King Bodacious, if it is possible, I would suggest taking a course or reading a book on statistics. Before last year, I had a similar view of polls as you do, and after taking a statistics course, I just how accurate many polls can be, even with a small sample size. That being said, I don't know the details of the poll conducted in your example, so I can't say if it is accurate or not.

I'm going to firmly take both sides here:

I've taken statistics, and given an unbiased diverse sample, polls with minute portions are very accurate.

HOWEVER, if you take a poll like that - often certain groups are more likely to answer than others. And many soldiers are unavailable to be polled, so most of the soldiers polled are likely to be in a fairly small localle or type of localle. So the sample is not necessarily representitive.

Same thing goes for telephone polls - calling people at home in the daytime, you're likely to come up against an inordiantly large number of women who stay home in the daytime, or people without jobs. If you're asking them what they think of Oprah, you're not likely to get a representitive sample of the population. Now if you call at dinner time, people who take the poll at dinner time are likely to have a strong opinion one way or the other, or else they'd hang up on you out of hand. A particularly good example of this is exit polling, which is notoriously unreliable. The people who choose to answer the polls tend to have a bent one way or the other, depending on what they think public opinion is... otherwise they won't admit they voted for the other guy. And then they can't figure it out when their samples don't represent the actual vote.

So polls CAN be accurate - however, the sample has to be good, and samples generally aren't. Good samples are hard to come by.

Same with polls on NS... terrible population sample. It'll tell you what generalites think, but nothing outside that.

This is MY last reply on the topic, and as neutral as I can possibly make it.

You asked me to read the whole thread, assuming that I had not. I assure you that I read every post, and that this is the timeline that I observed.

I see a poster asking for someone to explain statistics.
I see a lot of people explaining how statistics work.

I see a poster saying polls are 'bullshit' despite not knowing how statistics work.

I see a lot of people asking if the poster had taken a course in statistics.

The poster replied that he didn't, but that the polls were wrong.

I see a lot of people explaining more about how statistics work using examples.

I see the original poster denying their examples, and insisting that even though he doesn't actually know how statistics work, they are wrong.

I see people observing that the poster is ignorant of how statistics work, and since he has said from the outset that he does not know how they work and does not accept any explanations on how they work, that he is being willfully ignorant.

The original poster then characterizes this accurate observation as "sqacking", telling people not to post because they are not debating when, ironically enough, it is he who is not listening and is simply repeating his misinterpretation. The original poster also ironically tells people not to tell him he's wrong without facts (which they have given) and calls such posts "fictitious", dismissing them as not being "factual" despite having no training in compiling statistics.

The original poster again asks for education but dismisses every attempt to give it, calling such attempts 'bogus' because they apparently do not agree with a preconceived yet admittedly uneducated notion of how statistics work. The original poster continues to throw out all attempts to explain as not being "factual input" although he admits he doesn't know the facts anyway.

The poster is given a load of links that explain statistics, and ignores them completely.

The original poster has taken the term "ignorance of statistics" as a personal attack when all it means is that he has not yet been educated in this area, *as he himself stated*.

The original poster constructed a biased poll, with the implication that since it is POSSIBLE to construct a biased poll, that ALL polls are biased. This is like saying since Liz Taylor has violet eyes, proving that this color is POSSIBLE in the population, then EVERYONE has violet eyes.

The original poster again characterized people's observation that he didn't understand statistics as "name-calling". And yes, one or two people did call him an idiot about that time, but if one's argument is going to be "I don't understand it but you're wrong and I am not listening," this is not precisely an unprecedented observation to make. Polite? no. But to ask for information and deny it repeatedly based on "I feel it is not right" is not the sign of someone who is interested in learning. More in this in my closing.

The original poster then erroneously claimed that posters who were seriously discussing the thread said polls were 100% accurate, which is either a misinterpretation on his part or a gross and willing misstatement of facts. He was invited to point out which posts claimed all polls were 100%. He did not.

The original poster called an example of statistics and probability explained to him (the poker chips) dumb, despite acknowledging he did not understand statistics.

He refused to budge on the matter despite acknowledging his lack of knowledge.

The original poster equated all explanationsas "negativity", asked people to be rational when in fact he was displaying irrational behavior, seemed not to understand that calling other people's arguments "sqacking", irrational, inaccurate et cetera ad nauseum without knowledge of the subject is childish, and characterized other people's frustration with the tactic of "NOT LISTENING" as childish. Such "defense tactics" are the equivalent of a small child putting its fingers in its ears and chanting "LA LA LA" because they don't want to hear something that disagrees with their worldview. The original poster claims to have seen personal attacks where they were none, and in answering in an increasingly obtuse and hostile manner, invited but really did not get any in return.

I am going to give you the benefit of the doubt and simply close this thread as obviously you are NOT interested in discussing the matter, but I must warn you that this behavior smacks of trolling and could well gain you a warning if you choose to continue it in other threads.

<snip Kat's excellent and thorough post>Wow! That must have taken forever to type up! Good job! :)

yes I agree good job. let your fingres have a rest now, they have earned it.