If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#11
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
Mike Marotta wrote: Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. There are liars, damned liars and statisticians. I took 'stat' in college, and if there was one thing that I brought home from that class was when the professor stated "Anytime someone quotes statistics to you, they are lying. There are so many ways to make the statistics work in any foregone conclusion." Also, I am with the majority of the 93.4832% of the population who don't believe in polls. (No jokes about Poland please!) |
Ads |
#12
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"Mike Marotta" wrote in message ... On Feb 9, 3:29 pm, "Bruce Remick" wrote: Makes one wonder about the level of education reached among the people polled, and how many had to be told who Darwin was before responding to the question. Bruce, that was why I went to the HARRIS POLL website. Harris, Gallup, Pew and a couple of others are well-known for their statistical reliability. The "level of education" can be expected to be statistically representative of the USA. They would be representative for age, income, gender, race, etc. etc., all the significant variables. That's what makes Harris, Gallup, etc., worth paying for. If you read the print edition of USA Today, you will see that their polls often give the sample size, margin of error and confidence level. If I recall my stats class, you need to have 1054 samples to be 95% confident with +/- 3% error. I was pleasantly surprised to see all the hits when I googled "statistics help." Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. I still would be interested in learning if Harris only counted a response from individuals who already were familiar with Darwin and his work. I also wonder how much religion might have played a factor, and from there, demographics. Just the fact that someone would comission such a poll suggests to me that the was a built in bias and that the originator had an interest in seeing some interesting results. Always skeptical of polls and statistics. |
#13
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"Bruce Remick" wrote in message ... "Mike Marotta" wrote in message ... On Feb 9, 3:29 pm, "Bruce Remick" wrote: Makes one wonder about the level of education reached among the people polled, and how many had to be told who Darwin was before responding to the question. Bruce, that was why I went to the HARRIS POLL website. Harris, Gallup, Pew and a couple of others are well-known for their statistical reliability. The "level of education" can be expected to be statistically representative of the USA. They would be representative for age, income, gender, race, etc. etc., all the significant variables. That's what makes Harris, Gallup, etc., worth paying for. If you read the print edition of USA Today, you will see that their polls often give the sample size, margin of error and confidence level. If I recall my stats class, you need to have 1054 samples to be 95% confident with +/- 3% error. I was pleasantly surprised to see all the hits when I googled "statistics help." Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. I still would be interested in learning if Harris only counted a response from individuals who already were familiar with Darwin and his work. I also wonder how much religion might have played a factor, and from there, demographics. Just the fact that someone would comission such a poll suggests to me that the was a built in bias and that the originator had an interest in seeing some interesting results. Always skeptical of polls and statistics. ......as are 64% of others over 55. |
#14
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"Mike Marotta" wrote in message ... On Feb 9, 3:29 pm, "Bruce Remick" wrote: Makes one wonder about the level of education reached among the people polled, and how many had to be told who Darwin was before responding to the question. Bruce, that was why I went to the HARRIS POLL website. Harris, Gallup, Pew and a couple of others are well-known for their statistical reliability. The "level of education" can be expected to be statistically representative of the USA. They would be representative for age, income, gender, race, etc. etc., all the significant variables. That's what makes Harris, Gallup, etc., worth paying for. If you read the print edition of USA Today, you will see that their polls often give the sample size, margin of error and confidence level. If I recall my stats class, you need to have 1054 samples to be 95% confident with +/- 3% error. I was pleasantly surprised to see all the hits when I googled "statistics help." Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. There have been recent discussions about potential skewing of random surveys because the poller has to catch someone who is at home, has land-line telephone service, and isn't too rushed to answer a battery of questions. It's been argued that the sets of surveyed respondents in all polls are a bit heavy on seniors and other stay-at homes, while those who choose not to keep their land line and just use their mobile phone, or are out working, or cannot afford land line service are under-represented. I suspect that the net effect of each of these factors can tend to skew results toward the opinions of under-educated and/or old-time, fundamentalist thinkers. However, this effect is only important in surveys such as political polling where even a fraction of a percent can count. Regardless of whether it really is 47% or 52% or just 42% that don't accept evolution, that's enough to make me fear that the country is going to hell in a hand basket. (Okay, I don't believe in hell, I'm just trying to use terms that the ignorati understand.) mazorj "Caveat mensore" "Stats 101 taught me everything I need to know about polling pitfalls" |
#15
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"Bruce Remick" wrote in message ... "Mike Marotta" wrote in message ... On Feb 9, 3:29 pm, "Bruce Remick" wrote: Makes one wonder about the level of education reached among the people polled, and how many had to be told who Darwin was before responding to the question. Bruce, that was why I went to the HARRIS POLL website. Harris, Gallup, Pew and a couple of others are well-known for their statistical reliability. The "level of education" can be expected to be statistically representative of the USA. They would be representative for age, income, gender, race, etc. etc., all the significant variables. That's what makes Harris, Gallup, etc., worth paying for. If you read the print edition of USA Today, you will see that their polls often give the sample size, margin of error and confidence level. If I recall my stats class, you need to have 1054 samples to be 95% confident with +/- 3% error. I was pleasantly surprised to see all the hits when I googled "statistics help." Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. I still would be interested in learning if Harris only counted a response from individuals who already were familiar with Darwin and his work. Right away you run into trouble. Plenty of creationists and ID advocates think they know their Darwin. You'd have to ask a battery of questions about Darwin and evolution before you could get any idea of the accuracy of the responses of "Yes, I am familiar". For those who tested positive for knowing Darwin and evolution, can anyone doubt that the percentage of those who accept evolution would be significantly higher? I also wonder how much religion might have played a factor, and from there, demographics. Just the fact that someone would comission such a poll suggests to me that the was a built in bias and that the originator had an interest in seeing some interesting results. The biases of whoever commisioned the survey for what reasons don't matter. What matters is the survey design and in particular, whether the questions are framed in neutral terms or are "push poll" questions designed to steer responses in one direction. A reputable pollster will try to keep the questions neutral (they are the experts there) and prevent the client from putting his thumb on the scale. Even if the client is trying to keep his questions neutral, a good polling outfit will point out any problems and suggest better ways to frame the questions. Been there, done that. Always skeptical of polls and statistics. As we all should be - but only for the right reasons. That's why you have to read everything that is critical to a poll: Who commissioned it, how the sampling population was defined and respondents were selected, how they were contacted, the instructions given to the actual pollers, all the actual survey questions and branching rules thereof, how "unable to reach" and "refused to answer" cases were handled, how and why respondents and their responses may have been stratified, plus the usual statistical information like sample size, the claimed margin of error PLUS the confidence level for that level of margin of error (you hardly ever see that last one). IMO you're reading too much into the motives for the poll. In fact, sometimes polls like this are commissioned by churches and others who you might think have a bias in favor of favorable religious views, but they truly and only want accurate results. However, lacking the analysis that I just described (and am not going to do), I can't dispositively refute your concerns. |
#16
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"mazorj" wrote in message ... "Bruce Remick" wrote in message ... "Mike Marotta" wrote in message ... On Feb 9, 3:29 pm, "Bruce Remick" wrote: Makes one wonder about the level of education reached among the people polled, and how many had to be told who Darwin was before responding to the question. Bruce, that was why I went to the HARRIS POLL website. Harris, Gallup, Pew and a couple of others are well-known for their statistical reliability. The "level of education" can be expected to be statistically representative of the USA. They would be representative for age, income, gender, race, etc. etc., all the significant variables. That's what makes Harris, Gallup, etc., worth paying for. If you read the print edition of USA Today, you will see that their polls often give the sample size, margin of error and confidence level. If I recall my stats class, you need to have 1054 samples to be 95% confident with +/- 3% error. I was pleasantly surprised to see all the hits when I googled "statistics help." Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. I still would be interested in learning if Harris only counted a response from individuals who already were familiar with Darwin and his work. Right away you run into trouble. Plenty of creationists and ID advocates think they know their Darwin. You'd have to ask a battery of questions about Darwin and evolution before you could get any idea of the accuracy of the responses of "Yes, I am familiar". For those who tested positive for knowing Darwin and evolution, can anyone doubt that the percentage of those who accept evolution would be significantly higher? I also wonder how much religion might have played a factor, and from there, demographics. Just the fact that someone would comission such a poll suggests to me that the was a built in bias and that the originator had an interest in seeing some interesting results. The biases of whoever commisioned the survey for what reasons don't matter. What matters is the survey design and in particular, whether the questions are framed in neutral terms or are "push poll" questions designed to steer responses in one direction. A reputable pollster will try to keep the questions neutral (they are the experts there) and prevent the client from putting his thumb on the scale. Even if the client is trying to keep his questions neutral, a good polling outfit will point out any problems and suggest better ways to frame the questions. Been there, done that. Always skeptical of polls and statistics. As we all should be - but only for the right reasons. That's why you have to read everything that is critical to a poll: Who commissioned it, how the sampling population was defined and respondents were selected, how they were contacted, the instructions given to the actual pollers, all the actual survey questions and branching rules thereof, how "unable to reach" and "refused to answer" cases were handled, how and why respondents and their responses may have been stratified, plus the usual statistical information like sample size, the claimed margin of error PLUS the confidence level for that level of margin of error (you hardly ever see that last one). IMO you're reading too much into the motives for the poll. In fact, sometimes polls like this are commissioned by churches and others who you might think have a bias in favor of favorable religious views, but they truly and only want accurate results. However, lacking the analysis that I just described (and am not going to do), I can't dispositively refute your concerns. I guess they're not really "concerns". More like skeptical peeves which seldom affect me personally. Like you pointed out, the only way to truly judge the validity of any poll results is to have access to the actual questions asked, a recording of the actual session, personal background details of each pollee(?) along with what part of the country they lived in. Totally impractical and unlikely. It just peeves me when any poll results are announced and then are treated as foundations on which to build additional projections. But I still get much more upset to find that I'm out of beer. .. |
#17
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"mazorj" wrote in message ... "Bruce Remick" wrote in message ... "Mike Marotta" wrote in message ... On Feb 9, 3:29 pm, "Bruce Remick" wrote: Makes one wonder about the level of education reached among the people polled, and how many had to be told who Darwin was before responding to the question. Bruce, that was why I went to the HARRIS POLL website. Harris, Gallup, Pew and a couple of others are well-known for their statistical reliability. The "level of education" can be expected to be statistically representative of the USA. They would be representative for age, income, gender, race, etc. etc., all the significant variables. That's what makes Harris, Gallup, etc., worth paying for. If you read the print edition of USA Today, you will see that their polls often give the sample size, margin of error and confidence level. If I recall my stats class, you need to have 1054 samples to be 95% confident with +/- 3% error. I was pleasantly surprised to see all the hits when I googled "statistics help." Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. I still would be interested in learning if Harris only counted a response from individuals who already were familiar with Darwin and his work. Right away you run into trouble. Plenty of creationists and ID advocates think they know their Darwin. You'd have to ask a battery of questions about Darwin and evolution before you could get any idea of the accuracy of the responses of "Yes, I am familiar". For those who tested positive for knowing Darwin and evolution, can anyone doubt that the percentage of those who accept evolution would be significantly higher? So from what you say, it would seem near impossible to corral enough educated open-minded, unbiased individuals to ever conduct a meaningful poll. If you know Darwin you're out. If you never heard of him you're out. If religion plays a significant part in your life you're out. If you're an atheist you're out. If you own more than one Darwin medal you're out........... |
#18
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"Jud" wrote in message ... Mike Marotta wrote: Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. There are liars, damned liars and statisticians. I took 'stat' in college, and if there was one thing that I brought home from that class was when the professor stated "Anytime someone quotes statistics to you, they are lying. There are so many ways to make the statistics work in any foregone conclusion." Also, I am with the majority of the 93.4832% of the population who don't believe in polls. (No jokes about Poland please!) If your head is frozen inside a block of ice and your feet are on fire then statiscally, on average, you are perfectly fine. Billy |
#19
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"note.boy" wrote in message ... "Jud" wrote in message ... Mike Marotta wrote: Depending on where you live, you can probably take an accredited statistics class at a community college for $350-$500. It will take 13-15 weeks of commitment, but you will know way more than everyone else about how polling is done and the mental exercise will help stave off senility. There are liars, damned liars and statisticians. I took 'stat' in college, and if there was one thing that I brought home from that class was when the professor stated "Anytime someone quotes statistics to you, they are lying. There are so many ways to make the statistics work in any foregone conclusion." Also, I am with the majority of the 93.4832% of the population who don't believe in polls. (No jokes about Poland please!) If your head is frozen inside a block of ice and your feet are on fire then statiscally, on average, you are perfectly fine. Billy You must be an engineer of some type. :-D We'd kid ours with the a variant of that: "An engineer is someone who thinks that if your left foot is in a bucket of boiling water and your right foot is in a bucket of ice water, on average you're comfortable." The funny thing was that 81.717% of them agreed! |
#20
|
|||
|
|||
Darwin, Britain's Hero, Is Still Controversial In U.S.
"Bruce Remick" wrote in message ... "mazorj" wrote in message ... "Bruce Remick" wrote in message ... .... I still would be interested in learning if Harris only counted a response from individuals who already were familiar with Darwin and his work. Right away you run into trouble. Plenty of creationists and ID advocates think they know their Darwin. You'd have to ask a battery of questions about Darwin and evolution before you could get any idea of the accuracy of the responses of "Yes, I am familiar". For those who tested positive for knowing Darwin and evolution, can anyone doubt that the percentage of those who accept evolution would be significantly higher? So from what you say, it would seem near impossible to corral enough educated open-minded, unbiased individuals to ever conduct a meaningful poll. If you know Darwin you're out. If you never heard of him you're out. If religion plays a significant part in your life you're out. If you're an atheist you're out. If you own more than one Darwin medal you're out........... Maybe I misunderstood what you were driving at with "I still would be interested in learning if Harris only counted a response from individuals who already were familiar with Darwin and his work." My point was that in order to do that, you can't just ask the question "Are you familiar with Darwin and his work?" to screen and limit the participants to those who actually know enough about Darwin and evolution. Most anti-Darwinians fancy themselves as knowledgeable, as in "Know thine enemy" even though in the vast majority of cases, whatever they "know" comes from sermons and diatribes reviling him (and maybe a feature episode or two on the History Channel). And since most people don't like to admit ignorance even in an anonymous poll, you'll have another block of respondents who don't know jack about Charles but will say they do. What trips me up here is that you shifted your stated criteria from "only including those familiar with Darwin and his work," to excluding just about everybody because in your view they cannot be "educated open-minded, unbiased individuals" if they know Darwin, if they don't know Darwin, if they never heard of him, if they are religious or if they are atheist. Sure, that excludes just about everybody, but that's not what you asked in your original question. I took the original question to mean "Did they only poll people with enough knowledge to intelligently answer the questions" or did they interview "any warm body that answered the telephone"? In most surveys you want true random selection (no, not evolutionary random selection, just statistically random selection) because you want truly representative slices of all members of the overall population. So if you wanted to limit respondents only to people who have an accurate and adequate (even if only a layman's) grasp of Darwin and evolution, first they'd have to pass a moderately tough quiz on the topic. I don't know why you'd want to do it that way. The results would only confirm that "the vast majority of people who really know evolution think it's a valid scientific model." As I said, that's already a no-brainer. Most surveys want to know what the entire population thinks, not just one limited slice of it. And in reporting the results of the poll, you would have to state your findings with the limiting condition "Among people who have an accurate and adequate understanding of Darwin and evolution..." To keep this on topic, I would only rule you out if you had two or more Darwin medals. One is just a random selection by you in the evolution of your holdings. Two or more indicates some intended design (intelligent or not) in your acquisition of things Darwinian. ;-) |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Darwin Award | Tony Vella | General Discussion | 12 | February 17th 07 12:26 AM |
Panal approves controversial coin | stonej | Coins | 1 | February 2nd 06 07:06 PM |
New Austrailian commemorative coin is controversial | JSTONE9352 | Coins | 18 | December 13th 04 08:25 AM |