Can We Believe In Polls About God beliefs?
31 March 2007 by StardustIt doesn’t surprise me that there is yet another poll about god beliefs right around the time of another Christian mythology season. On the front page of MSN is a misleading headline that reads “Poll finds 90% believe in God” and the link is to a Newsweek article that states, “The latest NEWSWEEK poll shows that 91 percent of American adults surveyed believe in God—and nearly half reject the theory of evolution.”
What gets me about this particular poll is this little bit:
Although one in ten (10 percent) of Americans identify themselves as having “no religion,” only six percent said they don’t believe in a God at all. Just 3 percent of the public self-identifies as atheist, suggesting that the term may carry some stigma. Still, the poll suggests that the public’s tolerance of this small minority has increased in recent years. Nearly half (47 percent) of the respondents felt the country is more accepting of atheists today that it used to be and slightly more (49 percent) reported personally knowing an atheist.
Now, I admit I am not a math wizard, but if 49% of god-botherers claim to personally know an atheist, how does the percentage of atheists come out to only 3%? (Or there are a lot more atheists than these polls are admitting to.)
The last paragraph at the article where few readers will ever see states:“The NEWSWEEK Poll, conducted March 28-March 29, has a margin of error of plus or minus 4 percentage points for questions based on all registered voters and plus or minus 6 percentage points for results based on registered Republicans and Republican leaners. In conducting the poll, Princeton Survey Research Associates International interviewed 1,004 adults aged 18 and older.”
31 March 2007, on 7:48 pm
“Now, I admit I am not a math wizard, but if 49% of god-botherers claim to personally know an atheist, how does the percentage of atheists come out to only 3%?”
It’s because they say “But I know some atheists” in the same way that bigots say, “But I have several black friends.”
31 March 2007, on 8:52 pm
There seems to be a great difference between those who identify themselves with a particular religion and those who actually go to church. At least here in Australia, that seems to be the case.
“The 2001 census identified that 68% of Australians call themselves Christian: 27% identifying themselves as Roman Catholic and 21% as Anglican. Australians who identify themselves as followers of non-Christian religions number 5%. A total of 16% were categorised as having “No Religion” (which includes non-theistic beliefs such as humanism, atheism, agnosticism and rationalism) and a further 12% declined to answer or did not give a response adequate for interpretation. As in many Western countries, the level of active participation in church worship is much lower than this; weekly attendance at church services is about 1.5 million, about 7.5% of the population.”
http://en.wikipedia.org/wiki/Australia
As this statistics show, only about 7.5% of the australian population got to church on a weekly basis. We must be a mob of heathens.
31 March 2007, on 9:00 pm
It appears that 49% of that 3% (of us) knows everyone under the sun.
Which means us US atheists are far too busy being social butterflies, & are rushing about outing ourselves to all & sundry. Surprising how we get anything done, what w/all the social networking we’re doing, no?
Of course, if they’d asked the polling group if they blogged, that’d explain a great deal.
31 March 2007, on 9:27 pm
Now, I admit I am not a math wizard, but if 49% of god-botherers claim to personally know an atheist, how does the percentage of atheists come out to only 3%?
I guess a lot of god lovers happen to know many of the same atheists. (ie. an entire congregation of a church all know the same atheist.)
BTW: maybe the poll was taken at a church.
31 March 2007, on 10:16 pm
Here is a Pew Research poll.
http://people-press.org/reports/pdf/287.pdf
I have been trying to find a poll I read recently that stated the breakdown as to religious belief among the different Generations. I know it had ~20% of GEN Y as being Agnostic/Atheist and they also defined “spiritual” differently. The Newsweek poll above seems awful suspicious to me. I don’t trust it to be truly reflective of the facts. Pew Research usually has some very well researched and in depth papers on the polls they take.
1 April 2007, on 12:19 am
Is 1004 interviewees really a large enough sample to guage the entire population?
1 April 2007, on 12:26 am
Perhaps these religious kooks are right after all. If there are that many monumentally stupid people who don’t believe in science then the human race is NOT evolving but rather DEvolving, returning to a stupider, more primitive state.
I read the last report by the National Institute of Literacy that stated 90 million adults are functionally illiterate, meaning they haven’t the intellectual capacity to read or understand even simple phrases such as in a newspaper even if they can tell you the words. With more than 50% of adults being that dim I might wager these polls could be right.
What an absolute horror, as the American Empire falls its populace turns into masses of easily manipulated morons. No wonder the criminal profiteers have taken over so easily.
1 April 2007, on 1:45 am
Eh, you know how phrasing is everything with these polls. Like when one pollmaker got the majority of people who answered to say that they didn’t believe that the Holocaust had happened. It’s all in the wording.
1 April 2007, on 1:52 am
That poll is extremely suspect.
Question #12: which comes closest to your view about the origin of humans? Here is the Agnostics/Atheists breakdown:
God had no part: 45%
God guided the process: 27%
God created man in present form: 13%
Other/Don’t know: 15%
I would have expected the numbers to be split between “God had no part” and “Other/Don’t know” relating to the ratio of Agnostics/Atheists.
But these figures seem to imply that 40% of agnostics and atheists believe in God! The researchers should have thrown their data out right then!
1 April 2007, on 2:12 am
Follow the money! Who paid for the poll? Newsweek AND WHO ELSE?
When you have the answer to that question, you will probably see why the results seem skewed.
See my questions to Zogby (Polling) International:
1 April 2007, on 4:22 am
Personally, I’ve had a profound change of heart over the last few days. My view has now (i.e. since the Vatican has opened up to the possibility of allowing The Russian Orthodox Church to take possession, in an exchange with the the government, of a local Cathedral in a city very close to my current location and transform into a kind of Russian Orthodox A$$i$i) become much more tolerant of religion and other statue-worhipping type folks. The potential opportunities for per$sonal gain are endless !!
1 April 2007, on 5:54 am
Ahem, it’$ endle$$!
1 April 2007, on 9:02 am
Have you seen the movie “Idiocracy”? Polls like this one make me think we may be heading in that direction.
1 April 2007, on 10:12 am
As the mathematician in residence at GifS, I think I have been summoned…
Christopher said:
If the methodology of the poll was correct (random sample, non-loaded questions, ..), then yes, 1004 are enough. This is true in particular if the methodology wer explained (so that we can understand what the results actually mean), and if we were told both the error margin and the confidence interval, (which we are not!) Good Math, Bad Math had a very nice basic post on it. Statistics is a very powerful branch of math. Too bad it is often misused and prostituted.
Stardust said:
Oh, but this makes perfect sense. Say that there are 3% of atheist, and that they are distributed more or less homogeneously among different geographical areas, jobs, and social status (this is probably very non-true, but bear with me). So, if you meet personally a random person, the probability that he is not an atheist is .97. Most people that a believer knows personally will not be chosen randomly,
as they may be relatives or friends from church. But say that a believer knows personally 20 random people (the postman, the babysitter, their kid’s teacher,…), meaning people for whom there is not a reason to assume they have the same beliefs. The probability that none of them is an atheist would be
(0.97)^20 = 0.54
If you make that 25 instead of 20
(0.97)^25 = 0.47
So yes, it makes sense that half of the population knows one of us, even if we are only 3%. Isn’t math great?
1 April 2007, on 11:54 am
Zipi - that’s why we pay you the big bucks
So, if the following findings are accurate
The human race along with rationality and reason are fucked.
Sweden is looking better and better to me.
1 April 2007, on 12:26 pm
A question relating to the existance of “god” is biased in and of itself. The pollers may be surprised with the percentage of ghost believers if they askied the masses such a question. I agree that the USA could have a much more intelligent populace, but this poll is misleading and was probably taken with alterior motives (missionary).
To ask someone questions regarding evolution and the bible is tainted when the person probably knows virtually nothing about either.
Perhaps a more accurate poll would have been, “What do you think about philosophy?”, or “Do you make life decisions based on rationale?”, or “Do you understand the current prevailing theories on the origin of species, and if so do you believe them?”
Many people think that evolution means that apes suddenly turned into humans. In my experience, an openminded person who is explained the concepts of speciation over very long time periods has no choice but to “believe” in such a well supported theory.
The fact is that we are all atheists - evryone on the planet. Those who say otherwise are lying to try to fit in. we can’t help but base our decisions based on rationale, as this is gentically programmed into us. Try to convince your peers of this - its worth a shot.
1 April 2007, on 1:17 pm
jimmy dean - good points. polls only allow for yes or now answers and are often misleading.
I cross-posted this at my personal blog and tommykey brought up other good points. tommy wrote:
1 April 2007, on 3:25 pm
David W - A lot of people don’t understand the term agnostic. I know people and have talked to people who think they are agnostic yet believe in god or a “higher power”. I think they generally mean they are not Xian or don’t subscribe to any particular religion.
So yes, they probably have a problem with the wording if people don’t understand what they are identifying themselves as.
2 April 2007, on 12:05 pm
The main problem with polls like this is that they cannot and never will be close to accurate. You quite simply can’t take a random sample of the American population and say that is a direct ratio.
I’d have to give this the same answer I gave a lot of my word problems in math class…NOT ENOUGH VARIABLES.
There is no TRUE way to get a completely accurate census on personal beliefs in America unless you survey every single person in America and guarantee they they completely understand the words you used and are completely honest.
So basically the the 3% thing is bull.
2 April 2007, on 1:59 pm
The “honesty” part is why I don’t put much stock in most polls. I doubt many people lie outright, but I imagine even fewer people tell the whole truth accurately.
If a poll is taken live and the surveyor asks questions orally, then the subject could be inclined towards fabricating answers in order to please his questioner.
Written polls with multiple choice questions are not much better. When given 4 or 5 choices, what does the subject do when none of them really express his opinion? Pick the closest answer? Pick a random one?
2 April 2007, on 2:30 pm
The “honesty” part is why I don’t put much stock in most polls. I doubt many people lie outright, but I imagine even fewer people tell the whole truth accurately.
I kept my atheism under wraps for quite awhile before admitting it openly. People would ask me what church I went to and I would merely say, “oh, I don’t go to any church at the moment, I really am disillusioned with organized religion.” The longer I was away from religion, the more I studied other religions, mythologies and other cultures, the more open I became about not having any god beliefs.
Several of my good friends will say they are xian on a survey, or when they go into the hospital and are asked what religion they are, but they don’t even think about religion at all otherwise and they really don’t seem to give a hoot.
2 April 2007, on 9:25 pm
Ian said:
I challenge that. If you take a random sample, the questions are not biased, and you get people to answer honestly (and there are even ways to try to account for potential liars!), then you can trust the results. Nobody claims that it is a direct ratio. Instead you give your results with a margin of error and a confidence level. That has a precise statistical meaning and it is very accurate.
To give you an example, say that a poll states that 40% of the population has property X, with a margin of error of plus or minus 2 percent points and a confidence level of 95%. What this means is that you have a 95% certainly that the real percentage of the population having property X is between 38% and 42%. Nobody claimed that it was a direct ratio.
Better courses in statistics should be compulsory in high school.
Of course, this does not convince any fundie, because they mistrust everything that has to do with science, but it should be convincing for a rational person.
3 April 2007, on 12:38 am
And remember, 5 out of every 4 dentists prefer Crest toothpaste!
3 April 2007, on 11:39 am
I have always wondered about polls in the US. Not only are we geographically a large country, but we are really multiple mini-countries. The culture, beliefs, and education levels are relatively homogenous within each region (East Coast, West Coast, South, South West, North West, Midwest, etc.), due, in part, to history and the settlement of large groups of immigrants of different ethnic origins in different parts of the county. The differences between each region in geography and culture make me think that 1000 from each section would be more accurate. A poll taken in New England cannot reflect the prevailing views and beliefs of people living in the South. Each of those areas represents millions of people, and their results would not be interchangeable. I doubt they would even fall into the margin of error. A poll taken in Seattle could not reflect the views of people in Orlando. There are so many polar opposites in this country that the only way to get useful and valid statistics (if the math is done correctly and honestly as Zipi demonstrated), would be for the sample to be evenly distributed throughout the US and not to have been taken one afternoon on a university campus, for instance. I may be wrong, of course, as my mastery of math in a formal setting never extended beyond a D in Algebra II (I did excel at Geometry - 110%, ha ha - but that doesn’t seem to apply here…)
3 April 2007, on 5:03 pm
I have a theory that I’ve espoused here before: The mountains have created this problem, and keep it going.
When the earliest settlers moved west, the first mountain ranges they encountered were the Smokies, Blue Ridges and all the way north to the Poconos and Adirondacks.
The loners tended to find the back-of-beyond hollows and settle there. Not entirely loners, they began to have families. But they were the usual poverty-stricken families–too many kids (wonderful for farming but a real resource drain) and no education.
If one or both parents could read and write, it’s possible it was taught to the children. If they could only read, they may have just read to the children.
But if you’re poor and can only afford one book, it was definitely the bible. No Shakespeare there! And it was usually the repository of family records, as well.
I imagine most families lived too far from the nearest small town to go into very often. Thus the children were, in effect, prisoners of the educational level of the parents–usually the stern father.
From the 1600s until progress brought schools and education, this culture flourished out of sight.
This probably applies to the Ozarks, as well.
And America still has its pockets of mountain poverty!
What doesn’t fit into this model is the Mississippi Delta bible-belt insanity…
3 April 2007, on 8:11 pm
Bean said:
Very much so, and this is one of the common mistakes in polls. Geting a random sample is very difficult! For instance, polls are often done calling random telephone numbers. This is not an ideal method: homeless people or Amish are less likely to have a phone, unemployed or retired people are more likely to answer it, …
4 April 2007, on 12:40 am
Crest toothpaste also makes great lube
6 April 2007, on 3:03 pm
A new report just published in the UK shows the discrepancies between the figures quoted in the UK 2001 Census, the Government’s annual ‘British Social Attitudes Survey’ and the latest academic report entitled ‘Churchgoing in the UK’. Further details can be found here.