Survey: Churchgoers more likely to back torture

  • #1
djeitnstine
Gold Member
614
0

Main Question or Discussion Point

Interesting survey done

http://www.cnn.com/2009/US/04/30/religion.torture/index.html

I find it highly ironic. Although the sample numbers are not large its still eye opening. None church goers still do have morals.
 

Answers and Replies

  • #2
1,482
3
No kidding

If a man beats his male or female slave with a rod and the slave dies as a direct result, he must be punished, but he is not to be punished if the slave gets up after a day or two, since the slave is his property. Exodus 21:20-21
If thy hand or thy foot offend thee, cut them off, and cast them from thee: it is better for thee to enter into life halt or maimed, rather than having two hands or two feet to be cast into everlasting fire. Matthew 18:8-9
The same shall drink of the wine of the wrath of God, which is poured out without mixture into the cup of his indignation; and he shall be tormented with fire and brimstone in the presence of the holy angels, and in the presence of the Lamb. And the smoke of their torment ascendeth up for ever and ever. Revelation 14:10-11
 
  • #3
378
2
Interesting survey done

http://www.cnn.com/2009/US/04/30/religion.torture/index.html

I find it highly ironic. Although the sample numbers are not large its still eye opening. None church goers still do have morals.
I am not surprised and my reasoning beyond that statistics is that:
1) Majority Religious people might be conservatives
2) Religious people have evil definition and they might be supporting "Us against them"
3) Majority liberals might be non religious
4) God tells them that it's okay to kill your enemies (witches/etc)
It can be either all of them or none of them.

Religions are the reflection of ancient societies.. I am not against them but I also don't have blind faith in them. They will hardly ever meet with the present needs.


But, one thing to point out:

The analysis is based on a Pew Research Center survey of 742 American adults conducted April 14-21. It did not include analysis of groups other than white evangelicals, white non-Hispanic Catholics, white mainline Protestants and the religiously unaffiliated, because the sample size was too small.
 
Last edited:
  • #4
4,239
1
Party affiliation.

“The nationalist not only does not disapprove of atrocities committed by his own side, but he has a remarkable capacity for not even hearing about them.” -George Orwell
 
  • #5
Vanadium 50
Staff Emeritus
Science Advisor
Education Advisor
2019 Award
23,832
6,273
I'm amused that people are drawing conclusions from this - particularly conclusions that I suspect support what they "knew all along". RootX touches on the key point: 742 people were tested. If 8 of them change their answer from "sometimes" to "rarely" (or the other way), the purported effect goes away.

Everything depends on the answers of those 8 people.
 
  • #6
6,265
1,275
The question was about torture of suspected terrorists:

"The survey asked: "Do you think the use of torture against suspected terrorists in order to gain important information can often be justified, sometimes be justified, rarely be justified, or never be justified?"

This complicates matters greatly. What kind of "important information" do people imagine the question pertains to? For Americans right now it probably calls to mind an impending attack on the scale of 911: "Is it OK to torture a few people if that means a few thousand would be spared getting rammed by a hijacked airliner? What about sparing the grief of the many more people represented by their surviving families?" 54% of churchgoers called up a mental scenario such that they agreed it could be justified and 42% of non-churchgoers called up a scenario such that they agreed to it.

The main trouble is that it's a fairly vague hypothetical question that will never actually neatly fit any real world situation. To the extent anyone gives an answer it is an answer to the rohrschach scenario they happen to imagine in response to the question. It could be, I'm not saying it is for sure, that churchgoers, being family oriented, do a lot more brooding over the thought of wives whose husbands were suddenly blown away, and children whose fathers suddenly stopped coming home.

So, imagine yourself in an emergent scenario: you're some churchgoing small town sheriff who happened to stop and arrest a fleeing terrorist. You called the FBI and they're on their way, but in the meantime you know there's a hijacked jet in the sky planning on crashing into some densely populated area. You have the hijacker's accomplice in the back room, but he's just grinning and spitting at you.

"Time's a wastin', Clarice. Tick toc tick tock."
 
  • #7
Hurkyl
Staff Emeritus
Science Advisor
Gold Member
14,916
17
Interesting survey done

http://www.cnn.com/2009/US/04/30/religion.torture/index.html

I find it highly ironic. Although the sample numbers are not large its still eye opening. None church goers still do have morals.
To rephrase -- you know that the numbers are inadequate to draw any conclusions, yet you still wish to draw conclusions. :tongue: It sounds like you're just looking to bash religion....
 
  • #8
Astronuc
Staff Emeritus
Science Advisor
18,704
1,718
The analysis is based on a Pew Research Center survey of 742 American adults conducted April 14-21. It did not include analysis of groups other than white evangelicals, white non-Hispanic Catholics, white mainline Protestants and the religiously unaffiliated, because the sample size was too small.
Seems like a very small sample.

From the United Methodist Church of which my parents are members:
Our commitment to human rights is grounded in the conviction that each and every human life is sacred. Therefore the United Methodist Church endorses legislative and judicial remedies for the use of torture and illegal detention ...such as the appointment of special counsels [and] appropriate investigations.

#6120 "Opposition to Torture" Book of Resolutions (2008)
http://www.umc-gbcs.org/site/c.frLJK2PKLqF/b.5128831/k.521B/United_Methodists_Do_Not_Torture/apps/ka/ct/contactus.asp?c=frLJK2PKLqF&b=5128831&en=cuJTK5NLJmISL1PTIhJMK9OSKnI0I5MMJfIUIbPXKuJbF [Broken]

National Religious Campaign Against Torture
http://www.nrcat.org/index.php

http://www.pcusa.org/peacemaking/
http://www.pcusa.org/peacemaking/actnow/actnow.htm#stoptorture [Broken]

I'm quite sure the majority of Unitarian Universalists oppose torture as well as war.
 
Last edited by a moderator:
  • #9
LowlyPion
Homework Helper
3,090
4
It seems to me to be a weak correlation. The sample size is large enough I suppose to come to a meaningful conclusion, but I don't think the skew in responses between the extremes of non-church going to frequent attendee is sufficient to characterize it as anything but a potential tendency, certainly not a strong enough conclusion to warrant the article.

And this takes no account I think of the bias carried in the question itself.

Maybe ask people if they think that enhanced interrogation performed on untried detainees denied legal rights is fair and reasonable?
 
  • #10
DaveC426913
Gold Member
18,596
2,053
I'm amused that people are drawing conclusions from this - particularly conclusions that I suspect support what they "knew all along". RootX touches on the key point: 742 people were tested. If 8 of them change their answer from "sometimes" to "rarely" (or the other way), the purported effect goes away.

Everything depends on the answers of those 8 people.
Perhaps, but one can do that with any set of statistics. The question is: is the study rigorous enough to be valid?

If the answer is yes then the results (even if it's just 8) is statistically significant.
 
  • #11
Gokul43201
Staff Emeritus
Science Advisor
Gold Member
7,051
17
I second phrak's opinion on this. I think this is more about party affiliation, and the tendency to condone or denounce a position taken by someone that you favor or reject. The correlation is stronger when directly tabulated against party affiliation1.

1. http://people-press.org/report/510/public-remains-divided-over-use-of-torture

I'm amused that people are drawing conclusions from this - particularly conclusions that I suspect support what they "knew all along". RootX touches on the key point: 742 people were tested. If 8 of them change their answer from "sometimes" to "rarely" (or the other way), the purported effect goes away.

Everything depends on the answers of those 8 people.
So what? What does it matter that "everything depends on ... 8 people"? It's completely irrelevant, without analyzing the statistics, and moreover, unless I'm making a mistake below, it also appears to be wrong.

Of the 336 religious respondents, 54% chose 'often' or 'sometimes'. While of the 186 hardly religious respondents, 42% chose 'often' or 'sometimes'. For the "purported effect" to go away (just between these two groups; there's also a third in-between group), 12% of either group (or some combination making up the 12% difference) - that's 40 religious people or 23 of the other group - will have to switch sides. Where did you get the number '8' from?

On samples sizes2 of 336 and 186 respondents, the error bars at the 95% confidence level are +/- 6% and +/-8%. For the "purported effect" to go away, you are asking for people to switch sides, which we know will not happen with a confidence level in the high 90% range.

2. http://people-press.org/report/?pageid=1506 [Broken]

So, imagine yourself in an emergent scenario: you're some churchgoing small town sheriff who happened to stop and arrest a fleeing terrorist ...
Are you (the sheriff) churchgoing, or is the town? And you (the sheriff) can have only arrested a suspected terrorist. And then, you should multiply the probability of your responses to this emergent scenario with the frequency with which you expect to encounter such a scenario (and accordingly for any other scenarios you can think of) to determine if your final answer belongs in the 'sometimes' (or 'often') category or the 'rarely' (or 'never') category.

To rephrase -- you know that the numbers are inadequate to draw any conclusions, yet you still wish to draw conclusions.
True ... if by 'rephrase', you mean 'rewrite without regard to the meaning conveyed'. djeitnstine only said that the sample size is "not large", not that it was inadequate to draw conclusions. In this case, you can not draw the conclusion stated in the thread title at the 3sigma confidence level, but you can draw it at the 2sigma (around 95%) level.
 
Last edited by a moderator:
  • #12
Vanadium 50
Staff Emeritus
Science Advisor
Education Advisor
2019 Award
23,832
6,273
So what? What does it matter that "everything depends on ... 8 people"? It's completely irrelevant, without analyzing the statistics, and moreover, unless I'm making a mistake below, it also appears to be wrong.
8 people only have to "switch sides" from "sometimes" to "rarely".

But if you like, take the raw data from the Pew Forum. Looking at the graphics for the case where they compare the torture opinion to the churchgoing rate allows one to make a 3x4 table of number of respondents. One can then calculate the Chi-square of the experimental table under the assumption that there is no correlation between one's opinion on torture and one's rate of churchgoing. I get a chi-squared of 8.5 for 6 degrees of freedom.

That's not enough to reject it at the 90% confidence level, much less 95% or 99%.

The 8 is saying the same thing, just another way: ~700 people (some declined to answer) and 12 bins means you have ~60 people per bin. A change of 8 in a counting experiment of 60 is only a one standard deviation effect.
 
  • #13
drankin
It may also depend on what each person considers to be torture. I think it's relevant that conservatives, who tend to have a religious faith, would be more inclined to beat the snot out of someone if they believed they had information about an impending act of terrorism. I know I wouldn't have an issue with it.
 
  • #14
6,265
1,275
And you (the sheriff) can have only arrested a suspected terrorist.
You're right. My scenario is nonsense in any case where it's stipulated we only have a suspect.
 
  • #15
mgb_phys
Science Advisor
Homework Helper
7,774
12
It also depends on how the question was phrased - and what conclusion the questioner wants to draw from it:

Q: Do you believe the US should be allowed to kill terrorist suspects.
Turns into either "Majority calls for US military presence in Afghanistan" or "Majority support concentration camps"

Or from the "Yes minister" guide to polls:

Q: Are you worried about the rise in crime among teenagers?
A: Yes.
Q: Do you think there is lack of discipline and vigorous training in our Schools?
A: Yes.
Q: Do you think young people welcome some structure and leadership in their lives?
A: Yes.
Q: Do they respond to a challenge?
A: Yes.
Q: Might you be in favour of reintroducing National Service?
A: Yes.

Q: Mr. Woolley, are you worried about the danger of war?
A: Yes.
Q: Are you unhappy about the growth of armaments?
A: Yes.
Q: Do you think there's a danger in giving young people guns and teaching them how to kill?
A: Yes.
Q: Do you think it's wrong to force people to take arms against their will?
A: Yes.
Q: Would you oppose the reintroduction of conscription?
A: Yes.

Sir Humphrey Appleby: There you are, Bernard. The perfectly balanced sample.
 
  • #16
490
2
Perhaps, but one can do that with any set of statistics. The question is: is the study rigorous enough to be valid?

If the answer is yes then the results (even if it's just 8) is statistically significant.
You seem to be implying that being statistically significant would mean that it would be reasonable to assume the headline is accurate, which indicates a misunderstanding of the meaning of "statistical significance," which is simply an arbitrary thresholding on the difference between two sample means, and merely indicates whether or not those means come from the same distribution or not. In other words, it has nothing to do with the validity of making a generalization based on the sampling.

A statistical inference is only valid to the extent that the sample is representative of the entire population to which the result is generalized. If the sampling is not representative of the population, then being statistically significant, or insignificant, is irrelevant.

Vanadium's point was that the opinion of 8 people were enough to change the results in the sample size, which makes it intuitively obvious that the sample size was not large enough to be representative of the American (or world) population, which involves hundreds of millions or billions of people.
 
  • #17
Gokul43201
Staff Emeritus
Science Advisor
Gold Member
7,051
17
8 people only have to "switch sides" from "sometimes" to "rarely".
From what group?

But if you like, take the raw data from the Pew Forum. Looking at the graphics for the case where they compare the torture opinion to the churchgoing rate allows one to make a 3x4 table of number of respondents. One can then calculate the Chi-square of the experimental table under the assumption that there is no correlation between one's opinion on torture and one's rate of churchgoing. I get a chi-squared of 8.5 for 6 degrees of freedom.
That's a confidence level of about 80%. So you can reject (at the 90% level, or even the 81% level) the hypothesis that religious regularity is correlated to the tendency to choose between one of the 4 choices for allowability of torture. But let's say you're not looking to be able to determine the finer shades and nuances in opinions and want to look only at the coarser tendency.

So, if instead, you club 'often' and 'sometimes' into one group (those that believe torture is justified at least sometimes), and put 'rarely' and 'never' into another group (those that believe torture is justified no more often than rarely), then you get a table with 6 cells and a chi-squared value of 4.99 with 2 DoF. With this test, the null hypothesis that the choice between allowing torture more often than "sometimes" and less often than "rarely" is independent of religious regularity can be rejected at a 92% confidence level.

If further, you throw out the middle group of not-so-regular churchgoers, and rewrite your hypothesis to test only the difference in tendencies of the two extremes of the churchgoers (the very regulars and the nearly nevers), you can reject the independence at the 97% confidence level.
 
  • #18
Gokul43201
Staff Emeritus
Science Advisor
Gold Member
7,051
17
Vanadium's point was that the opinion of 8 people were enough to change the results in the sample size, which makes it intuitively obvious that the sample size was not large enough to be representative of the American (or world) population, which involves hundreds of millions or billions of people.
I am not possessed of such a strong intuition that this is obvious to me. Can you provide a rigorous proof for this intuitively obvious truth?
 
  • #19
Hurkyl
Staff Emeritus
Science Advisor
Gold Member
14,916
17
True ... if by 'rephrase', you mean 'rewrite without regard to the meaning conveyed'....
I think you're putting words in his mouth.
 
  • #20
490
2
I am not possessed of such a strong intuition that this is obvious to me. Can you provide a rigorous proof for this intuitively obvious truth?
With inferential statistics, you are trying to reach conclusions that extend beyond the immediate data alone. For instance, we use inferential statistics to try to infer from the sample data what the population might think. ...
http://www.socialresearchmethods.net/kb/statinf.php

In choosing a sample, it is therefore crucial
that it be representative. It must not overrepresent one kind
of citizen at the expense of others. For example, something
would be wrong with our sample if it happened to be made up
entirely of Florida residents. (Recall the controversy
surrounding presidential voting in Florida in 2000.) If the
sample held only Floridians, it could not be used to infer the
attitudes of other Americans.
Random samples, especially
if the sample size is small, are not necessarily
representative of the entire population
. For example, if a
random sample of 20 subjects were taken from a population with
an equal number of males and females, there would be a
nontrivial probability (0.06) that 70% or more of the sample
would be female. (To see how we obtain this probability,
click here.) Such a sample
would not be representative, although it would be drawn
randomly. Only a large sample size makes it likely that our
sample is representative of the population. For this reason,
inferential statistics needs to take into account the sample
size when it attempts to generalize results from samples to
populations
.
http://cnx.org/content/m10185/latest/

Be wary of statistics that have small sample sizes, unless they are in a peer-reviewed journal. Professional statisticians can interpret results correctly from small sample sizes, and often do, but not everyone is a professional, and novice statisticians often incorrectly interpret results.
http://owl.english.purdue.edu/owl/resource/672/05/

See pages 13 and onward for ways to determine the minimal sample size necessary for the results to be generalized to a larger population.

http://books.google.com/books?id=0kajTYfT_U0C&pg=PA12&lpg=PA12&dq=external+validity+sample+size&source=bl&ots=U_SmQLYDh0&sig=58GnXdUTk3xRcpyPYJNp5JhWK5M&hl=en&ei=VHv7Sb7MC9KGmQektYS9BA&sa=X&oi=book_result&ct=result&resnum=8#PPA13,M1

Without even doing the math do you honestly think that a sample size of 742 is large enough to generalize to a population size of 300,000,000 ?
 
  • #21
Gokul43201
Staff Emeritus
Science Advisor
Gold Member
7,051
17
I think you're putting words in his mouth.
This is mind-boggling.

You put words in his mouth. I quoted his exact words. And you tell me that I am putting words in his mouth? Wow!
 
Last edited:
  • #22
Gokul43201
Staff Emeritus
Science Advisor
Gold Member
7,051
17
See pages 13 and onward for ways to determine the minimal sample size necessary for the results to be generalized to a larger population.
You have made the assertion that the sample size is too small. I believe the onus is upon you to calculate this minimum sample size and tell us what it ought to be.

Without even doing the math do you honestly think that a sample size of 742 is large enough to generalize to a population size of 300,000,000 ?
Without doing the math, I think very little about your question. The size of the sample you need depends on the hypothesis you are testing and the confidence level that you desire to test that hypothesis to. If your hypothesis is that there are no Americans with two legs, you can reject that hypothesis with a sample of only one person. If your hypothesis is that there are as many (to say, the nearest million) Americans shorter than 10ft in height than those that are taller than it, you can reject that hypothesis to a 99% confidence level with a sample of fewer than 10 people.
 
Last edited:
  • #23
chroot
Staff Emeritus
Science Advisor
Gold Member
10,226
34
All of you who are interested in this topic should check out this online ebook by U of Manitoba sociology professor Bob Altemeyer.

The Authoritarians by Bob Altemeyer

He has amassed a great deal of survey results, mostly from the endless captive audience attending his undergraduate lectures. His results are often astounding.

Bob Altemeyer said:
Why are high RWAs extra-punitive against law-breakers? For one thing, they think the crimes involved are more serious than most people do, and they believe more in the beneficial effects of punishment. But they also find “common criminals” highly repulsive and disgusting, and they admit it feels personally good, it makes them glad, to be able to punish a perpetrator. They get off smiting the sinner; they relish being “the arm of the Lord.” Similarly, high RWA university students say that classmates in high school who misbehaved and got into trouble, experienced “bad trips” on drugs, became pregnant, and so on “got exactly what they deserved” and that they felt a secret pleasure when they found out about the others’ misfortune.
- Warren
 
  • #24
65
0
The old argument "Is it justified to torture/kill a few in order to save many" is a tricky one, but I will stand up and say "NO", while most would probably say "yes". I personally couldn't live with that inhuman suffering I have caused another lifeform. I believe in throwing up my arms and surrendering to fate. If millions die, so be it. We all must die sometime - does it ultimately matter "how" and "when"?

Nobody deserves torture. You don't fight evil with even worse evil.

Personally I think my stance is very much aligned with the New Testament moral stance - which is ironic, because I consider it all a fabrication (good story, though!).

It doesn't surprise me that "the other half of humanity" would condone without any qualms the use of torture and brutality and even murder in the name of "crime prevention".
 
  • #25
6,265
1,275
All of you who are interested in this topic should check out this online ebook by U of Manitoba sociology professor Bob Altemeyer.
What does RWA stand for?
 

Related Threads for: Survey: Churchgoers more likely to back torture

Replies
9
Views
3K
  • Last Post
Replies
20
Views
8K
  • Last Post
3
Replies
52
Views
6K
  • Last Post
Replies
17
Views
5K
  • Last Post
Replies
19
Views
6K
  • Last Post
Replies
4
Views
2K
  • Last Post
Replies
5
Views
2K
  • Last Post
2
Replies
34
Views
2K
Top