Religion & Torture Trends
In what is sure to be my most controversial post of the year, I thought it was interesting and ironic that a recent Pew Research poll showed a correlation between religious service attendance and the tendency to support torture against suspected terrorists. White evangelical protestants clocked in at the highest approval rate (62% said torture was "often" or "sometimes" justified), while those not affiliated with a religious institution had a 40% approval rate.
Anyone want to crank out the correlation coefficient on this? I suspect (by eyeballing it), that it's around .3, but I'm about to sit down and crank out a final, so I'll have to drag out my statistics textbook later.
Anyone want to crank out the correlation coefficient on this? I suspect (by eyeballing it), that it's around .3, but I'm about to sit down and crank out a final, so I'll have to drag out my statistics textbook later.
Labels: Rabid Conservatives, Rabid Liberals
posted by Toney at 11:00 AM
30 Comments:
I don't understand why "white evangelical protestant" is comparable to the entire universe of "unaffiliated" people. Shouldn't you ask only unaffiliated white people, or ask all evangelicals, to be sure that the correlation is between religion and views on torture, and not race and views on torture?
I agree that there may be an omitted variable here but I don't know if it is race. White evangelical protestants are also highly likely to a) be Republican and b) like the television show "24" where torture is used to great effect.
Also there are many torture polls suggesting that up to 46% of Americans support it in some cases--too lazy to look up source right now. Makes me think this data is not all that interesting.
"...a recent study shows that berkeley residents are the most likely to run around naked and shit in public parks..."
there, now the post is fair and balanced.
I read the study and now realize the second part of my comment was totally asinine.
I don't think this poll is meaningful, because the topic is obviously a controversial political hot potato.
Instead, I think it merely reflects a cascade of correlations: (A) regular church-goers are likely to be Republicans; (B) Republicans are likely to be sympathetic to the perspective of the prior administration; and (C) the prior administration has defended its harsh treatment of detainees as a justifiable necessity.
So tempted to make generalizations about evangelicals. But I shall refrain. You could just substitute "Republican" for "evangelical" and "Democrat" for "non-church going" (or whatever it was) and likely get the same results. Nothing shocking here. OK, one generalization - can't help myself. Lack of self-reflection and reluctance to question authority are common characteristics of evangelicals and Republicans. I.e. "if the guv'ment tells me torture is fine, it must be fine"
that last comment is just completely ridiculous.
Montana (a fairly red state) just passed a law saying that all guns manufactured in the state, for intrastate use, are immune from federal regulation.
I do believe that is questioning authority.
And what exactly does "lack of self reflection" mean?
Montana did that??? God, I love red states!
Don't get me wrong -- when it comes to personal geographic preferences, I'll take paved roads and non-creationist primary education every time. But when it comes to making my heart beat with patriotic glee, gimme' the Montana's "leave me alone so I can shoot guns and go fishing"* every time!
*query: does this quote contemplate one activity, or two?
I think what is the more frightening aspect of all of this is that clearly a unbelievably large percentage of Americans, of any religion, support torture.
Look at the question. One of the options is "in very rare circumstances." Usually, the largest group chose that option, which still counts as supporting torture.
Alot of people believe that there are times where that would be needed. Hell, even if I can't think of one, I would be tempted to mark that option on a survey, just to cover my bases.
The ticking time bomb scenario comes to mind. That doesn't mean I will go all Jack Bauer on your ass.
Surveys can have a lot of errors depending on the sample population of the respondents. 742 is a very small sample for the US population and thus this research has a very very high "margin of error" in statistical terms. The problem is the media likes to sensationalize a research on religion especially when it is so negative without really checking the validity and reliability of the research. Another research should cross-validate the data on the news with a bigger population sample from different parts of the US. As for me and most statisticians this recent kind of research is most likely unreliable, invalid, and has a very wide range of "margin of error" aside from other variables that led to the conclusion. Most likely the one who did this research has a hidden agenda/motive against religious people, particularly to Evangelical and Catholic Christians.
"742 is a very small sample for the US population and thus this research has a very very high "margin of error" in statistical terms."
What does the size of the US population have to do with the margin of error?
Perhaps, as a statistician, you can explain that. In statistical terms.
You know, sometimes I really hate the irrational and inflammatory liberal politics (see, e.g., 3:47) that can characterize our school.
Non-affiliated survey respondents tend to be clumped at either end of the torture scale (often be justified, or never), while religious persons tended to cover the middle ground (sometimes be justified, or rarely). If anything, that shows that affiliated people tend to hold more complex, more nuanced positions, than non-affiliated. It doesn't mean they're reactionary or un-critical, and it doesn't make them moral sheep.
Actually, the US population has everything to do with the margin of error.
If the population of the US were 10, then a sample pool of 8 would be fairly reliable. However, if the US population were, say, 300 million, the sample pool of 8's margin of error would be astronomical (close to 100%).
What doesn't matter is the motive or agenda of the people behind the survey, so long as scientific norms are followed.
I agree the sample pool is small, though ~750 is enough to get the margin of error down to 3.5%. The problem is that this survey actually sub-divides the sample pool into several smaller groups between 94 & 336 in size. For the purposes of the survey, the margin of error of each is independent from the rest. A margin of error of a group of 94 is very large indeed (around 10%). I think the fact each group is further broken down into one of four answers probably makes this even worse. Anyone want to correct me?
"If the population of the US were 10, then a sample pool of 8 would be fairly reliable. However, if the US population were, say, 300 million, the sample pool of 8's margin of error would be astronomical (close to 100%)."
This is wrong.
Look at the formula for the standard error of a sample statistic using a simple random sample. If you think the population size matters, please tell me why it appears nowhere in that formula.
Here's the formula for the standard error of the sample mean using a simple random sample:
http://en.wikipedia.org/wiki/Standard_error_(statistics)
It's SD/sqrt(n), where SD is the sample standard deviation, and n is the sample size.
Where is the population size in this formula?
Sorry anonymous, you're wrong (mostly). You refer to standard error, but you should be looking at the margin of error, which relies on sampling error.
Margin of error is often calculated under the assumption that the population is infinite, but this assumption is only reasonable (apparently) when the sample pool is less than 5% of the population. If you can't see the effect of the population size on the ridiculous example that I gave, then I just don't know what to tell you (well, I do, but I'm trying to be nice).
ARG (though a less than reliable pollster) has an easy online interface for calculating the margin of error of a poll here.
This has just turned into a bad case of social "science" majors trying to justify their 4(+) years of undergrad existence.
This comment has been removed by the author.
Whatever. I have degrees in electrical engineering and math, which makes me much snobbier when it comes to petty issues like this.
Wow. You guys don't need anonymous commenters to be snarky anymore. You're all about taking ad hominem attacks at eachother now.
As a frequent anonymous poster, I find this development disturbing. The role of snarky sarcastic commenter has traditionally been the province of the anonymous. Sadly, named posters have used recent events to usurp this time-honored and finely tuned balance of power.
Stop it.
let's chalk it up to finals grumpiness. all is well in less than 2 weeks...
speaking of that... can't we have a witchhunt or something?
haha- yea, that helped 1L year. never did figure out who the dude (female?) was. good ol' trusti...
chances are he's someone who didn't donate to the class campaign though. should be easy to spot with the whole asterisk thing in the grad book...
haha jk
oh boy. i managed to spend a whole evening avoiding studying by following that witchhunt. good times.
was it really 1L year? where have the years gone? [sniff]
I know who Trustifarian is. Please put $5 in my locker, and a note including your locker number, to find out the truth.
Are you forgetting that great maxim of the common law, Mcwho? No man should profit from his own wrong.
oh snap
I wasn't in law school at the time but remember hearing about this Trustifarian character. What ever happen with that situation? Did people ever figure out who it was, were they expelled, criminal charges brought?
Post a Comment
<< Home