RE: Athiest nations are more peaceful?

RE: Athiest nations are more peaceful? (Full Version)

All Forums >> [Community Discussions] >> Dungeon of Political and Religious Discussion

Message

Asherdelampyr -> RE: Athiest nations are more peaceful? (6/4/2009 8:24:32 PM)
quote: I'm saying there is no causation argument in the original article. I'm saying that based on the supplied evidence no causation argument can be made. All there is a correlation between two statistics. One could cause the other or vice versa or both could caused by one or more external cuases. I can see your point here I think It would be the same as saying, that there are less blondes in war torn areas, so obviously blondes are more peaceful ((fake example, I have no idea what the blonde population is anywhere))

DomKen -> RE: Athiest nations are more peaceful? (6/4/2009 8:42:48 PM)
quote: ORIGINAL: Asherdelampyr quote: I'm saying there is no causation argument in the original article. I'm saying that based on the supplied evidence no causation argument can be made. All there is a correlation between two statistics. One could cause the other or vice versa or both could caused by one or more external cuases. I can see your point here I think It would be the same as saying, that there are less blondes in war torn areas, so obviously blondes are more peaceful ((fake example, I have no idea what the blonde population is anywhere)) To complete the analogy it could be blondes are more peaceful or blondes are more likely to flee from or die in a war torn area or the reason the area is war torn is because someone is killing all the blondes. A statistical correlation between two things simply cannot tell you anything more than the two things measured are linked in some way.

Asherdelampyr -> RE: Athiest nations are more peaceful? (6/4/2009 8:47:16 PM)
I am not sure that it can even tell you that much It can tell you that it may be linked, but not neccisarrily that it is linked

Kirata -> RE: Athiest nations are more peaceful? (6/4/2009 8:50:36 PM)
quote: ORIGINAL: Asherdelampyr I am not sure that it can even tell you that much It can tell you that it may be linked, but not neccisarrily that it is linked Specifically, correlation is the degree to which two or more things tend to vary together, and the Dunning-Kruger Effect explains why some people continue to insist that it means something else even after this has been pointed out to them. K.

DarkSteven -> RE: Athiest nations are more peaceful? (6/4/2009 8:55:36 PM)
First off, take a look at the countries' map here. The very peaceful places include Canada, Chile, Australia/NZ, Spain, Japan, and Scandinavia. The big nonpeaceful place is the former Soviet Union, where religion has been stamped out thoroughly. China's not looking too good, either, and the religion there is the state. I think that North Korea, also not a very devout nation, tops the list as well. Never accept someone's interpretation of the data. Look at it yourself. 90+% of the "conclusions" reached are usually made up.

DomKen -> RE: Athiest nations are more peaceful? (6/4/2009 8:57:49 PM)
quote: ORIGINAL: Asherdelampyr I am not sure that it can even tell you that much It can tell you that it may be linked, but not neccisarrily that it is linked In this case the correlation is very high. If you check the original article you will find that the author reports the P value, the probability of achieving the same result under the null hypothesis, at 0.001 or less. IOW the odds of this being coincidence are less than 1 in 1000.

Kirata -> RE: Athiest nations are more peaceful? (6/4/2009 9:35:12 PM)
quote: ORIGINAL: DomKen In this case the correlation is very high. If you check the original article you will find that the author reports the P value, the probability of achieving the same result under the null hypothesis, at 0.001 or less. IOW the odds of this being coincidence are less than 1 in 1000. Well, not so fast. Consider this (largely hypothetical) example: A study finds that divorce occurs more frequently when at least one of the partners is born in January, February, or March. The correlation is very high. However, an unrelated study finds that the incidence of schizoid personality traits is higher in people born during Winter months. Subsequent research clarifies that divorce correlates with at least one of the partners having schizoid personality traits, no matter when he or she was born. The first correlation was precisely a coincidence (to be polite). It neglected to consider approximately half of the available data (i.e., divorce in the Southern Hemisphere). Similarly, the correlation that started this thread neglects to consider approximately 60% of the data for "peaceful" countries and approximately 40% of the data for "violent" countries. K.

CruelNUnsual -> RE: Athiest nations are more peaceful? (6/4/2009 9:56:20 PM)
quote: ORIGINAL: DomKen quote: ORIGINAL: Asherdelampyr I am not sure that it can even tell you that much It can tell you that it may be linked, but not neccisarrily that it is linked In this case the correlation is very high. If you check the original article you will find that the author reports the P value, the probability of achieving the same result under the null hypothesis, at 0.001 or less. IOW the odds of this being coincidence are less than 1 in 1000. Which still says nothing about the methodology used to classify the groups that achieves those p-test results nor the validity of the underlying data. [Eg the study is a snapshot at the date of the current study, and says nothing about the history of a country. Its military spending/personnel data is garbage because it doesnt take into account the extent to which a country is defended by its allies.] Further, the article says NOTHING about correlation. They dont present a correlation coefficient, much less a p-test of the correlation coefficient. Ie nothing they have presented says HOW MUCH of the bellicose tendencies are explained by religion. . Why not? the data is there, they could have examined many more tiers of "peace" scores instead of just a single "yes or no" breakdown to develop a robust correlation coefficient. My guess is they dont because once a more robust analysis is done there really is very low correlation.

DomKen -> RE: Athiest nations are more peaceful? (6/4/2009 11:08:37 PM)
quote: ORIGINAL: Kirata quote: ORIGINAL: DomKen In this case the correlation is very high. If you check the original article you will find that the author reports the P value, the probability of achieving the same result under the null hypothesis, at 0.001 or less. IOW the odds of this being coincidence are less than 1 in 1000. Well, not so fast. Consider this (largely hypothetical) example: A study finds that divorce occurs more frequently when at least one of the partners is born in January, February, or March. The correlation is very high. However, an unrelated study finds that the incidence of schizoid personality traits is higher in people born during Winter months. Subsequent research clarifies that divorce correlates with at least one of the partners having schizoid personality traits, no matter when he or she was born. The first correlation was precisely a coincidence (to be polite). It neglected to consider approximately half of the available data (i.e., divorce in the Southern Hemisphere). Similarly, the correlation that started this thread neglects to consider approximately 60% of the data for "peaceful" countries and approximately 40% of the data for "violent" countries. K. I'm just using his numbers. The author is the one reporting a P of less than 0.001. Your example is not a coincidence. There is a correlation between divorce and month of birth. The cause may be tied to incidence of schizophrenia or all three data points may be correlative with the actual cause being some unidentified fourth variable.

CruelNUnsual -> RE: Athiest nations are more peaceful? (6/4/2009 11:19:33 PM)
quote: ORIGINAL: DomKen I'm just using his numbers. The author is the one reporting a P of less than 0.001. and once again, P says nothing about correlation, unless its a P-test of rho, and no rho was presented.

Kirata -> RE: Athiest nations are more peaceful? (6/4/2009 11:22:07 PM)
quote: ORIGINAL: DomKen Your example is not a coincidence. There is a correlation between divorce and month of birth. The cause may be tied to incidence of schizophrenia or all three data points may be correlative with the actual cause being some unidentified fourth variable. With "month of birth," yes. But, the hypothetical study did not report a correlation with month of birth. It reported a correlation with the months of January, February, and March. And it was dead wrong. The finding was purely due to the coincidence of Winter happening to fall in those months in the Northern Hemisphere. quote: ORIGINAL: DomKen I'm just using his numbers. The author is the one reporting a P of less than 0.001. Yeah I realize that, and I don't mean to pick on you. I don't deny it looks impressive. But when you consider that it's based on only half of the data.... pffffft. K.

DomKen -> RE: Athiest nations are more peaceful? (6/5/2009 12:25:09 AM)
quote: ORIGINAL: CruelNUnsual quote: ORIGINAL: DomKen I'm just using his numbers. The author is the one reporting a P of less than 0.001. and once again, P says nothing about correlation, unless its a P-test of rho, and no rho was presented. P does say something about correlation. P is the probability that the result could have occured under the null hypothesis. Rho might be worth calculating but since I'm sure from looking at the data that some of the countries would be tied in one ranking or the other then calculating rho involves using Pearson's coefficient which is a lot of trouble for a data set with 180 entries. You are of course welcome to do the calculation but if P is below 0.001 then the odds are that rho will have a noticeable absolute value.

CruelNUnsual -> RE: Athiest nations are more peaceful? (6/5/2009 1:09:47 AM)
quote: ORIGINAL: DomKen quote: ORIGINAL: CruelNUnsual quote: ORIGINAL: DomKen I'm just using his numbers. The author is the one reporting a P of less than 0.001. and once again, P says nothing about correlation, unless its a P-test of rho, and no rho was presented. P does say something about correlation. P is the probability that the result could have occured under the null hypothesis. Rho might be worth calculating but since I'm sure from looking at the data that some of the countries would be tied in one ranking or the other then calculating rho involves using Pearson's coefficient which is a lot of trouble for a data set with 180 entries. You are of course welcome to do the calculation but if P is below 0.001 then the odds are that rho will have a noticeable absolute value. No, it says nothing about correlation, it only tests the significance of the data. Note they didnt even give their hypotheses for their P. I calculated the rho for the 44 countries that matched in name between downloads of both sets of data. There may be a few missing as a result of different names/misspellings. (BTW--I used the rho between Atheist and the peace score, I forget if the original combined "atheist + not religious", but think it was atheist only. Care to guess what it is? Edit: I also calculated the rho for atheist + not religious and the score, which you would expect to be more meaningful since so few people will declare themselves fully atheist. You can guess that one also.

DomKen -> RE: Athiest nations are more peaceful? (6/5/2009 8:45:19 PM)
Strangely the data sets I looked at contained well in excess of 100 nations not 44.

Lorr47 -> RE: Athiest nations are more peaceful? (6/6/2009 12:05:14 AM)
quote: ORIGINAL: LadyEllen quote: ORIGINAL: DomKen quote: ORIGINAL: LadyEllen The problem with this of course - not mentioned in the comments to the article as far as I read, is that when the shells start falling and the bullets are flying, there aint a single atheist to be found. No wonder then that the religious are more often at war. E Bull. I was an atheist throughout my military service even while being shot at. Bet you wouldnt swear that on the Bible though eh? E Why not? It is just another book of fiction; like the Koran. (That should earn me a one of those edicts calling for my head.) There are plenty of atheists on a battle field because the religious force us to be there.

CruelNUnsual -> RE: Athiest nations are more peaceful? (6/6/2009 6:38:35 AM)
quote: ORIGINAL: DomKen Strangely the data sets I looked at contained well in excess of 100 nations not 44. Strangely, you apparently didnt look at the intersection of the two sets.

DomKen -> RE: Athiest nations are more peaceful? (6/6/2009 9:44:30 AM)
quote: ORIGINAL: CruelNUnsual quote: ORIGINAL: DomKen Strangely the data sets I looked at contained well in excess of 100 nations not 44. Strangely, you apparently didnt look at the intersection of the two sets. Actually I did. Still over 100 nations not 44.

CruelNUnsual -> RE: Athiest nations are more peaceful? (6/6/2009 10:18:04 AM)
quote: ORIGINAL: DomKen quote: ORIGINAL: CruelNUnsual quote: ORIGINAL: DomKen Strangely the data sets I looked at contained well in excess of 100 nations not 44. Strangely, you apparently didnt look at the intersection of the two sets. Actually I did. Still over 100 nations not 44. Then you arent looking at the data set that includes the atheist/religious/not religious breakdown, which has far fewer than 100 and only 52 intersect. I manually edited the other 8 in, and, as you would expect with so many data points, the rho is unchanged. And you still havent ventured a guess. Edit: A belief that prostitution can sometimes be justified has a signficantly better correlation to being "peacful" than atheism. Im sure I could pick out many statistics in the WVS that statisically correlate better than atheism, some that have no "common sense" relationship.

DomKen -> RE: Athiest nations are more peaceful? (6/6/2009 8:49:02 PM)
quote: ORIGINAL: CruelNUnsual Then you arent looking at the data set that includes the atheist/religious/not religious breakdown, which has far fewer than 100 and only 52 intersect. I manually edited the other 8 in, and, as you would expect with so many data points, the rho is unchanged. And you still havent ventured a guess. Edit: A belief that prostitution can sometimes be justified has a signficantly better correlation to being "peacful" than atheism. Im sure I could pick out many statistics in the WVS that statisically correlate better than atheism, some that have no "common sense" relationship. I am looking at that data. Based on which set you draw the data from you can have as few as 57, so not the 60 you claim, or well over 100. http://www.worldvaluessurvey.org/ As to guessing a rho for an incomplete data set that would be foolish. Produce the complete data sets including all the rankings and I'll see if your rho is accurate or not.

Arpig -> RE: Athiest nations are more peaceful? (6/6/2009 9:06:21 PM)
GHA!!! you two. This arguement has descended to the incomprehensible. WTF is a rho?

Page: << < prev 1 2 3 [4] 5 next > >>

Collarchat.com © 2025

Terms of Service

0.0625