Thursday, June 4, 2009

pax atheism?

I recently saw the latest global peace index rankings.

The thing that stood out was the countries with the low scores ("most peaceful"). They're countries well-known for low levels of religious belief. So I thought I'd try to get some data. I grabbed numbers on "no religion" from wikipedia, and looked at the relationship. There were a lot of countries really squashed up at the low end of non-religous percentage (left side), so I did a transform. Unfortunately a large proportion of "0% atheist" countries make a log-transform impossible - so I did a cube-root instead.

So here's my graph of Peace Index (Low = Peaceful) vs non-religious percentage (click for larger)




Interesting, eh?


Issues:
- "wealth" will be a confounding variable (poorer countries will on average be less peaceful and more religious)
- a few of the non-religious percentage figures are a bit rubbery, and my treatment of ranges (taking the upper end) less than ideal.

I'd like to improve this plot by getting other/better figures (e.g. I'd also like to try some of these), giving ranges where ranges were given and giving less weighting by the width of the range (make them smaller or paler or something, if the figures are really uncertain). I'd like to use other figures, and see what adjusting for wealth (say GNP per capita or somesuch) does. Just looking at some of the wealthy "western" countries, it looks like the effect remains, but would be weaker.

I'd also like to label the points!

But this is a good start, and I've spent all the time I can on it for now.

3 comments:

Anonymous said...

If you click the name of the nation in the GPI list, and scroll down to the bottom of the nations page, there is a culture section.

In that section is "importance of religion in national life"

Qualitative assessment of the level of importance of religion in politics and social life. Ranked 1-5 (very low to very high) by EIU analysts.

Source: EIU Analysts; Year: 2008

Since you wouldnt have to deal with percentages, it would be clearer to use it in a chart.

Skipinder said...

Correlation, causation. Without some treatment of confounders that's pretty much meaningless except as a statement of fact.

Anonymous said...

Dear Ecstathy,

Just to introduce myself, my name's Ellie and I'm writing on behalf of the Global Peace Index (GPI).

First of all we would like to sincerely thank you for previously including the GPI in your blog in the past.

As you'll be aware, the GPI is a ground-breaking piece of research in the study of peace, which not only ranks nations by their peacefulness but also seeks to identify the drivers of peace. Now in its fourth year, the Index is produced by the Institute for Economics and Peace, a global think tank dedicated to the research and education of the relationship between economic development, business and peace. It is collated and calculated by the Economist Intelligence Unit.

We're getting in touch to alert you to the fact that the next Global Peace Index will be launched on June 8th 2010. We'll be providing you with further information closer to the time - unless you tell us that you'd rather not hear from us again - but thought you might appreciate the heads-up in the meantime.

You might want to join us on Facebook too:

http://www.facebook.com/#!/pages/Global-Peace-Index/92288319748?ref=ts

and follow us on Twitter

http://twitter.com/GlobPeaceIndex

If you've got any thought or questions about the Index, our methodology, or our organisation, please get in touch. We hope you find our research interesting and informative, and look forward to sharing this year's findings with you in a few weeks' time.

Best Wishes,

Ellie Kirby - updates@visionofhumanity.org