
Monday, January 04, 2010

Tuesday, March 10, 2009
It’s becoming apparent that there’s a divide between people who are quantitatively minded, and people who aren’t. I don’t think people in the latter category can understand quite how viscerally people in the former category react against the misuse of numbers.
Nobody likes being bullshitted. I think that’s universal. However, not everyone has a feeling for whether or not a numerical statement is bullshit. The numbers quoted by Amnesty are bullshit, that’s become quite clear. The difference is between those who think that the qualitative argument - that domestic violence is a big problem that particularly affect women - is all that matters, and those who think that the accuracy of the quantitative description of that problem also matters.
I can’t stay on the fence here. Quantitative accuracy matters. The modern world has been constructed by people who cared about the value of quantitative accuracy, from antibiotics to the contraceptive pill to the internet. If you take advantage of these things, you have no place bashing people who are obsessive about quantitative accuracy: your way of life depends on them.
And let’s be quite clear: when people who care about quantitative accuracy criticise your figures, it’s not because they oppose your political positon. It’s not because they’re your enemy. It’s because you’ve got your numbers wrong. Quantitvely-minded people care about this in the abstract, regardless of the political context. In doing so, they are upholding an important value that transcends politics. Truth matters, and nobody should be criticised for upholding that value. - [source]
Wednesday, May 14, 2008
Via the excellent blogs of Dr. Goldacre and The Heresiarch , I read about a new poll by the Rowntree foundation of blessed memory : apparently, religion has become the "new social evil [sic]". How very exciting.
The Times reported the matter (article here) which makes the claims that there was a "widespread belief that faith - not just in its extreme form - was intolerant, irrational and used to justify persecution."
and that
"The researchers found that the "dominant opinion" was that religion was a "social evil"."
Actually, the poll shows nothing of the sort. You can read the study website here, the executive summary here, the complete results here and the "Voices of unheard groups" section here. [The latter is an attempt to compensate for the vast over-representation of white, middle class respondents to the Internet survey, by creating focus groups (with people with LDs, ex-offenders and others). This is discussed in the methodology section.]
The main survey first: there were ~3,500 respondents, with views falling into twelve broad categories, the first 6 described as "dominant", the remainder described as "not dominant but important" . "Religion" comes in at 9, less important than gender equality but more important than immigration (!). However, as is clear from both the summary and the discussion on the main report (pages 30-31), this actually covers both "religion" and "the decline of religion".
Although the "decline of religion" is the smaller category, the "religion" category covers four areas: "erosion of secularism", "the most divisive agent in our society", "undermining rationalism", and "religious extremism". As the quoted responses show, only the middle two categories exemplify the "religion poisons everything" attitude which the Times article implies is so prevalent. I simply don't think the data justify the statements that "widespread belief that faith - not just in its extreme form - was intolerant, irrational and used to justify persecution.", "Britain has had it with religion", or even that ' "dominant opinion" was that religion was a "social evil".' I also think the Times article would be of more practical value if the word "many" were quantified. Anywhere. The "unheard groups" section told a rather different story*: criticisms of religion were not directed at religious faith, but at "the ways in which religion was organised and practices were felt to be problematic". Unfortunately, this fascinating statement was not quantified nor supported with quotes, so we cannot know which aspects of organisation and practice were so problematic. "Religious extremism" [page 19] predictably comes in for criticism **, but more surprisingly so does "an absence of religious guidance in relation to the challenges of contemporary society". Further, "it was felt that religious leaders across faiths should provide some moral leadership, but that this wasn't happening". In both the studies, it's worth noting that government, the media, and big business come in for criticism as well as religion, but the Times chooses not to mention them. Britons are rather less sceptical about the Great Sky Fairy who Intelligently Designed the World than they are about the Great Socialist Fairy who Intelligently Designed the Economy.
And finally, what article about religion would be complete without a quote from Terry Sanderson of the National Secular Society, who was:
I don't understand why someone in favour of a secular society would try alienate people like me, who are religious and support a secular society - but then again, we theists are notoriously stupid.
-------------------------------------
* It is also desperately sad: I defy anyone to read the whole "Voices of unheard groups" report and not be moved. ** Voices Of Unheard Groups, page 19. Hands up anyone who is in favour of "religious extremism"?
Wednesday, April 02, 2008
This week I'll be blogging about The Economist's "Anglo-Saxon" attitudes survey. You can find the survey itself here. I was particularly struck by the throw-away comment:
Do you favour the death penalty for murder? (Yes, always; Yes, sometimes, depending on the circumstances; No, the death penalty is wrong; Don't know)
Britain | USA | |
Yes, always | 21 | 26 |
Yes, sometimes | 53 | 50 |
No | 24 | 20 |
Don't knows | 2 | 5* |
* NB: yes, the American data does add up to 101%: suspect either a rounding error or typo.
UPDATE: The Economist retracts; see next post.
Sunday, March 30, 2008
Kudos to the Economist for actually publishing the raw data of their "Anglo-Saxon attitudes:
a survey of British and American views of the world" survey on the Internet, rather than just cherry-picking random stats and hanging a story from them, as the rest of the British media does. You can read it here, and the article based on it here.
I'm going to blog about it this week, particularly regarding the questions on "crime and punishment" and "religion". I'm a bit puzzled by a couple of sentences in the article, that don't seem to be borne out by the data.
Newspapers and magazines definitely should do this more often.