Thursday, January 19, 2012

Reject statistics to protect prejudice

Nearing the end now of Joel Best's Damned Lies and Statistics and he divides people into four groups:

1) Awestruck: totally uncritically accepting any statistic, simply blown away by big numbers.
2) Naive: 'slightly more sophisticated' than Awestruck, but still accepting. Sincere, and generally assume the sincerity of those who create statistics.
3) Cynical: refusing to believe any statistics, think the numbers are too easily manipulated.
4) Critical: thoughtful, trying to evaluate the numbers and distinguish between good and bad statistics.

Best says that the Naive are the biggest group, but lately I've been coming across a lot of the Cynics. Discussing some social problem, I introduce statistics that challenge or support some argument, and the other person immediately dismisses them as useless, biased, even part of a conspiracy to make the government look good. I like Best's description:
Because the Cynical suspect that "you can prove anything with statistics," they can justify ignoring all the numbers - particularly those that challenge their beliefs.... They may be surprisingly sophisticated when pointing out the flaws in numbers they don't like, although they rarely examine their own side's figures with the same critical eye.
I find this too - the cynics replace statistical evidence with prejudice. Instead of rejecting statistics in favour of having no opinion, they reject it in favour of retaining their old, unsupported assumptions.

To me, statistics is exciting because it can help me to get past my own personal experiences, and therefore my own prejudices. The discovery that we are not stuck with our own personal experiences or, at most, the perspectives of scores or hundreds of authors we read, is a liberating one. With statistics we can view millions of data points, even billions, and track great global trends. Is Ireland a dangerous place? Person A visits Ireland and is immediately mugged. Person B visits and finds it safe and cordial. Only statistics, pulling back for a wider view, can say which is the more representative experience.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.