Saturday, May 08, 2010

Statistical competency

A friend posted this on facebook the other day, and I've been thinking: yes, lack of knowledge of even basic statistics leads people to completely moronic opinions. (See the comments on that article for an example.) On the other hand, when you have a little bit of statistical competency but not a lot, you get things like this:
For the first time in history, several countries have transferred directly, from a non-market to a liberal market economy. They stayed at the same technological level, and kept full statistical records of the effects. It is now possible, for the first time, to estimate the free-market contribution to death rates. These effects are significant, and huge numbers of people are involved. In Russia crude death rates rose from 10,7 per 1000 in 1989 to 15,8 per 1000 in 1994. Comparing Europe in 1500 with Europe in 2000 will not show the effects of the free market: society and technology were fundamentally different in 1500. But Russia in 1994 had the same technology, the same urbanisation, the same infrastructure as Russia in 1989. What changed was that a centrally-planned regime collapsed, and a liberal market system replaced it.
Yes, everything stayed the same except the regime collecting the statistics, so this is a perfectly valid inference. Read more.