05 November 2005

Statistics and Trends

The human brain is remarkably good at finding trends, even where none exist. If we have a small amount of evidence that we think points in one direction, our brain will then have a tendancy to see data that supports the claim, while not noticing data that refutes it. For example, suppose someone working on human archaeological remains dies. People will laugh and scoff and someone will inevitably make a crack about the curse of the mummy. No one will believe it, and the rumor will go away. Then a second person dies. Suddenly the story is dragged back out, and someone goes through the obituaries of the world for the past X years and finds the names of everyone else in contact with the remains who died, and suddenly a full-blown curse is "established." Never mind how many people worked with the body that didn't die. It's a curse! All the deaths prove it!

It's easy for people to believe in it, and it makes a good story. To rigorously disprove it, we'd have to pick some other object, not necessarily archaeological, say my car. How many people have come into contact with my car since its "discovery" (creation) 18 years ago? How many of those people have died? Make it a percent, or a rate. Now do the same with some other random object. And another, and another. What's typical? What's statistically acceptable as random chance? NOW compare it to the "mummy's curse" deaths.

A lot of work, neh? Easier to just believe it. Not quite an unfalsifiable statement, as it's possible to test, but not a good choice of hypothesis to be disproveable.


It is not known how many people have worked on the Oetzi project - and whether the death rate is statistically high. ...

Dr Loy's brother Gareth said the two had never talked about a curse - and that Tom Loy had been in poor health, with a condition that caused his blood to clot.

An inquest into Dr Loy's death was inconclusive, ruling out foul play but unable to determine if he had died of natural causes, an accident, or both, Gareth Loy told The Australian newspaper.

An unnamed colleague of Dr Loy scoffed at the idea of a curse, the newspaper reported: "He didn't believe in the curse. It was just superstition. People die."

(BBC)

1 comment:

EKENYERENGOZI Michael Chima said...

The fossils of two copulating lovers were found recently in India dated 65 million years.

Is this scientifically possible?

Can we have fossils of our sex organs when we all know that the human male organ has no bone? And skeletons have no sex organs?