Tuesday, October 24, 2006

A Pair Of Lancets

The noun Lancet means a blatantly false statistical claim - here are two examples.

First, it seems we need to kill off about a quarter of mankind (my ellipsis):
The Earth’s natural resources are being used 25 per cent faster than the planet can renew them, analysis by WWF indicates.

Measurements of crop yields, carbon-dioxide emissions, fishing and the use of forests suggest that Mankind’s ecological footprint is too big to be sustained.

This rests on the premise that technology doesn't improve productivity - which it clearly does since Brit society is rich enough to pay these two zoologists to Lancet:

By tracking the fortunes of 1,313 species of vertebrates from around the world, the report indicated that there had been a 30 per cent slump in wildlife since 1970.

Tropical species, including mammals, reptiles and birds, were the most badly hit of the 695 land-based animals monitored. They declined by an average of 55 per cent, while the populations of temperate creatures have, overall, remained stable since 1970.

But weren't the creatures declining by 55% just paying the penalty for being intemperate? Anyway, the authors save the killer fact until the end:
Wildebeest have declined by 20 per cent in the past 30 years because of encroachments on their migration routes by farmers.
Slam dunk - we're doomed.

The other Lancet is closer to home:

Study Suggests Link Between Cell Phone Use, Low Sperm Count

If you're a man thinking about having kids, you might want to cut back on your cell phone use.That's the preliminary finding of a study done by The Cleveland Clinic, which found statistically that men who used their cell phones the most had poorer sperm quality than those who used them the least.

The lowest average sperm counts seemed to be in men who had the most cell phone use (more than four hours a day); those who didn’t use cell phones seemed to have the highest. Although the sperm count appears to go down with increasing cell phone use, the difference in numbers wasn't significant.

We all know that correlation doesn't imply causality, so it's just as likely that men with low sperm counts use cell phones more. I can thing of several reasons why, none suitable for a family blog. But more likely the sampling was screwed by this (my ellipsis):
The new study (exclusively) included 364 men being evaluated for infertility between September 2004 and October 2005.
So this correlation has been found for men being treated for infertility, hence the study tells us nothing about men in the general population.

This means that if you're not being treated for infertility, it's fine to engage in cellphone abuse. Just lay off the Wildebeests.