Boffins prove boffinry does not improve statistical ability.
"Similarly, as average IQ in Western societies increased throught the 20th century, so did rates of atheism, he said."
The average IQ is by definition 100.
How did they manage to measure an increase in 100?
Regardless, their statistical correlation between the rise in average *intelligence* (not IQ) and in western Europe suffers under the same problems as climate modelling: our sample set is one. One world, one climate; one Europe, one population. Do they have any data showing the converse of this, ie that increasing religiousness in Western Europe correlates with decreasing intelligence?
This is not statistically significant data!
It is pretty obvious that human beings will show more intelligence as human knowledge and schooling increase and improve. Anything that occurs at the same time as this may be a) related or b) coincidental. We do not have anywhere near enough data to prove either of these.
Statistics for Dummies is available in all good bookshops on the High Street and on-line.
(And they say it's religion that makes people stupid.)