Posts

Showing posts with the label Survey data

How to abuse statistics (again)

Silly statistics, to be taken with a very large pinch of salt.

London to New York - how?

What a silly question!