Abstract
The paper describes various faulty assumptions made by both users and producers of statistical software. It serves as a warning to those who use the various software packages that it is often easier to produce misleading results than it is to properly analyze data.
Full text
PDF



