Statistical Mistakes

Misuse of statistical inference in medical research

The author

jrakiyThank you for visiting my blog. My name is Jonas Ranstam, and I am a medical statistician and statistical reviewer that each year review a large number of manuscripts submitted for publication to scientific journals (see my personal homepage for more detailed information).

The reason that many scientific medical journals have statistical reviewers is that medical research heavily relies on statistical inference for estimation of treatment effects, complications, side effects, risks of hazardous exposure, etc. This is a quantitative research area with a highly specialized methodology, and it is generally considered important that editorial offices make sure that the findings presented in published reports do have the empirical support that their authors claim, and that limitations and uncertainties are clearly presented.

The statistical review often reveals serious mistakes and misunderstandings. I maintain this blog to support researchers, reviewers, editors, and readers with references to publications describing the nature of such statistical mistakes, how often they occur, and with guidance on how to avoid such mistakes.