The exposure of fabricated numbers in published papers by eagle-eyed readers has been due to sporadic serendipity. I am going to describe a semi-automated method that you can take away with you to do some sleuthing. I am going to describe what I found when I analysed over 4500 papers.
