Deevy Bishop: Editors behaving badly?. It starts with this (and goes down from there!):
... I found that for 32 papers co-authored by Matson in this journal between 2010 and 2014 for which the information was available, the median lag from a paper being received and it being accepted was one day.
Scott Burns in Dallas Morning News: Tortured data will confess to anything. Mentions Uri Simonsohn and p-hacking. And, of course, torture of data.