along those lines one should also look at this NRDD paper:
http://www.nature.com/nrd/journal/v11/n3/full/nrd3681.html
Rather long but has lots of very useful citations.
If you ask industry people it's a know problem that 50% of their own
experiments can't be reproduced but it's even worse if you
Just came across it and thought it's quite relevant with regard to the
latest furor over fraud in crystallography:
http://www.nature.com/nature/journal/v483/n7391/pdf/483531a.pdf
"53 papers were deemed 'landmark' studies ... scientific findings were
confirmed in only 6 (11%) cases".
Mean num
