Weak statistical standards implicated in scientific irreproducibility

The NIH should require statistics training in the same way the require ethics training for all life sciences graduate students.