Failures to replicate evidence of new discoveries have forced scientists to ask whether this unreliability is due to suboptimal implementation of optimal methods or whether presumptively optimal methods are not, in fact, optimal. This paper reports …