I fail to see the point of a study that can't be replicated. Assuming the replication is performed correctly, failure to replicate pretty much means either deli

I fail to see the point of a study that can't be replicated. Assuming the replication is performed correctly, failure to replicate pretty much means either deliberate fraud (unlikely) or an honest "mistake" (or whatever word you prefer to use to describe a situation where you thought you measured and described a real-world effect in a manner that's in any way useful, but in fact you did not).