Comments on: De-discovery round-up (plus a correction)

By: Jack

Jack — Wed, 06 Jul 2011 05:48:04 +0000

Thanks for the links to the two other articles – quite fascinating (and actually quite worrying) stuff.

As a layperson I always thought science would replicate as a matter of course. I still believe this happens and quickly too in respect of ‘significant’ perceived threats – but replication of exact methodology – as referred to by Gleiser – does not seem to be insisted upon.

It is the replication of results in the original research that are the focus – and others will try their own techniques to try and do this – take XMRV for instance.

Are exact methods and conditions EVER replicated in science – or even attempted I wonder. IS such a thing even possible? Should such attempts even be made?

After all if through tried and tested techniques the same results cannot be reproduced or are even believed to have been possible in the orginal research (if said research is questionable); then how far should science have to go?

All very interesting. I should like a follow-up piece if at all possible Mr Zimmer.

By: jackd

jackd — Fri, 01 Jul 2011 17:30:13 +0000

Would it be feasible, I wonder, for a journal to run a regular (annual?) follow-up issue, which would summarize replication studies of results previously announced in the same journal? The difficulty would be marshaling the resources for review of the replication papers themselves, I suppose.

By: Michael Meadon

Michael Meadon — Fri, 01 Jul 2011 16:55:13 +0000

As I understand it, there were three unsuccessful replications, which were all described by this one rejected paper. So, in a sense, you’re both right. (As Smith no doubt knows).

By: Joshua Zelinsky

Joshua Zelinsky — Fri, 01 Jul 2011 16:37:54 +0000

One almost wonders if there should be something like “The Journal of Non-Replication” that accepts only papers that failed to replicate results.