Should You Lose Your Job If You Can’t Replicate Results?

The question of the title was begged by the recent resignations of top leaders of the OPERA (Oscillation Project with Emulsion tRacking Apparatus) facility.  This facility reported that it had detected faster-than-light neutrinos in September of 2011.  The results were scrutinized – as is supposed to happen – by other researchers, the equipment was checked, and an independent team was unable to replicate the OPERA findings.  The leading explanation for the not really faster-than-light particles is that the measurements suffered from technical challenges that led to the odd figures.  There has been no suggestion of any misconduct, and the researchers were appropriately cautious in announcing their results.

However, some of the OPERA researchers were apparently not happy with how the announcement was handled and/or how the results were checked prior to announcement. Those concerns seem to have contributed to the vote of no confidence that persuaded project leaders to step down.  Some may argue that this was a case of bad experiment management.  I think it worth noting that the honest mistakes made by the project may not be cause for termination, but they can displease your colleagues on the project enough to make it harder for the group to function.  It’s not enough to follow protocol, you need to be a good people manager.

I have to wonder, though.  Would a similar situation with a group of postdocs led by a research scientist that had done something similar lead to a vote of no confidence?  If it had not reached the official level of approval (like this vote did not), would the lead scientist step aside?  I didn’t think so either.  And that’s too bad.