Saturday 5 July 2014

STAP retractions are both a failing and a triumph of science

It was looking inevitable and this week two high profile Nature articles on “STAP” (stimulus-triggered acquisition of pluripotency) stem cells were finally retracted in Nature:

Several critical errors have been found in our Article and Letter (http://dx.doi.org/10.1038/nature12969), which led to an in-depth investigation by the RIKEN Institute. The RIKEN investigation committee has categorized some of the errors as misconduct (see Supplementary Data 1 and Supplementary Data 2). Additional errors identified by the authors that are not discussed in RIKEN’s report are listed below.

...

We apologize for the mistakes included in the Article and Letter. These multiple errors impair the credibility of the study as a whole and we are unable to say without doubt whether the STAP-SC phenomenon is real. Ongoing studies are investigating this phenomenon afresh, but given the extensive nature of the errors currently found, we consider it appropriate to retract both papers.

Nature cover the retractions in an editorial, “STAP retracted”, which runs with the tagline,

“Two retractions highlight long-standing issues of trust and sloppiness that must be addressed.”

You can get a sense of those issues from the retraction statement and the editorial, which concludes:

“we and the referees could not have detected the problems that fatally undermined the papers. The referees’ rigorous reports quite rightly took on trust what was presented in the papers.”

They also highlight “sloppiness” in science, manifest as a “growth in the number of corrections reported in journals in recent years”. (Something not helped, in my opinion, by high profile journals such as Science and Nature burying so much of the important methods in Supplementary Data, which is rarely reviewed or edited as critically as material in the main text body.)

You can read more about those issues in the editorial and elsewhere, such as the Faculty of 1000 blog. The STAP papers, their initial irreproducibility and eventual retraction highlight potential failings of the current scientific system, which places far too much emphasis on output quantity and impact rather than (true) quality and integrity.

However, they also highlight the tremendous success of the scientific system.

The fact is, the experiments were repeated, the failure to reproduce results was documented, suspicions were raised and investigations made. Science works because, ultimately, you cannot fake it. Whatever data you make up, whatever results you misinterpret, whatever sloppiness leads to “conclusions [that] seem misleadingly robust”, the truth will out eventually. You cannot hoodwink nature.

And that is why science remains far and away the best (probably only) method we have for establishing the truth about reality. The system maybe flawed, it may waste money and it may lead poor unsuspecting suckers chasing wild geese, but eventually it will self-correct. So, whilst I would never put my trust in individual scientists (unless they have earnt it) or results, and I remain skeptical of every new claim, I still emphatically trust science itself.