Friday, March 22, 2013

Pre-review publication is not outside the scientific method



I've been thinking about this post for a while, and the departure of Martin Ferguson provides opportune moment to actually write it.

Back in November, some scientists conducted research (which is their right), and included some of their research in a submission to a government inquiry (which is their right), even though the research hadn't completed the peer-review process (more on that in a minute).

The research appears to find high concentrations of methane at ground level around CSG fields – and the submission included sufficient information that anyone else with appropriate equipment and expertise could try to replicate it. Ferguson, an apologist for industry, was http://www.smh.com.au/environment/minister-slams-unscientific-report-on-gas-leak-20121120-29nj5.html angry.

From the story: “Mr Ferguson said he believed the study … abandoned usual scientific practice”.

Bollocks.

He criticised the study's public release, before it had been peer-reviewed, saying that in "the scientific community that is not regarded favourably".

Bollocks.

“Conduct yourself in a professional way and focus on the outcome, not short-lived media opportunities”, he is quoted as saying.

Bollocks.

1. Did publication “abandon usual scientific practice”?

No. Scientists are free to do what they will with their data. Peer-review and publication aren't “science” per se. Peer review exists to provide a quality control mechanism before dissemination. The usual scientific practice – the boring stuff of making observations, conducting experiments, constructing hypotheses, and providing enough information about the work to permit replication – remains intact.

2. Releasing results is “not regarded favourably”?

That's a political response, not a scientific one. There even exists an entire scientific archive – Arxiv.org – that allows scientists to publish “pre-press” versions of their papers. Anyone can download those papers. A great many scientific releases I receive, including from the Australian Science Media Centre, which joined in criticizing the scientists, end with "this research has been submitted to journal X".

When the Higgs-Boson results started to emerge last year, Arxiv (among other places) received all the pre-press stuff. Did the Large Hadron Collider researchers do something “not regarded favourably”? What utter nonsense.

3. Short-lived media outcomes

The data was given to the government inquiry, not to the press. There never was any reason to accuse the scientists of being media whores.

Publication is publication, science is science

Somehow, in the public's mind, a piece of science that isn't science has been incorporated into “scientific method”. It's not part of “the method” - it's a publication process. The scientific method – observation, hypothesis, prediction, experiment – works even if you don't publish.

Publication permits replication; and peer review is simply an evolution of the editorial process, because no one person has enough knowledge to distinguish between good science and bad.

A decision to publish information ahead of peer-review is neither uncommon nor beyond the pale. It's just that in this case, it was something the minister – a tireless advocate for rich industries that are big enough and ugly enough to take care of themselves – didn't want to hear.

No comments: