Tuesday, November 27, 2012

AusSMC: goodbye. You're not there to censor science

I have just asked the Australian Science Media Centre to no longer consider me a member.

Here’s why.

First let me set down my scientific credentials: I have none. I am a journalist with a strong interest in science, and – I hope – a functioning sniff-test on what I will and won’t write about. Their job was to educate me, but it seems I have to turn tables.

My only credential is that The Register (http://www.theregister.co.uk), for whom I write (http://search.theregister.co.uk/?author=Richard%20Chirgwin), gets millions of hits in any given week (sometimes on a good day), and my science stories do well enough that nobody tells me to lay off science stories.

The Australian Science Media Centre has seen fit to upend a very public bucket (http://www.abc.net.au/environment/articles/2012/11/27/3639625.htm) on two Southern Cross University scientists for “media coverage by press release without a peer reviewed scientific paper to back it up?” asking “whether releasing preliminary data to the media is ever warranted”.

Let’s start with the hypocrisy. For the AusSMC to set “peer review” as the benchmark for “talk to the press” blithely ignores its own patron, the Baroness Greenfield, who just as blithely ignores peer review (for example http://www.guardian.co.uk/science/the-lay-scientist/2012/feb/27/1) when it comes to her theories about social media and brain development.

Also, there’s this exceptionally silly statement.

“The scientific process goes something like this: a researcher constructs a hypothesis, runs experiments to test their hypothesis, gathers data, interprets the results and then puts the lot through peer review”.

Bollocks. Nonsense. That’s how the scientific publication process goes. Science is messy. I’d suppose the most exciting words that an elder professor can hear from a PhD candidate are:

“That’s odd…”

And I have an example, here (http://sydney.edu.au/news/science/397.html?newsstoryid=3204). Speaking at the Australian Institute of Physics recently, CUDOS’ Dr Ben Egglestone was more frank about the researcher’s puzzlement. Actually, the first thought was that a bit of apparatus (presumably expensive) was broken.

The observation came first; after which came the hypothesis; after which the experimental test. After which, the paper.

Back to the CSG issue.

The first public discussion of the work by Dr Damien Maher and Dr Isaac Santos was not, as far as I can see, this press release (http://www.scu.edu.au/news/media.php?item_id=6041&action=show_item&type=M) as asserted by AusSMC, but rather this (http://www.climatechange.gov.au/government/submissions/closed-consultations/~/media/government/submissions/csg/CSG-20121109-CentreForCoastalBiogeochemistrySCU.pdf - PDF) submission to a Department of Climate Change’s inquiry.

I see nothing remotely improper about a scientist contributing to a government inquiry, even pre-peer-review.

It seems the press release was issued after the Sydney Morning Herald noticed the submission and put together this story: in other words, the press release was probably intended as a media summary after every man and his dog started calling up the University.

Which brings us to the question “whether releasing preliminary data to the media is ever warranted”?

To be polite, don’t be silly: is the world now to start censoring its scientists solely on the basis of whether a journal has accepted a particular item of research for publication? Sure, it’s good business for the big journals, but as a journalist, I acknowledge no obligation whatever to protect their business model.

In the specific instance of the SCU submission to the government inquiry: the document repeatedly makes clear that it is presenting preliminary results. The researchers say that their measurements are incomplete – they “provide evidence for significant but unquantified” emissions, and call for “baseline studies” before new projects are commenced.

Ahh, someone or other complains, but they didn’t release the raw data, so nobody else can test it! No: because the raw data is off with a publication undergoing peer review. It’s stuck in the “process” that the AusSMC is promoting.

More broadly, suggesting an extension of peer review from its proper(ish) role – ensuring that the science is sufficiently rigorous to justify publication in a specific journal – into a pre-publication self-censorship is an awful idea.

First, keep in mind that “peer review” isn’t magic. It means “this result is robust enough to warrant publication” – after which the real business of “replicate it or rip it to shreds” begins. The journals do not replicate an experiment before they publish: that is the job of other scientists, after they’ve got their hands on the data.

Apply a “pre peer review” gag? So that no scientist can ever answer the question “what are your current research interests?” So that all science journalism is forever beholden to the embargoes and fanatical media management of the large publishers? So that journalists can see nothing, read nothing and know nothing except by the grace of the journals – while laying out $20k in annual subscriptions?

Should Cornell University pull Arxiv because a journalist might download a document that hasn’t yet been peer-reviewed?

To think that the Australian Science Media Centre wants to filter “science” through the lens of the “science publisher’s” world view is a depressing thought indeed.

Goodbye. I don’t wish the AusSMC bad luck, because – to paraphrase Archie Goodwin – even with good luck, it won’t get much of an epitaph.

No comments: