To hypothesise or not to hypothesise?

Are there are times, perhaps, when one should not communicate with wider audiences - at least not via the media - unless one does so very, very carefully? The mainstream media (and I'm talking about journalists who are not science correspondents) have little understanding of the scientific method. As most are arts / humanities graduates, to them one "expert" opinion is just as good as any other.

So when an "expert" scientist hypothesises, to the mainstream media their hypothesis has the same weight as actual peer-reviewed evidence. Consequently, as scientists, we need to flag very carefully when we are publicly stating hypotheses, versus reporting actual results.

Two recent examples: "Twittering brains withering, expert warns" - Baroness Susan Greenfield makes a remark in the House of Lords, stating the hypothesis that social networking may influence neural development.

Was Baroness Greenfield reporting the peer-reviewed results of an actual study into this phenomenon, showing neurological degeneration resulting from exposure to social networking sites? No, she was hypothesising - the "evidence" she offered in support of the idea (she had spoken to a teacher of 30 years standing who suspected kids may be becoming less socially skilled - but is that down to more time online, or other social changes in family dynamics?) was anecdotal.

However, because she is (1) a professor of neuroscience, (2) a Baroness and (3) President of the Royal Institution, the media reported what she said as "fact" rather than hypothesis.

Don't get me wrong - I have a lot of time for Baroness Greenfield (anyone who manages to sit in the Lords, do a lot of outreach, run the RI and *still* maintain a strong publication record in their field deserves much respect!). She may well be right; but until that hypothesis is suitably tested with evidence, it needs to be clearly labelled as such, for the good of public trust in science.

To be fair, perhaps she did label it as such, but the label was stripped by the media in their reporting. Here's the Guardian coverage, which quotes her as saying: "It might be helpful to investigate whether the near total submersion of our culture in screen technologies over the last decade might in some way be linked to the threefold increase over this period in prescriptions for methylphenidate, the drug prescribed for attention-deficit hyperactivity disorder". So clearly she is stating a hypothesis to be tested, at least on that aspect.

But given the media's confusion between hypothesis and evidence, is it perhaps sometimes better to say nothing? Here's the Daily Mail coverage: note the phrase "an eminent scientist has warned", suggesting eminence is just as good as evidence...

If you read the readers' comments on the Daily Mail story, they show that the audience are not stupid - several people suggest that there may be other explanations that need to be tested and explored. And the comments there and on the Guardian story show that some people are consequently suspicious of science and the motives of scientists when hypotheses are reported as evidence. Hence the need to be very, very careful when engaging the media on that ground.

Second example:

In a similar vein, Dr Aric Sigmund's paper in The Biologist garnered much coverage along the lines of "How using Facebook could raise your risk of cancer".

With headlines like that, you might be forgiven for thinking that his paper was reporting a careful epidemiological study actually quantifying such a link.

But it was not. To be fair, The Biologist is a peer-reviewed journal, but one that publishes review articles, rather than primary research. Sigmund's paper does review evidence of how lack of social interaction can alter gene expression etc, but the suggestion that using Facebook could increase cancer risk is hypothetical at present.

The Guardian's tech blogger Charles Arthur does give some sensible commentary here.

And here is the excellent dissection of the story by the NHS Choices "Behind the Headlines" webpage, which is a very welcome resource - just a pity it is not more widely publicised (and perhaps more widely used by journalists).

Ok, who really has the responsibility here? Clearly fault lies with journalists who are not familiar with the scientific method and happily interchange hypothesis and evidence. But knowing that problem exists, I think we have to be very careful when communicating on that ground.