Initial reports are not a done deal

By Grant Jacobs 11/01/2012 2

In offering his opinion on a recent research paper reporting that plant micro-RNAs can be found in human blood, Larry Moran closes ’How do you convey the idea that all scientific results are preliminary until they have been confirmed by others?’

I’ve previously written on a related issue, conveying the state of play in media reports of scientific work.

Scientific reports are an argument for a case, the starting point of a line of thinking for the wider scientific community, not the end point. The end point is when, after much cross-examination of the ideas presented in the initial report and subsequent work, the scientific community as a whole reaches a general consensus that the ideas presented are indeed the best current model for the particular thing under study.*

One reason some writers might choose to use definitive (uncritical) statements when writing about scientific reports for articles meant for a wide audience is concern over losing their readers’ (or editors’) interest.

Looked at this way, we can consider Larry’s question as more just raising the preliminary nature of new work, but how this might be best presented in writing an article for a wide audience.

My local newspaper, the Otago Daily Times, presents a section running historical or interesting articles from the files, typically from 50, 100 and 150 years ago. One I can recall** presented a initial report on a multiple murder case from (I think) 100 years ago. Having few facts at hand, the journalist elected to present the murders almost entirely as well-written intrigue.

There is more than one way to skin a cat. There’s nothing wrong with presenting things as exploring what might be.

What might help to present the tentative nature of initial reports positively is to simply take that standpoint from the onset.

If an article starts with a ‘definitive’ style of presentation, stating things as just being so, then later adds caveats to what has been written it can weaken the piece. The reader might feel a little cheated of the time they’ve invested in reading the article. The writer may wind up inviting readers to question everything you’ve written, the good along with the bad, triggering that common remark that scientists have no idea what they are talking about.

Perhaps lead the reader on a journey of curiosity, rather than ’it’s like this, but maybe not really’. I’m sure readers have seen the latter before and experienced first-hand how annoying it can be. By starting with a questioning, investigative approach nothing has to be ‘undone’ later.

Of course, scientific papers also have ‘facts’, so they ought to have something that ‘is’. There is the data and then the interpretation of the data.

That’s another distinction I don’t see often enough in media reports. Statements saying what was measured and separately what interpretations were drawn.


* Note that’s best current model, not something fixed in stone.

**  Sorry I don’t have a link for this; I’m not sure it’d even be on-line (I read it in print).

Some related articles on Code for life:

Scientists can’t write?

Note to science communicators–alleles not ’disease genes’

Banished from science writing. Words, that is.

Media thought: Ask what is known, not the expert’s opinion

When the abstract or conclusions aren’t accurate or enough

Of use of the active voice by scientists

XMRV prompts media thought: ask for the ’state of play’

2 Responses to “Initial reports are not a done deal”

Site Meter