Yesterday I posted an analysis (/pmr/2011/10/21/open-access-works-articles-matter-not-journals/ ) of the accesses to the papers from our special symposium (“Visions of a Semantic Molecular Future). In it I claimed that it no longer matters where you publish but what you publish. It was a single anecdote, but I think a very important one. If anyone has similar cases then I think we can make a fairly strong case.
I reported that in SIX DAYS 1247 had accessed the article by John Wilbanks. Now “Accessed” != “read in total” but it means that at least a human has clicked on the link to the paper and presumably allowed some photons from it to be processed. (It’s a bit complex because these papers are still provisional and you have to click through to the full PDF whereas with a final (Open Access) paper you can read the whole thing.
How long does it take to “read” a paper? I’d guess between a minute and an hour. A minute?? Yes – because that’s what many of today’s readers do. I can’t give the link, but I recall Allan Renear describing a study of how young scientists read the literature skipping rapidly from here to there – perhaps 30 papers an hour. I certainly read some papers in a minute (unless the production is so awful that my brain hurts. Why do we have double-column PDF on screens devised for viewing movies. It’s grotesque – but that’s why we pay publishers to produce stuff we don’t want).
So let’s assume that 200 people a day have read John’s paper. Do we have any evidence?
Yes – from the twittersphere. Yesterday there were 40 tweets with the string “Openness As Infrastructure”.
40. Some were retweets (RT) but still people had taken the effort to tweet this. Why? Because they wanted to tell their friends and the world about this paper. They aren’t bots because you can see they are beautiful people, with real identities. Most simply RT’ed, but I fished out the comments:
“Science was like a wiki! ” [a quote from John’s text]
“A very interesting article for open data folk and others”
[Twitter only gives you this much after you have included the title and the @wilbanks]
If people had thought there was something wrong with the article then they’d have said so.
So I think we can conclude this is massive public interest and approval of the article.
In J.Cheminformatics!
The last place you would expect to see an article from John.
Now I frequently hear people say “Oh, don’t publish there – no one will read it and it’s not got a decent IF” (IF == Journal Impact Factor, remember). So here’s a very relevant comment from Mike Taylor, the sauropodovertebrologist:
Mike Taylor says:
October 21, 2011 at 8:42 am
VERY interesting analysis. I wonder if there is a way to move this beyond the anecdotal, and somehow quantify the relative importance of writing a good article and placing it in a good journal.
I recently wrote (http://svpow.wordpress.com/2011/10/17/collateral-damage-of-the-non-open-reviewing-boycott/#comment-11558) that “If you work in an institution where they count up the IFs of the journals where your papers are published, then it may be advantageous to forego actually getting read and cited, for the benefit of rubbing up against other papers that, despite the disadvantage of being in a walled garden, still accumulate many citations. Promotion by the company that you keep rather than by your own achievements.” I’m hugely encouraged to think that we may be further along to road to fixing that system than I’d realised.
I think that for younger scientists this is tragically true. If it had been someone less well known who submitted the article would people have tweeted and read it? I don’t know. Possibly, but slower and in smaller volume. But unfortunately the great-and-the-good of science (with honourable exceptions such as PloS people) are generally saying “you must publish in the highest impact journals.” I hear this all the time. It’s a massive demonstration of the mediocrity of current academic morality.
Because people who cannot read the literature – the Scholarly Poor – suffer. Some die because of it. We don’t know who, but I am certain this is true. I want to see how we can measure this and expose this gross abdication of social responsibility.
Meanwhile, what about peer-review? In @wilbanks we can ask the questions:
- Is it in scope for the journal? Conventional thinking would say no, and some of the reviews for the other papers were somewhat conventional. I didn’t see John’s reviews. But we had agreed with/as editors that all the material was “inscope”. And the readership has confirmed this!
- Is it sound? The twittersphere thinks so, and so do I.
- Is it interesting. Same answer.
- Is it novel? I think we’d have heard “same old stuff” if it wasn’t.
- Was it plagiarised? As before
The point is that “peer-review” isn’t a single dimension. If I am reading a clinical trial I want to know that the stats are valid and the methodology sound. If I am seeing a crystal structure I want to know it wasn’t a duplicate of work published a year ago by the same authors. Although I DO want to see independently repeated scientific experiments as they calibrate the discipline. But for much of the rest the twittersphere is often hugely better than the conventional peer-reviewers.