Rich butts in:
Myxoma is transmitted mechanically by mosquitos and does not replicate in the insect.
Interesting episode (as always). There is a fascinating paper out there which shows videos of mosquitoes probing for blood vessels. Very interesting to see: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0050464
Steen writes: (at virology blog)
As discussed in great length on http://microbe.tv/twiv/twiv-38…, the paper described here has not yet passed formal peer review.
I for one could not care less. I know that the author of this blog peer-reviewed the paper, which provides more confidence than knowing that two or three unspecified anonymous scientists reviewed a published paper. I would rather read a paper without the cognitive bias introduced by knowing that it was accepted by a prestigious journal.
If a serious controversy arises around a preprint that might result in more stringent formal peer review of that paper, which will probably be a good thing.
Someday a preprint will attract public attention and will later definitively be shown to be wrong. This may damage the public trust of researchers and of science communicators. In the mean time, slow dissemination of results and uncritical acceptance of formally peer-reviewed papers are much bigger problems.
Thank you for publicizing preprints. I hope you revisit this paper on TWiV once it inevitably gets published somewhere.
As I was listening to episode 388 (What Could Possibly Go Wrong?) and the discussion around pre-print publications, as a (I like to think science-savvy) layperson, I would have to agree with Kathy. If you are blogging about a paper that has not yet been peer-reviewed, it is worthwhile pointing that out, rather than solely relying on a link to the study. That, at the very least, will let the reader know that they should be looking at the paper through a somewhat more critical eye.
I also have to disagree a little bit with Vincent regarding poor-quality science published through pre-print servers. Vincent mentioned that we would not see an increase in bad papers because scientists want to put their best work forward. While that may apply to the majority of researchers, there are people who are just bad scientists (or even hucksters trying to game the system) that would not otherwise get their papers into a decent peer-reviewed journal. This allows them to get their work indexed and available for citation without worrying about their paper being withdrawn or retracted had it somehow worked its way past peer-review. A layperson may not know what caveats to apply to these pre-print publications, and they may give a veneer of credibility to crank ideas.
Granted, I’m not too familiar with pre-print servers and how submissions work, so these concerns may be moot. Perhaps you could comment on that aspect?
Wasn’t Wakefield’s research published under peer review?
In that very narrow crack that I’m qualified to look through, I see malarkey published in name journals.
There’s also the same author publishing very similar work in a number of different journals.
Something that I find really rotten are the cites that when looked up don’t support the statement. And that’s for facts that just are not applicable. There’s also the cites that when looked up result in cites. Tracking them down results in an anecdote. How are these papers reviewed if the cites were not checked?
I’m listening right now to your discussion of Vincent’s use of a preprint in a blog post, and I have to step up to agree with him. I think Kathy’s right that a disclaimer would be ideal, but I also think Vincent is right that it would make a blog post unwieldy. I’ve had to choose concision over total detail a few times in blog posts, and I reconcile it by saying that when someone asks, I can commend their attention to detail and publicly explain why their caveat didn’t make the final cut.
The real problem I see here is that the public has been led to believe that scientific results, peer reviewed or not, are settled fact just because someone respected talked about them.
Vincent is a peer reviewer or editor for several different publications, and was elected President of ASV by his peers. What is the difference between his reading the paper for the purposes of peer review and reading the paper for the purposes of writing a blog post about it? The main difference that I see is that with the blog post, Vincent risks (small) damage to his reputation if the paper is ultimately false.
If he were asked to review it, he would read it and analyze the experiments, and then decide whether or not the designs and outcomes made the results sufficient evidence to support the conclusions. Isn’t that exactly what he did for the purposes of his blog post, but out in public rather than anonymously?
Certainly a full review would come with viewpoints from additional scientists, but can we not just take Vincent’s blog post and say, “Okay, that’s Reviewer #1’s opinion.”? Sure, he could be wrong, but that’s true of peer review, too. I think this is what peer review is going to look like in the age of preprint archives, and we had better get used to it.
Would there be a difference had the paper been peer reviewed and then written about? Vincent would then just be Reviewer #4, or 5, or 6,000. Peer review misses falsehoods both intentional and unintentional. Plenty of peer reviewed papers are discovered to be wrong by the community–but the veneer of peer review has given the public confidence in these results, the same confidence that they would have in a blog post written by a prominent scientist. Andrew Wakefield’s paper was peer reviewed, and that gave it legitimacy that it didn’t deserve.
Maybe the problem here isn’t that there’s a hazard in believing Vincent’s blog post. Maybe the problem is that we haven’t educated the public enough for them to be widely aware that the same hazard exists in believing any scientific result in isolation. Peer reviewed or not.
Leighland Feinman, PhD (aka John Skylar)
Vincent et al,
I was interested in the dialog among the group regarding the use of pre-print servers to publish work prior to peer review. While I appreciate the desire for scientists to get their data “published”, I wonder about potential conflicts when submitting a manuscript of previously “published” data to a peer-reviewed journal. As you know, many journals require that authors indicate where the submitted data has previously been presented (e.g. meetings) or that they document that the data has not been previously submitted for publication or published elsewhere. I seems to me that submitting a paper to a pre-print server prior to submitting to a peer-reviewed journal could be viewed as double-dipping, particularly if both publications are cited in a CV or if claims for publication date priority are put forth. We all know that meeting abstracts are not treated as proper peer-reviewed publications (hence the separate section in one’s CV) and are not particularly relevant when questions of who published when arise. I certainly see the value in pre-print servers for reporting of studies, but I also see the potential for considerable abuse by those who are trying to pad their CV. I guess institutions will need to add an additional section to their CV template for non-peer-reviewed articles and journals will need to find a way to screen pre-print servers to ensure that submitted studies are novel and not previously published, if that is a criterion for submission.
Note that these are my opinions and do not necessarily represent those of the Federal Government or Battelle.
Mike Holbrook, PhD
Research Leader, Battelle
High Containment Supervisor
National Institute of Allergy & Infectious Diseases
Integrated Research Facility (NIAID-IRF)
A few comments about the preprint discussion in TWiV 388.
The comparison of conference presentations and preprints reminded me of this Retraction Watch post about thousands of retracted meeting abstracts:
The previous TWiV (Quaxxed) gave two reasons why preprints in biology are more of a danger than preprints in physics.
First, a paper about a disease or a cure can be worth a lot of money. Lawyers can get rich. Drug companies can get rich. Investors can get rich even if the paper is later discredited, just by betting on changes in stock price. This is why journals started requiring competing interest declarations.
Second, publishing bad statistics can cause real-world consequences.
You discussed the subgroup of MMR vaccine recipients that showed a higher risk of autism. To me it sounded like an example of the old saying, “1 in 20 two-sigma results is wrong.” But there are people who take those reports seriously.
When I write online I occasionally test my readers by saying something misleading, linking to my source, and asking them to spot how I misrepresented it. I’m hoping to teach skepticism. These experiments have been failures. As some of the TWiV panel said, blog readers are unlikely to click links and less likely than that to understand the page they landed on.
This is why you should distinguish preprints from peer-reviewed papers rather than hoping readers will click the link, see the word preprint, and understand how a “preprint” differs from a pre-publication copy of the peer-reviewed final draft.
I found your podcast somewhere around 2010. I’m very nervous in writing you, but am more grateful than you can know for your efforts to bring science and virology to someone like me (55, few classes past high school, factory worker, mom). I’m just a little concerned though. Since I’ve been listening for awhile, I’m used to the light, intelligent and respectful way you handle all of the discussions brought to the table. So, when the issue of whether it would have been better had you notated “pre-print” in your blog commentary on the Dengue virus antibodies, you seemed defensive. You’re a wonderful host and facilitator, probably a great teacher too. But you have a following now. And people Google, taking things out of context, believing everything they read as gospel. Not your fault and, who knows, you and the TWIV team might be Virology’s answer to Neil deGrasse Tyson or, possibly, Carl S!
To the team, I love you all. I’m grateful you all attend these podcasts with respect and enthusiasm. But even an uneducated science buff like me knows that “pre-print” means the piece hasn’t been “printed”. Hence the “pre” prefix.
PS. Intermittent rain and clouds all day, 40 f, 4 c – ugh!
In TWiV 384, Rich picked Dune (which I agree is among the best sci fi novels of all time), and said he is always looking for new sci fi. I’m here to help! (although please don’t feel obligated to read this whole thing on air) I have a few recommendations for you, Rich:
Ancillary Justice (and its sequels) by Ann Leckie – This is a MUST read. Ancillary Justice is the only novel to win the Hugo, Nebula and Arthur C. Clarke awards. It is spectacular. It has clones and sentient machines and gender politics, and is brilliantly written. You should be reading it right now.
Parable of the Sower by Octavia Butler – published in 1993; Butler was one of the great American sci fi writers but is generally under-read because of her race and gender; I highly recommend all of her books, but this is my favorite.
The Fifth Season by N.K. Jemisin – one of the best books I read last year. It’s not typical space ship sci fi, but it’s not swords and sorcerers either. Jemisin is another author under-read because of her race and gender, but this is her best book to date. The plot twists will rip your heart out.
God’s War (and sequels) by Kameron Hurley – This is not for everyone. Hurley pulls no punches on gender, violence, and individual monstrosity. The writing is not quite as polished as her more recent books (which are more fantasy than sci fi, and I don’t know if you read that genre). But if you want to go for a hell of a ride, Hurley will take you there.
The Passage (and its sequels) by Justin Cronin – if you like government-created vampires, this series is for you. Much better than your average novel about government-created vampires.