We experienced technical difficulties in the first segment. Google was not switching the camera to the person who was speaking, and Bob’s jokes were not generating laughs. These two phenomena are in all likelihood connected.
This week on the “Virtual Skeptics”…
- Bob completely makes up a news story;
- Eve reveals who debunked Ancient Aliens
- Sharon wonders about the qualifications for “news journalists”;
- and Tim shows how Facebook is peeking at our skeptic homework.
Here’s the first part with Bob’s segment:
After Google just hanged up on us, we restarted here:
Much, much more including the full text of Bob’s piece below…
- Brian Gregory – Host and creator of Virtual Drinking Skeptically
- Eve Siebert – Editor and blogger at Skeptical Humanities
- Bob Blaskiewicz – Committee for Skeptical Inquiry’s Conspiracy Guy web columnist, JREF Swift Blog contributor, and blogger for Skeptical Humanities
- Sharon Hill – Editor of Doubtful News and author of the Committee for Skeptical Inquiry’s Sounds Sciencey web column
- Tim Farley – JREF Research Fellow, creator of What’s The Harm and and blogger at Skeptical Software Tools.
- A news report about plagiarism: http://arstechnica.com/science/2012/10/research-fraud-exploded-over-the-last-decade/
- The original study they stole it from: http://www.pnas.org/content/early/2012/09/27/1212247109.full.pdf+html
- Retraction Watch, a resource the researchers consulted to determine causes of retractions: http://retractionwatch.wordpress.com/
- Charles Seife’s report on Jonah Lehrer’s journalisitic misdeeds http://www.slate.com/articles/health_and_science/science/2012/08/jonah_lehrer_plagiarism_in_wired_com_an_investigation_into_plagiarism_quotes_and_factual_inaccuracies_.html
- Ancient Aliens Debunked http://ancientaliensdebunked.com/
- Doubtful News story, with comments from the filmmaker Chris White http://doubtfulnews.com/2012/10/ancient-aliens-debunked-by-biblical-literalist/
- Jason Colavito’s review http://www.jasoncolavito.com/1/post/2012/09/reviewing-ancient-aliens-debunked.html
- Sitchin Is Wrong http://sitchiniswrong.com/
- Texas man makes interesting find while fishing – a giant tooth http://doubtfulnews.com/2012/10/texas-man-makes-interesting-find-while-fishing-a-giant-tooth/
- Media trips up on Amityville house-for-sale story http://doubtfulnews.com/2012/10/media-trips-up-on-amityville-house-for-sale-story/
- Facebook scans private messages for URLs & counts them as Likes http://nakedsecurity.sophos.com/2012/10/08/facebook-scans-private-messages-like-counter/
- Tim’s original article on why you should be careful about links was called Robots Don’t Get Sarcasm
- Privacy Fix is a new Chrome plugin for privacy settings on websites https://www.privacyfix.com
- Congratulations Seth Kalichman of University of Connecticut who just got $100,000 from Bill Gates to build an anti-vaccine surveillance & alert system http://www.grandchallenges.org/Explorations/Pages/GrantsAwarded.aspx?Topic=Vaccines&Round=8&Phase=all
Bonus Material! Bob’s story. With bonus jokes, just to spite his fellow panelists.
This week, I am talking about what lying, snivelling, rotten deceivers scientists can be. A study published last week in the Proceedings of the National Academy of Sciences by–and the authors have awesome superhero names–Ferric Fang, R. Grant Steen, and Arturo Casadevall, the most interesting man in the world, examined in some detail the reasons behind retractions of articles in scientific journals. This study had a couple of advantages over previous studies that examined retractions. It had a larger sample size than previous studies, the researchers pooled information about retractions from a wider variety of sources than had earlier researchers, and they seem to have aggressively followed up on the reasons behind retractions, and we’ll see why this is important in a minute.
The most interesting finding, and the one that made this newsworthy was the substantial increase in the number of retractions due to misconduct rather than simple error in comparison to previous studies. Looking at all the biomedical and life-science research papers in PubMed that had been retracted before May 3, 2012. A mere 21.3% of retractions were due to error. Instead, a whopping 67.4% of retractions were due to misconduct. To break that down, 43.2% of the retractions were due to fraud or suspected fraud, 14.2% were due to multiple publications of the same article, and 9.8% were due to plagiarism.
The study confirms other observations that the frequency of retractions is rising in recent years. Now, I read about this first in an Ars Technica piece that I am linking to in the show notes, and the headline was “Research Fraud Exploded Over the Last Decade,” but I want to qualify that. That does not seem to be exactly what the authors are saying. Detection of fraud seems to be on the rise at the very least. Retractions seem to be a very recent publishing phenomenon, at least in the data set of the 25 million PubMed articles that they examined. All of the retractions in this study were logged after 1978.
Now, recently in my rhetoric and argument seminar I have recently been working with students on a similar topic, about how an author can lose credibility, and the example that I used was science writer Jonah Lehrer. And what jumped out at me while I read this study was how closely the misdeeds of scientific publication tracked the types of concerns that have been raised by journalist regarding Lehrer’s work. Leher, as you know, was a popular science blogger whose work and integrity has been called into question following a plagiarism scandal. A journalism professor, Charles Siefe, was called in by Wired to investigate claims made against Lehrer, and his findings were published on Salon.com. And what he found was that Lehrer was systematically recycling his old work, was recasting press releases as interviews, faking interviews, and occasionally just plagiarizing other writers.
I made a graphic of the relative frequency of the types of infraction that Lehrer committed and the infractions that were leading to retractions, and what it shows very clearly is that I can make a graph in Excel.
I understand that I’m not really comparing likes here. Lehrer is one guy, and the authors of this study are looking trends in biomedical publishing as whole. Nothing that Lehrer has done could be considered an “accident”; his misdeeds seem to be deliberate. However the Lehrer case is analogous to something that is strongly reflected in the study, and that is when an author or a lab is implicated in misconduct, it often leads to fact-checking of their other results and published studies, and that can lead to a whole series of retractions. As a result, a mere 38 labs which had over 5 retractions were responsible for 35% of retractions. Indeed, when you look at the overall rates of retraction in the PubMed database, you get about 8 thousandths of a percent. So biomed publishing as a practice seems to be overwhelmingly clean.
One major obstacle, however, that necessitated this study and gets in the way of assessing the health of biomedical publishing is the way in which retractions are announced. Some journals were announcing that there had been unspecified “errors” and these supposed errors were recorded as such in previous studies of retraction rates. This research team, however, when classifying fraud, followed up on those cases and concluded that “incomplete, uninformative or misleading retraction announcements have led to a previous underestimation of the role of fraud in the ongoing retraciton epidemic.” One example of this, a retraction announcement in Biochemical and Biophysical Research Communications (I actually have their swimsuit issue) stated that “reports were derived from experiments that were flawed in methodological execution and data analysis.” The investigation that Harvard filed with the Office of Research Integrity, however, reported “many instances of data fabrication and falsification were found.” So when the editors of Biochemical and Biophysical Research Communications failed to communicate data fabrication and falsification to their readers, they were…fabricating and falsifying data about biochemical and biophysical research. The authors of this study make the reasonable recommendation that standards for reporting retractions should be established and used. Makes you wonder why they bothered retracting it, since it seems to have been in keeping with editorial policy.
Of special interest to both of our viewers, is the prominence that Andrew Wakefield’s bogus, misleading, horrible chronicle of sociopathy and fraud in the Lancet, at the top of their list for the most cited retracted article. His string of manufactured case studies was cited some 758 times, the highest of any retracted garbage study in PubMed, but very often it was cited as a source of controversy not as if it were in any way useful.
So what is driving scientific misconduct in these few cases? The study raises these questions and floats some hypotheses, including what they call “underlying counterproductive incentives that influence scientists.” They note a type of winner-take all the limited research dollars, jobs and prizes mentality that may be at fault. In Andrew Wakefield’s case, it is clear that he did it because he is worse than all other people. They recommend ethics training and increased editorial vigilance to curb dishonest practices.
The Virtual Skeptics is an independent production of Doubtful News, WhatsTheHarm.net, Skeptical Humanities, and Brian Gregory. Our logo was designed by Sara Mayhew at SaraMayhew.com. Our theme music is by Tremor and is used with permission.