Readers may roll their eyes at the various excuses authors use — including flooded labs and “my laptop was stolen” — when their data are unavailable for further scrutiny following questions. But here’s a case in which a stolen laptop is a real story.
On April 5, Daniel Bolnick, the editor-in-chief of The American Naturalist, posted an expression of concern for three studies published in 2018 and 2019:
This Editorial Expression of Concern serves to notify readers of The American Naturalist that the Editorial Board has identified data archiving and statistical concerns regarding three previously published papers.
The statement continues, noting that author Denon Start — who has had two awards from his PhD rescinded, and whose employment status is unclear — “no longer has access to these data:”
The Board has determined that the Dryad repositories for these papers do not contain all the data necessary to reproduce the reported analyses. The author is unable to correct this because he no longer has access to these data, for valid reasons that the Editorial Board has independently confirmed. It is the policy of this journal that we will publish an Editorial Expression of Concern when we are made aware of deficiencies in the published data repository, which the author(s) are unable to correct in a reasonable time. A committee appointed by the Board has also identified potential errors in reported statistical analyses in Start 2018a and Start 2019 and in graphics in Start 2018a and are awaiting Corrections from the author for those two papers. The Editorial Board emphasizes that authors who actively participate, in good faith, in corrections to their published work should not be penalized or harassed for their engagement in correcting the scientific record.
The papers had drawn comment on PubPeer last August for missing data and “unexpected patterns.” One, titled, “Ontogeny and Consistent Individual Differences Mediate Trophic Interactions,” has been cited 15 times, according to Clarivate Analytics’ Web of Science. It was flagged after Start measured how much insects moved around for an experiment in the paper, repeated the same experiment 10 months later, and obtained very similar data both times. An anonymous PubPeer commenter said that “the raw data deserves scrutiny.” But while most of the paper’s data were stored on Dryad, a data repository, the data points for those experiments were missing.
The other papers, titled, “Keystone Individuals Alter Ecological and Evolutionary Consumer-Resource Dynamics,” and “Individual and Population Differences Shape Species Interactions and Natural Selection,” have been cited 10 and four times, respectively. They, too, were flagged on PubPeer for missing Dryad data on PubPeer (here and here).
Bolnick — who has written about various investigations at the journal and tweeted about this case — told us he began looking into the papers in October. When he did, he learned that the reason Start gave for the missing data — a stolen laptop — checked out.
In February of last year, Michael Wayne Pickering, then 41, pleaded guilty to prowling and stealing from “at least four dozen cars at more than seven different trailheads and parking lots stretching from Mount Rainier National Park to Third Beach on the Pacific coast, to trailheads in the interior of the Olympic Peninsula’s national park and forest,” according to a press release from the U.S. Attorney’s office in the Western District of Washington. Though Start isn’t mentioned in the plea agreement, Bolnick said that the Department of Justice confirmed to him that Start was a victim.
In an email to Retraction Watch, Start apologized for the missing data, and said he takes “full responsibility for data loss–both the main file and its back-up were stolen,” and that he “should have had a third copy.”
During a PhD at the University of Toronto, Start earned two awards from the Canadian Society for Ecology and Evolution (CSEE) for his thesis. Both of them were rescinded last June. CSEE president Stephen Heard declined to comment on the rescinded awards “upon legal advice.”
The awards were withdrawn after CSEE reached out to students at the University of Toronto and asked them to describe their personal and professional experiences with Start. One student claimed that Start did not properly credit others in his papers, according to an individual familiar with the matter who asked not to be named for fear this story could affect future job prospects. A spokesperson for the University of Toronto said they could not answer our questions “due to privacy constraints.”
In an email, Start said he was never given a reason for the awards being rescinded:
To be clear, CSEE never asked for my side of the story, never informed me of their decision until after they had tweeted it out, and never responded to multiple follow-up emails asking for an explanation and/or meeting.
Start — who asked us not to list his current employer — is also a past recipient of the UC Davis Center for Population Biology’s postdoctoral research fellowship. But last October, Jeffrey Ross-Ibarra, a professor at the school, posted a screenshot on Twitter of an email stating that “Denon Start is no longer affiliated with the [Center for Population Biology Program effective September 3, 2020, and has accepted another position.” When we asked Ross-Ibarra questions about the tweet, Ross-Ibarra declined to comment.
As for the three papers in The American Naturalist, Bolnick told us that Start said he “intends to write Corrections as needed,”but that “he has also had nearly a half year at this point to do so and nothing is forthcoming.”
PubPeer commenters have flagged 16 papers by Start. Many of the comments are by the same anonymous sleuth, and usually for concerns about missing data.
In emails, Start said that he takes exception with the anonymous complaints that have been sent to journals:
To be entirely clear: I would have been happy to deal with any comments coming through normal channels (informal emails, non-anonymously through the editor, via the editor where only they know the persons ID) that raise valid criticisms of my work. This has never occurred.
He also said that editors at several journals have asked him to respond to whistleblower concerns about his papers, and that:
In all cases these complaints were either outright rejected or resulted in very small corrections (e.g. at Ecology Letters) that did not in any case qualitatively change any of the results in these papers. This is in contrast to the ~10 complaint documents each of which was ~10-20 pages to which I was asked to respond.
Two other papers by Start received editor’s notes — similar to expressions of concern — last month. Those studies, published in Proceedings of the Royal Society B, were also flagged on PubPeer for missing data.
In January, The American Naturalist amended its policy for critiques, noting that they “expect that criticisms of published papers be written as Comments and submitted to us through Editorial Manager, to be subject to review through our editorial software that appropriately archives all steps in the process.” Those comments could be anonymous, but the journal preferred that people submit criticisms through official channels.
That didn’t happen with these papers, as Bolnick told Retraction Watch:
Some concerns were anonymously posted on PubPeer. Some were informally submitted to me by an individual who I know but who wanted to remain anonymous to the broader community. Some concerns were raised publicly and non-anonymously on Twitter. Some concerns were raised by an Associate Editor. None were brought to my attention through official channels (submission of a Comment through Editorial Manager, which is our preferred means of getting reports of concerns.
Like Retraction Watch? You can make a one-time tax-deductible contribution or a monthly tax-deductible donation to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
The “stolen laptop” reminded me of the situation with Sook Shin at Oklahoma University who lost years worth of research on prostate cancer that had never been backed up anywhere. There was never any real follow-up so I guess it was never retrieved. I was always a bit skeptical of that story, that a researcher would never back up those kind of data and then just leave the laptop in a car while eating lunch at Panera. But, the truth is often stranger than fiction, so who knows.
I cannot speak to scientific data (been lucky with that so far) but I lost the only copy of the last three chapters of a novel I was writing when someone broke into my house and stole my laptop. I had trusted the laptop manufacturer’s claim that you could use the second internal hard drive as backup, but not if the whole machine is gone! So it really can happen. (Gosh, it was painful rewriting those chapters. You are always sure that you had some perfect turn of phrase that you just can’t quite recall.)
Not a stolen laptop story, but …
Back in ’81 I was bringing back to school all of the raw data from mine and from another student’s Master’s thesis (we were both working on different sets of variables from the same complex psychology experiment that we both ran). I had both sets of data in paper folders in my book bag hidden out of view behind the front seat of my 1972 VW Beetle. Before I got to campus I stopped at a corner pizzeria to get a slice of pizza to bring it to school. Well, in less than two minutes that took me to buy my lunch -and in plain daylight- someone broke into my car and stole my book bag (those vent windows in old VWs were very easy to pop open). Fortunately, for the two of us, our theses were optional and we were able to graduate as planned. We also had already generated one conference presentation from the data. But, now all of the underlying data from the study, including a substantial amount that had not been analyzed, were gone and the planned journal article that would have been based on those data was never written. Since then, and after a couple of other losses of data and documents due to computer malfunctions, bad floppies (anyone remember those?), etc., I compulsively make multiple copies of everything!!
Good learning lessons! I routinely email to myself the drafts of papers as backups… Also, I email from my work email address to my personal email address (or vice versa), so it is backed up in both an inbox and a separate outbox. “Belts and suspenders” approach.
Dropbox, Google Drive, OneDrive… God created the cloud for a reason.