Recently, a biostatistician sent an open letter to editors of 10 major science journals, urging them to pay more attention to common statistical problems with papers. Specifically, Romain-Daniel Gosselin, Founder and CEO of Biotelligences, which trains researchers in biostatistics, counted how many of 10 recent papers in each of the 10 journals contained two common problems: omitting the sample size used in experiments, as well as the tests used as part of the statistical analyses. (Short answer: Too many.) Below, we have reproduced his letter.
Consider this: Fragments of a PLOS ONE paper overlap with pieces of other publications. The authors used them without credit and without quotation marks.
This sounds an awful lot like plagiarism — using PLOS‘s own standards, even. But the journal isn’t calling it plagiarism. They’ve labeled this an instance of “text overlap,” a spokesperson told us, based on the amount of material that the paper shares with others.
The last author — Carlo Croce, who has two retractions under his belt — denies that he plagiarized, and says that his university has cleared him of a plagiarism charge from an anonymous whistleblower.
PLOS fixed this case last year with a correction notice — not the common course of action for a case of confirmed plagiarism. Take a look at the notice for yourself:
A Cancer Cell paper that caused a flurry of activity on the website PubPeer is under investigation, after the last author announced on the site that he’d requested a correction from the journal.
The 2012 paper sparked a lively dialogue last month on the post-publication discussion site, as commenters questioned Western blot images in which some bands appeared to be duplicates. The last author responded, noting he had alerted the journal to a “mishap,” and a correction would be forthcoming. However, some commenters remained unsatisfied, and questioned why the correction was taking so long to appear, as well as the explanation for what went wrong.
A 2014 Cancer Cell paper became the subject of an erratum in January 2015, shortly after PubPeer members began criticizing the data. However, many issues brought up by commenters weren’t addressed in the correction notice, including a figure that might be two experiments spliced together to look like one.
The paper, led by Guido Franzoso at Imperial College London, claims that a new cancer drug called DTP3 kills myeloma cells “without causing any toxic side effects,” according to a press release from the school. Guido Franzoso is the founder of Kesios Therapeutics, a drug company which is set to begin clinical trials on DTP3.
The correction indicates that Western blots were cropped badly, which omitted several panels discussed in the text, while an “extra time point” was included accidentally. An antibody was also omitted from the description of the procedure.
PubPeer commenters have noticed additional issues, such as a criticism of figure 3D, which were not included or changed in this correction.
Here’s the correction for “Species- and Cell Type-Specific Requirements for Cellular Transformation:”
We were apprised recently of errors made in the assembly of Figures 2B, 3A, 4A, 4B, and 5G, resulting in the incorporation of incorrect representative images in these figures. These errors occurred during the electronic assembly and have no bearing on the conclusions of the study. The corrected figures are shown below. The authors apologize for any possible confusion this might have caused.