Retraction Watch - Scientists behaving badly

05/21/2015 10:55

by Ivan Oransky


Scientists behaving badly: Andrew Gelman explains why we should care. []


Peer review is under attack, reports Tom Spears of The Ottawa Citizen, who has done his own sting of predatory journals.


Should there be standard protocols for how researchers attempt to reproduce the work of others? Kerry Grens explores the question. Journals including Science and Nature have joined forces to create guidelines designed to improve reproducibility.


Rolf Degen has an update on a case of suspected plagiarism by six professors of sports medicine in Germany.

Nanotechnology is making double-blind peer review an option. Our decision to offer double-blind has been driven by concerns from sections of our community that biases, such as those against female authors or researchers based at less prestigious labs and institutions, could play a role in the review process.

Who needs science when we have anecdotes?

A married Yale cardiology researcher smitten with a junior scientist told her that she was choosing the wrong man since he was in a position to 'open the world of science' to her, The New York Times reports.


A Northwestern researcher has agreed to pay $475,000 to settle claims of grant mismanagement. The university had already paid nearly $3 million over related claims.


A concert pianist had demanded that a review of a 2010 concert he gave be removed from internet search results under the European ˜right to be forgotten' law. Could scientists found guilty of misconduct be next?


Inside Higher Ed covers the Fazlul Sarkar lawsuit.

Reuters has done away with comments on its news stories.


Sorry if I seemed to be legally liable in any way: Sorry Watch on an apology stemming from a request to retract a Washington Post column.


A Harvard vice provost authorized a study that photographed faculty and students in class without notice, The Crimson reports.


Star Trek medicine: A scientist's apology for basic research.



[Note: Genuine research scientists -- whether doing basic or therapeutic RESEARCH -- don't falsify or modify their data, especially since negative data is often correct and allows the scientist to rethink his/her theory properly in order to determine what is, in fact, fact. They LIKE negative data as it eventually helps them decipher the truth! Genuine research scientists also do not design or execute their experiments in order to advance politically or serve some private master. Genuine scientists also follow the traditional Scientific Method (which is often mis-defined these days) -- which has FOUR steps, the last one being the requirement that one's data from one specific experiment be REPLICATED, DUPLICATED by INDEPENDENT laboratories around the world before they are considered scientific FACT. Otherwise, one's data remains as mere unevaluated speculation. Why? Because quite often there are artifacts (intended or not) in the design and execution of one's experiment that cause fake untrue data to be produced (e.g., there might have been impurities in the chemicals from the manufacturer used, or the assays used might have been incorrectly calibrated, or the technician might have rushed the experiment to get home early, etc.). The requirement for replication/duplication assures that such artifacts are caught so that all the researchers can modify their theories appropriately. A mere 'proof of principle' derived from just one or a few experiments just doesn't make it! It's also in the public interest to determine if there are artifacts since without such verification such false data can unfortunately then be used by other researchers who assume it is correct, to create false drug models, as well as be applied to innocent humans in clinical trials (which clinical trial data itself would be rendered false) and human patients who would be seriously harmed! It is long past time that disingenuous scientists be held legally liable for proven misconduct -- rather than just a slap on the wrist. The article first appeared here -- DNI]