Monday, November 9, 2009


Are scientists honest? A blind study

J.L. López-Vázquez, J. Conrado and J.L. García-Berlanga
Plácido Foundation, Navalmoral de la Mata, Spain.

Scientists have long suspected that many colleagues do not faithfully reflect their experimental methods in published articles, omitting important issues for reproducibility or, worse, deliberately hiding ethically unacceptable dirty tricks that were used through their research aimed to enhance its impact. In order to find whether such practices were common among researchers, we induced a cataleptic pseudo-hypnotic state in the minds of graduate students, postdoctoral fellows and senior investigators from several public institutions in Japan, the United States, Germany, UK, Israel and Spain. A cohort of 333 individuals were examined. Catalepsy was induced by moderate alcohol ingestion (2% v/v in bloodstream, Garrafón Enterprises, Cazalla de la Sierra, Spain), monitored drug administration (700 mg intravenous barbiturates, Shootes-Barbate Pharma, Cadiz, Spain), and visioning of a video displaying a routine session at the European Parliament (3 min). Confessions were recorded with confidentiality so that they could not be used against the individuals. 67.3% of graduate students claimed to have teased on results and cheated on their supervisors. Irreproducibility of expected results was the main cause for this behaviour. Being distracted by sex-appeal of opposite sex labmates was a common trigger for experimental failure. In other cases, proper incubation times were neglected as researchers were entertained chatting at the computer. Fake results were most commonly elaborated by using Adobe Photoshop manipulation of blots and micrographs. One individual regretted having cooked with rice the rabbits he was using to produce polyclonal antibodies before bleeding them. Other, frustrated by not being sure of having mistaken samples #36425 and #36426 in a high-throughput screen, trashed the whole experiment and made up all the dataset out of his imagination. Postdoctoral fellows cheated less (34.5%), but often redacted incomplete or misleading experimental details on the Materials and Methods section of their scientific reports intentionally. A Japanese postdoc attempted to flood their competitors’ labs by changing the real scale of his experiments from microliters to liters. A researcher from Israel only used Kosher reagents to hamper reproducibility of his experiments in foreign labs. Other case dealt with adding irrelevant obnoxious components to buffers, like mercapthanes, just to cause disgusting odours in the competitors’ labs. A remarkable confession revealed that Bioinformaticians mostly drive conclusions from their experiments by discarding all the data that they do not understand, which amounts to 99.6% of the output of their research. Surprisingly, a high percentage of senior researchers made use of voodoo dollies to appease or punish referees and journal editors. Some wet their pants when they receive an e-mail from the journal. A widespread unethical practice was to lock postdoctoral fellows at the lab with a limited supply of food (see figure) when the journal asked for experiments to be performed in a limited time.

We found three cases of fake submissions, e. g., the principal investigator pretended to have submitted a paper and he made the negative comments of two false reviewers himself in order to manipulate the feelings of his fellow researchers in a particular direction. The good news is that about 10% of researchers had never failed to produce honest rigorous and exemplary Science. However, we cannot be sure of this result. For a start, you would have to trust our study.

1 comment:

  1. The figure displayed in your paper clearly reflects the uncomfortability I often feel when I hear somebody in the lab saying "Estaré plaqueando, en la cabina". The similarities between modern lab research and classical short films are quite creepy.