Misconduct in science

 

Sir, – I write in reply to the letters from Rev. Patrick G. Burke and Dr. Cormac O’Raifeartaigh (June 4th) regarding my science column on misconduct in science (“Fraud is now the biggest enemy of science”, June 2nd).

First, let me reassure Rev Burke that science continues to make great progress, despite the problem of misconduct, and the pronouncements of mainline science can be trusted. However, the growing problem of misconduct needs to be tackled urgently.

I am surprised that Dr O’Raifeartaigh seems reluctant to accept that we now have a serious problem with misconduct in science – published results that cannot be replicated by other scientists, selectively picking data to support the hypothesis, inadequate statistical analysis and even outright fabrication of data.

This misconduct problem is now well evidenced and widely acknowledged. For example, in a large study published in PLOS One in May 2009, Daniele Fanelli reported that, whereas only 2 per cent of scientists admitted to outright fabrication of data, 33.7 per cent admitted practices such as dropping data based on a “gut feeling” or selectively reporting data that supported the hypothesis. Tellingly, about 70 per cent of scientists said that they had seen colleagues doing this. I also refer Dr O’Raifeartaigh to an overview of the problem published in the Economist in October 19th, 2013.

Misconduct ranges across the scientific disciplines but seems to be most prevalent in biomedicine. Richard Horton, editor in chief of the prestigious medical journal the Lancet, said in 2015, “The case against science is straightforward: much of the (biomedical) scientific literature, maybe half, may simply be untrue”.

As a science columnist, I must report my subject, warts and all. Happily, I can mostly report wonderful progress, but I can’t sing hallelujahs the whole time.– Yours, etc,

WILLIAM REVILLE,

Emeritus Professor,

School of Biochemistry

and Cell Biology,

University College Cork.