Fudging the answers is as old as science itself. Not a few experiments which produced classic breakthroughs have proved strangely difficult to reproduce. Fortunately they came up with the 'right' results, so no more need be said. Analysts disagree about how frequently it occurs now.

Some would argue that it's rare because self-correcting mechanisms, like the anonymous peer-review process, discourage cooking the results. Others claim that recently publicized cases of scientific misconduct are only the tip of the iceberg. The fact is that we don't know how widespread scientific misconduct is.

Readers of Voodoo Science by Robert Park may not be optimistic. Dr Robert Park is professor of physics at the University of Maryland. He also heads the Washington office of the American Physical Society. His views on science are thus of public interest. In his book, Robert Park attacks what he sees as a tendency towards 'junk' science. Not only is this prevalent in the media, in advertising and commerce, but also within medicine and science. A recent string of revelations suggests that the junk content in mainstream science is significant.

It would be surprising if scientific cooking of the books did not occur. Scientists are human, and humans under pressure are prone to adopt underhand methods. Is there any reason to suppose that scientists are less human than accountants?

University laboratories may appear tranquil on the surface. There has always been the academic imperative to be first to publish a new discovery, but modern universities are exposed to additional pressures. Sponsors, whether industrial or governmental must be fed with results, otherwise the funds may dry up. University Legal Departments are on the watch for lucrative patent applications. Pressure from this source contributed to premature disclosure of the original Cold Fusion research by Fleischmann and Pons, with damaging consequences.

Prominent scientists may feel that their reputations enable them to cut corners and get away with it. Who will be foolhardy enough to question their results? Recently, it has come to that.

In 1999 the Lawrence Berkeley Laboratory at Berkeley California announced two new chemical elements following work on their 88inch Cyclotron by Victor Ninov and a team of experimenters. A paper published that year in Physical Review Letter described the discovery of elements 118 and 116. The announcement was described by US Energy Secretary Bill Richardson as "a stunning discovery". Equally stunning was the retraction of the paper less than two years later, followed by the dismissal of Ninov from the Berkeley Laboratory in May 2002. No misconduct was attributed to Ninov's co-workers but questions were also raised over earlier work by the same scientist at the German Research Institute at Darmstadt.

In industrial science, technology and medicine the pressures are still greater, as are the potential rewards. The consequences can be still more serious.

In October 2001, the US Office of Research Integrity announced that Dr Steven Arnold, Tulane University, had '…admitted to scientific misconduct and conceded that there were no original data or corroborating evidence to support the conclusions reported in the Science paper'. ORI's investigation had focussed on a paper published by Arnold and others in Science July 25th 1997, subsequently retracted. The paper claimed that certain environmental chemicals showed a huge increase in activity as oestrogen mimics when tested in combination. The work had seriously misled the US Environmental Protection Agency in formulating policy on endocrine disrupters.

In September 2001, following a series of papers on high temperature superconductivity of fullerene compounds, the Bell Laboratories (NJ - USA) were widely tipped for a seventh Nobel Prize. In May 2002 the American Association for the Advancement of Science announced evidence of data manipulation in five separate papers. 'The groundbreaking papers include a Bell Labs physicist as lead author… Schon is the only researcher who co-authored all five papers in question.' Inquiries continue.

Action is in hand to improve ethical standards in scientific and medical research. It remains to be seen how effective this will be. Meanwhile the reputation of 'pure' research looks a trifle tarnished.

References

Robert L. Park Voodoo Science - The road from foolishness to fraud Oxford University Press ISBN: 0-19-850745-3