Interesting talk covers fraud in science issue

Morpeth Rotary Club's Rhona Dunn gives the vote of thanks after the talk by President Andrew Hamnett, right.
Morpeth Rotary Club's Rhona Dunn gives the vote of thanks after the talk by President Andrew Hamnett, right.

Club President Andrew Hamnett, a retired university vice-chancellor, welcomed an official visit by the Rotary Club of Longbenton with Killingworth before giving a talk on academic fraud in science.

He said the main aspects could be grouped under the headings of fabrication, falsification and plagiarism.

Fabrication is where all of the findings are made up, falsification is where the data is changed or doctored to make it look better and plagiarism is passing off the work of others as your own.

An obvious question is: Why would anyone try fraud in science as the chances of detection must be high?

Information is published and any part of it can be checked and verified.

Funding bodies look carefully at proposals and scientific credibility. The research then goes to a journal and is assessed by the editor and specialist journalists.

The test is verification by the scientific community – for example, are the results repeatable?

The problem should be self-correcting but it is not.

Two components work against each other, pursuit of the truth and pursuit of recognition as a scientist.

They may omit data that does not fit or just tell lies when they are defending a theory and seeking recognition.

There may be fraud with the intention of furthering their career.

Up to the time of World War II it was still possible to do good research without much funding, but that is no longer possible. The researcher tends to join a group led by a famous guru who has attracted research funding.

Anything original that they discover is ascribed to the great guru who gets the praise and the awards. It is possible to use the guru to get a better post in the system, but if you want the guru’s good will, it is not a good idea to find results that oppose his favourite theory.

Discovery of fraud does not happen much. It is unusual to say that a particular method does not work.

The jungle telegraph plays a part. Scientists tend to go to conferences a lot, which is where they find out what works and what doesn’t.

Word gets around and the new method dies without doing much damage.

If a world-shaking report appears, all would rush to verify it and fraud would soon be found out.

This was the case in the ‘breakthrough’ of using ‘cold nuclear fusion’ to generate electricity.

A more careful scientist using more complex experiments found it did not work.

Damage was done to the original scientist and to the credibility of science as a whole.

A survey done on research findings in a medical journal found that two thirds of the results could not be reproduced.

Authors generally chose the best results, not the typical results.

One claimed a yield of 80 per cent on a useful product when the actual yields were 20 per cent to 50 per cent.

The total percentage of crooked scientists has not really changed in the last 50 years, but there has been a dramatic increase in the number of scientists and therefore also in the number of frauds.

A cancer researcher got wonderful results on skin transplants with no rejections, but no-one else could reproduce this, including himself. He fabricated the data to show the result he wanted.

A scientist in the USSR during a decline in agriculture in the 1920s and 30s showed greatly increased wheat yields using a discredited genetic theory and rose to the top of Soviet agricultural science.

He doctored all of his data to fit in with what the political masters wanted.

Another got remarkable results with electricity conducting polymers. His method was to listen in to management conversations about what they wished would happen and fabricate his results to fit.

A local scientist was known to be especially inept and had a reputation for wrecking equipment. He was fizzing with ideas that did not work and trimmed all results that were too high or too low.

He had his data eagerly accepted by his manager, who was also discredited when it all came out.

The famous Gregor Mendel had his results on the genetics of peas re-examined by a statistician in the 1930s. It was found that he had trimmed his results to produce data which was far too good – he had picked out observations that aligned with his expectations and ignored the rest.

Plagiarism is very common. In a race to get research findings published before that of a rival, a scientist rushed out the work of a subordinate who recognised whole sections of a paper she had originally written.

The paper did not have the names of the hospitals and doctors where research work was supposed to have been done and patients were quoted as having recovered from illnesses when they had not.

It was later found that every piece of data had been fabricated.

Whistle-blowers are often badly affected. The subordinate lost a job and career.

A vote of thanks was proposed by Rhona Dunn and this was followed by a final toast to the Rotary Club of Longbenton with Killingworth and Rotary the world over.