By Melissa Jackson
BBC News education reporter
One in 10 research scientists is under pressure to tailor findings to suit the work's sponsor, a UK survey suggests.
Women are under greatest pressure
Women are more likely to be targeted than men, according to the poll of 358 scientists carried out by two unions.
Unions say the findings were "extremely worrying" and called for research to be properly financed, and for an end to fixed-term contracts for scientists.
The Royal Society, the UK's national academy of science, is drawing up guidelines to combat the problem.
More than 10% of scientists have been asked by their commercial backer to tailor their research conclusions to meet the sponsor's requirements, according to the survey of university and government laboratories.
Research, carried out jointly by the Association of University Teachers (AUT) and the public service union Prospect, found that women were under even greater pressure.
However, most (84.5%) of the 358 respondents (58% male and 38% female) said they had never been asked by a sponsor to skew their research.
A total of 7.9% of those who took part in the online poll said they had been asked in general terms to tailor their conclusions to the funder's preferred outcome.
A further 1.2% of the total said they were asked to tailor their results so that they might obtain further contracts, and another 1.7% said they had been discouraged from publishing their findings by their backer.
When the figures were broken down, 11.5% of women (compared with 6% of men) said they had been asked to tailor conclusions to suit their sponsor's preferred outcome; 1.5% of women (compared with 1% of men) were asked to do so to obtain further contracts and 2.3% of women (compared with 1.5% of men) had been discouraged from publishing their findings.
Prospect's head of research and specialist services, Sue Ferns, says the findings reinforced union concerns.
"Given that all the survey's respondents considered that their key role was to provide impartial and objective advice, any evidence to suggest some members feel under pressure to modify their results is extremely worrying.
"Prospect has been arguing for some years that the contract culture is a real barrier to developing a long-term strategic approach to science, and it is disappointing that our warnings over the dangers of commercialisation and loss of independence are still going unheeded in some quarters.
"Any request to falsify results brings science into disrepute, threatens the integrity of scientific advice to government and damages public trust in government itself.
"Science, above all else, is about a pursuit for the truth."
An AUT spokesman said: "These findings are worrying and indicate a possible problem when research projects involve some commercial money.
"The fact that many researchers are also on fixed-term contracts and whose continued employment also relies on the funding of the research is not good for those staff, or for the long-term future of British research.
"One message we think government and employers should take from this is to end the practice of fixed-term contracts and properly finance research."
The Royal Society is equally concerned about the survey results.
Sir Patrick Bateson, chair of the Royal Society working group on best practice in communicating research results, said: "It is clear that some researchers are influenced by their affiliations, be they to funders, sponsors or employers, when carrying out or reporting their work.
"In many cases these biases are introduced unknowingly, but can be avoided if researchers become more aware of the potential problems.
"There are also occasions when biases, for instance on the selection of evidence, are deliberate, and such practices are clearly undesirable.
"The Royal Society will shortly be publishing recommendations to overcome some of the problems of affiliation bias when research results are communicated to the public."
The survey looked at other issues relating to scientists' work, including job satisfaction and volume of work.