Ensuring science is conducted with integrity requires a supportive culture, not draconian rules and sanctions.
Last year, in a much-publicised case, two researchers at Jinggangshan University in Southern China were dismissed after 70 scientific papers they had published in an international journal were found to contain fabricated data.
The university blamed the falsification on the researchers' "lack of moral integrity". But critics also pointed fingers at the intense pressure put on Chinese scientists to compete with other researchers and raise their university's status.
For example, Jinggangshan University was reported to have offered prizes of 5,000 Chinese yuan (US$733) for any scientist publishing in a recognised international scientific journal. Other universities are said to have offered twice this figure.
In this, the developing world is not alone. Many scientists in developed countries struggle with a 'publish or perish' system and its associated pursuit of international rankings.
Sensitive to criticism of its strategy, Jinggangshan University has now stopped financial rewards for scientific publications. The move is welcome recognition that, although competition remains important in science, encouraging it can backfire if personal gain and institutional status is put before scientific endeavour.
Participants in the Second World Conference on Research Integrity in Singapore last month recognised the twin dangers: of scientific fraud and misbehaviour on the one hand, and of addressing the symptoms rather than the causes on the other.
The conference acknowledged such fraud is on the rise globally, and endorsed efforts to compile a "guide for professionally responsible research practices" that is due to be published shortly in the Singapore Statement on Research Integrity.
This is welcome international recognition that scientific misconduct is increasingly widespread. Whatever the rigours of peer review, the scientific system still relies on trust. Where this trust is abused, the scientific process itself becomes corrupted.
Equally welcome is the meeting's recognition that heavy-handed policing is not the solution. Science flourishes under recognised codes of conduct, but not under draconian rules and regulations, which may, for example, constrain informed speculation and intellectual risk-taking.
Sensibly, the draft Singapore statement focuses on principles, rather than methods of enforcing them.
Ensuring public trust
Ensuring that research is honestly conducted — and reported — is important for the integrity and robustness of science itself. But it is equally essential if science is to generate and maintain public support. This means that any code of conduct must include openness and transparency.
Recent disputes over the validity of scientific data used to support calls for climate change action shows this most plainly. To the relief of the scientists involved, various enquiries have cleared them of charges of deliberate falsification, and found no evidence of scientific misconduct.
But the lack of transparency from researchers at the UK's University of East Anglia, whose emails disclosed a reluctance to share data with critics, has damaged public trust. This in turn has fanned scepticism, and weakened support for urgent action on climate change.
Science must not only operate with integrity; it must also be seen to operate with integrity. This is an increasing challenge in a world where electronic communication records introduce a new dimension to public scrutiny. And it is one that needs reflecting in any code of ethics.
Culture of integrity
None of this will be achieved by harsh rules and punishments (even though sanctions remain important for significant transgressions). Rather, we must develop a culture of good practice, including open communication, which ensures scientific knowledge remains robust, and sensitive to public perceptions.
We must start to build this culture of integrity in the educational system, particularly in universities. Future scientists need to learn that success depends not only on what you achieve, but how you achieve it.
And the broader scientific community must also develop mechanisms for building and sustaining this awareness. Here, responsibility rests on individual scientists, the institutions that they work in, and those that sponsor their research.
If the proposed Singapore Statement on Research Integrity can support good practice at all three levels, it will become a significant document.