Science is a human activity, and as with any other human activity there is both good science and bad science. There are also good scientists and bad scientists. The need for high integrity in the science system is obvious in that the scientific method builds on peer-reviewed and published results. So when scientific fraud of a gross nature occurs, the science community is perturbed and trust in the whole scientific endeavour is threatened. No branch of science is immune from such fraud and periodically yet another dismal example is exposed.
Currently the scientific literature is full of commentary on a series of fraudulent papers in social psychology produced by a Dutch psychologist. Understandably, there is a lot of concern about such obvious and distressing fabrication. But, as Jennifer Crocker in her commentary in Nature (10 November 2011) points out, we need to reflect not on these dramatic cases but rather much more on the subtle process of how any scientist can start down the slippery track towards such malfeasance.
A number of internet surveys have suggested that transgressions in scientific behaviour are more common than we might care to believe. Integrity is critical to any profession, but there is no absolute way to prevent such goings-on which in all likelihood start with behaviours that we might not readily think of as unacceptable. We need to be sure that all those entering scientific careers understand fully the spectrum of unacceptable behaviours, for example data trimming and data selection or its over-interpretation. We need collective vigilance to maintain the highest standards of behaviour. After all, trust both in published data and in data from our own and our collaborator’s laboratories is essential to the scientific endeavour. Ultimately, there is no way to ensure honest behaviour in science except by watching one’s own backyard.
But beyond issues over the reliability and integrity of data, there are other behaviours we must reflect on. For example, there may be the perverse incentive to ‘adjust’ scientific findings to ‘accommodate’ commercial funding. This has been of particular concern in recent years and is something the public is always suspicious about. Similarly, maintaining the confidentiality of data and its interpretation in review processes is of utmost importance. Linked to this, integrity in peer review processes, whether reviewing submitted papers or sitting on grant assessment committees, is essential. There is also the danger of interpersonal dynamics affecting the objectivity of scientific review. This is a particular challenge in a small scientific community like that in New Zealand where conflicts of interest are common and competition for scarce resources is inevitably very intense. Indeed, every time we name a colleague we do not want to review a grant or a paper, we highlight the complex dynamics and relationships that exist within our scientific networks.