As Chinese research expands, who is looking out for faked results?
The investigation of research misconduct is always fraught with difficulty, even if the necessary protocols and experienced expert committees are fully in place. In China, they are not. If the nation is to get to grips with the problem of misconduct as it becomes a substantial scientific power, that situation has to change.
Chinese research agencies do have structures for investigating misconduct allegations, but in the absence of open discussion and independent press scrutiny, few researchers have much faith in them. The rapid and open exchange of information over the Internet has some potential to fill the void, but it also carries risks (see Nature 441, 392–393; 2006). It could readily break down into a dangerous game of unregulated accusation and counter-accusation, shedding no light on actual misconduct.
The power of the Internet in identifying scientific fraud was amply demonstrated last year in the case of Woo Suk Hwang, the discredited South Korean cloning researcher. Online portals discussed suspicious images and data in Hwang's papers, ultimately leading Seoul National University to pursue an investigation that exposed Hwang's fabrications. And Internet postings of allegations that Jin Chen faked digital-processing chips contributed to his dismissal from Shanghai Jiaotong University last month.
The Internet can play a particularly important role in countries such as China and South Korea that do not have adequate systems for investigating misconduct allegations. That isn't to say that countries with systems in place are totally on top of the problem, but at least they have developed some of the institutions and protocols needed to handle it.
Organizations charged with assessing allegations of scientific misconduct do exist in China, and on paper the system appears functional — but there is no evidence that it really works. China lacks an independent press to report on such matters. The very size of the country and subsequent disparate implementation of policies set in Beijing make matters worse.
In addition, the cultural importance of ‘saving face’ in Chinese society makes the full-frontal public attacks that tend to characterize Western misconduct allegations almost unthinkable. There are no effective provisions to protect whistleblowers, so it is hard to believe that anyone who observes misconduct would summon the courage to report it to the authorities.
It is in this climate that New Threads, a Chinese-language Internet site run by a single researcher based in San Diego, has come to play a significant role in the monitoring of scientific conduct. This arrangement is deeply problematic, however.
In China's recent history, ‘bottom up’ accusations have often been abused by the authorities to persecute perceived enemies of the state. This was especially true during the Cultural Revolution, when simply pasting a poster on the wall calling someone a ‘bourgeois’ could destroy their livelihood. The threat of innocent people being branded as ‘pseudoscientists’, either by a jealous rival or by the state, further clouds the misconduct picture in China.
The only real solution to this problem is a great deal more complex than hooking up to an Internet connection. It requires the establishment of independent offices in Chinese research agencies, rather like the inspector general's office at the US National Science Foundation, or the Office of Research Integrity at the US health department. The system can only operate effectively if it offers protection to whistleblowers. It also requires a new generation of scientists to be educated in what constitutes proper scientific conduct. And it needs to ensure that investigations give anyone accused the opportunity to demonstrate their innocence.
China is struggling to come to terms with these kinds of requirements in society at large, as well as within the scientific community. For a multiplicity of reasons — of which the desire for scientific progress is just one — addressing them ought to be the government's greatest priority.