Academics in China, India and other developing nations are more at risk of their researchers falsifying or fabricating data, according to a new study.

The paper found the risk of data falsification or fabrication was ¡°significantly higher¡± in China, India, Argentina and other developing countries, compared with countries such as the United States, Germany and Australia.

The , published on?preprint service bioRxiv and authored by researchers at Stanford University, Leiden University, the University of Washington and Johns Hopkins University, looked at 346 papers?containing ¡°problematic image duplications¡± ¨C from a sample of more than 8,100 published in the journal Plos One. The study?tested four hypotheses commonly associated with data fabrication: pressures to publish, researchers being unrestrained by forms of social control such as peer scrutiny, working in countries lacking scientific misconduct policies, and researchers being male.

These image duplications were subsequently placed into three categories according to whether they reflected ¡°unintentional error, questionable practice or outright scientific misconduct¡±.

³ÉÈËVRÊÓƵ

While the researchers were not able to determine whether the image duplications analysed in the study were produced intentionally or by mistake, many of the variables they tested ¨C in association with the hypotheses ¨C were ¡°significant predictors¡± of questionable practice or outright scientific misconduct.

In relation to the ¡°pressure to publish¡± hypothesis, the research finds that ¡°early-career researchers, and researchers working in countries where publications are rewarded with cash incentives were at higher risk of image duplication¡±.

³ÉÈËVRÊÓƵ

The correlation with cash incentives, the paper states, may not?necessarily be ¡°taken to imply that such incentives were directly involved in the problematic image duplications¡±, but such incentives ¡°may reflect the value system in certain research communities that might incentivise questionable research practices¡±.

Instances when researchers were mentored, had their work scrutinised by their peers, or were working in larger research teams, are ¡°less likely to engage in misconduct¡±, the authors say. Countries with national and legally enforceable policies against scientific misconduct were ¡°significantly less likely to produce image duplications¡±, they add in the paper, titled ¡°Why do scientists fabricate and falsify data? A matched-control analysis of papers containing problematic image duplications¡±.

Daniele Fanelli, co-author of the paper and senior research scientist at Stanford¡¯s Meta-Research Innovation Center, told ³ÉÈËVRÊÓƵ that of the hypotheses supported, cases of either unintentional or deliberate academic misconduct were more likely to be found in countries that had ¡°less of a history of scientific research compared to more industrialised countries¡±, and that possibly those more industrialised countries ¡°put in place high-level and very strongly defined [misconduct] policies¡±. He added that it was the global academic community¡¯s ¡°duty¡± to ¡°do what we can to help these countries catch up¡±.

¡°Researchers working in China, India and other developing countries run a higher risk of errors and/or misconduct in their work,¡± he said. ¡°Some of these are of great importance in terms of the volume of research they produce, so¡­the international research integrity agenda should consider it a priority to ensuring higher standards of research practice in these countries.¡±

³ÉÈËVRÊÓƵ

Asked why researchers carried out misconduct, Dr Fanelli said the ¡°bottom line is they can get away with it¡±, when they¡¯re operating ¡°in an environment that hands out cash bonuses [like China] and arguably sends the wrong signals¡± or if ¡°they lack research integrity as individuals¡±.

john.elmes@timeshighereducation.com