Cooked data, faked science

The scientific-academic world may seem far removed from corner cutting, unscrupulous markets, but impressions can be misleading. Thanks to the compulsion to ‘publish or perish’, some scientists are not just taking recourse to plagiarism and stealing other’s ideas, but are brazenly falsifying and fabricating data. One such recent incident has caused red faces in the Chandigarh-based Institute of Microbial Technology (IMTECH), a flagship lab under the CSIR. A senior scientist there has not only been summarily sacked, even his pension benefits have been reportedly refused, which is pretty harsh by government service standards in this country. But then, the scientist in question and his doctoral student have been charged with serious scientific misconduct; the data in the seven papers they co-authored in a US-based scientific jourl are all said to have been faked. Three papers have already been retracted, and the other four will soon meet the same fate — which is said to be a record of sorts. The scandal has left the IMTECH authority badly shaken, as it is a highly regarded lab for works like India’s first patented drug to dissolve blood clots. It is being seen as a blot on the entire Indian scientific community, with IMTECH director Girish Sahni remarking in anguish: “What happened is symptomatic of the way our science will progress if number of papers published is used as the yardstick for rewards and promotions.”

The fraud was spotted by a US-based scientist in whose lab the IMTECH doctoral scholar had pursued post-doctoral research. He wrote a complaint to IMTECH that the offending papers lifted ideas and work done at his lab. The IMTECH authority instituted a probe and discovered to their horror that the entire data published by the team had been cooked up. “There are no data available underlying this study and the published results are fabricated,” the investigation concluded. The matter even reached Parliament last year with Lok Sabha members querying the Minister of Science and Technology about the allegations. But the reaction of the editor of the prestigious, peer reviewed jourl PLOS ONE, where the papers were published from 2013 onwards, is telling of the rot setting in scientific publishing. Pointing out that the three retracted papers were handled by three different editors ‘which may have made it more difficult to catch any fraud’, the editor and eminent Dutch microbiologist Willem van Schaik admitted that neither he nor the reviewers ‘suspected a thing’. Despite considerable editorial effort and prolonged peer review, and even with hindsight now, Schaik laments that it is ‘difficult to find which data have been fabricated’. It has been established in the enquiry that the doctoral student faked the data, but his supervisor is also being held responsible for failing to cross check the data in his capacity as team leader and senior scientist.     

Of late, there has been much soul-searching in scientific circles over which way research and publishing is going. Questions are being asked as to who should be blamed ‘when fake science gets published’. The sad truth is that a clever, if unethical and misguided, research student can easily take the supervisor for a ride. And since the supervisor, being a senior scientist, rarely gets into the nitty-gritty of the research work — he or she can be easily deceived, or what is worse, chooses to ask no searching questions. If the research supervisor cannot catch any hanky-panky, chances are that a reviewer will similarly fail. As for research jourls, there is a major flaw which several experts have noted. Earlier, jourls gave space to reports about failing to replicate some result or problems with some study or experiment, which helped in solid, if incremental, progress in a particular research. But over the years, such jourls too are being accused of falling to the lure of catchy headlines announcing some breakthrough or the other. An eye to the market may bring forth a lurid tabloid headline, but should the same motive apply to a serious scientific jourl? The problem of detecting fake science has reached such a pass that a Stanford varsity study last year even devised an ‘obfuscation index’ that can help catch fabricated scientific research before it is published. The researchers found that scholars faking data tend to use more obfuscatory language and jargon; since they know their conduct to be unethical, they avoid first person references and positive terms for their data — showing a disinclition to attract too much notice! Interestingly, Pew Research Center, one of the major funders of high profile surveys, faced some discomfiture recently when two of its researchers devised a statistical test to show that one in every five surveys is full of concocted data. What is true of scientific research is also true of studies in the arts and humanities. Worried experts are calling for ‘better scholarly communication and academic values’ to check the toxic culture of exaggeration, plagiarism and falsification in research.