Honesty for Sciences 1.01

Honesty for Sciences 1.01

Scientists practice the integrity of research in the first semester of the university as much as towards the end of the career. Despite all the regulatory efforts, we end up deciding for ourselves how much we want to follow the rules, balancing reward and punishment. This became very clear to me as an overachieving postdoc in the adjacent lab at my alma mater lost her career overnight for falsifying data. And in our time when ignorance and scientific fraud strive to finish off our planet, it is imperative to cultivate research integrity.

Responsible research is vital for various reasons. For researchers, it is an important way to maintain the ethical balance which puts the interest of our society over the personal gain. It also provides a peaceful mind that is clean of guilt and the fear of being caught. On the level of the society, the integrity of researchers helps to make the research reproducible and to correct mistakes. And in a responsible community it helps to strengthen the links of trust.

To show integrity in the research means to report true findings. But also to share the raw data so that our peers can walk through the process themselves. It is also imperative to not conceal important information or under-represent the negative outcomes. And the researchers ought to correct or retract erroneous research. Simple? Yes, and a lot of researchers honour the code. But as often the case with honour codes, some do not.

The spectrum of ethical breaches is very broad and the severity varies with cases. A widespread problem in sciences, for instance, is that up to 70 % of studies cannot be reproduced. While the reasons are many, one of the main is the biased editing of the manuscripts. It omits negative findings and over-represents the positive ones. This, by now, commonplace bias looks harmless on a small scale. But on the level of national and international research, it wreaks chaos and sucks up billions of euros as the problem grows.

One step further are studies which turn out to be wrong and have to be retracted. Sometimes these mean to cheat the taxpayers and the community of peers. Just consider the infamous champion Yoshitaka Fujii, who fabricated at least 183 fraudulent papers and still holds the title for the most falsified publications between 1980 and 2011.

Yoshitaka is not alone and there are many who forge their data or results to secure the next grant. This gives the rest of researchers a bad name and, ironically enough, pointing out these frauds is the first line of defence for pseudo-scientists who claim to uncover the conspiracies of the “mainstream science”.

Sometimes when a paper is retracted, it is bad luck though (or rather a lack of data combined with the need to publish fast). Google Flu started 2008 and was very accurate predicting flu outbreaks with machine learning methods. These were written up in a striking Nature publication in 2009. Only 2013 the bubble has burst as the algorithm missed a seasonal flu outbreak in a spectacular fashion. Hat off for the Google scientists – they did not deny the mistake. They retracted the paper and capitalized on promoting Google Flu as a monument to Big Data pitfalls.

So how do you enforce responsible research? The system working so far was the peer review. The peers trying to understand and reproduce studies often are the ones exposing the breaches of the honour code. Plus the presence of someone looking over our shoulder gives us the push to publish honestly.

Ultimately, the media are a huge integrity booster. Sometimes media tend to spread hoaxes faster and better than any fraudster intended. But then, the media expose the research to a huge audience and it lands in the lap of those who can uncover the inconsistencies. There are even some indications that this effect is responsible for the lower reproducibility of studies that were covered on TV compared to the not covered studies.

More importantly, it pays to cultivate honesty in research institutions. The institutions also help to create an ethical culture in their halls. Often bad science emerges where the establishment looks the other way. And the way that managers and scientific directors treat research ethics is decisive for the careers of their employees. These tend to follow the leader’s example out of personal gain or respect.