What if science is made up

Loss of confidence : How science harms itself

Is the success story of science endangered in today's world by “own goals”? This question concerns us all. Science is a determining factor in all areas of life worldwide. It permeates the private sphere, the profession, determines state and economic action, the entire society.

All of the major topics that concern us in personal conversations and the public on a daily basis can largely be traced back to the successes of science, whether health, longer lifetimes and demographic change, the digital revolution from the Internet to mobile phones to artificial intelligence or questions of globalization and Sustainability.

However, the opportunities that lie in these developments are only used and the dangers can only be avoided if we also exhaust the possibilities that science currently offers us. And further developments in science are coming at us with giant strides. To give just one example: genetic engineering and stem cell research will certainly enable us in the foreseeable future to be able to change ourselves as living beings in a targeted manner. One can therefore rightly speak of the fact that people have moved from an object to a subject and become a self-designing project. That will be a big ethical challenge.

Findings are being questioned - for example by Trump

Against this background, a central question is whether trust in knowledge gained through science threatens to be lost and to what extent science itself is to blame. Today, scientific findings are increasingly being called into question. That ranges from everyday life to big politics. Children are being vaccinated less and less, fake news is widespread and the most powerful politician in the world, the American President, questions scientific knowledge about climate change caused by human activity and makes this the basis of his politics. How did that happen? Does science no longer lead to the right knowledge? Or is the loss of credibility itself to blame, so are science own goals?

When it comes to bad scientific practice, science itself is very briefly concerned only with plagiarism or incorrect authorship. This is borne out by the inquiries addressed to the ombudsman for science of the German Research Foundation (DFG). According to the 2016 annual report, there were a total of 87 inquiries, 21 percent of which were authorship questions and 11 percent plagiarism questions. Only eight percent were concerned with a substantive dispute, related to requests for data falsification or manipulation.

Plagiarism - relatively easy to solve

Now the questions of plagiarism and authorship are about a necessary hygiene of science, which is comparatively easy to solve. But what about the reliability, the quality of scientific results according to your own standards?

A good overview of the quality problems in medicine can be found in an article by Cornelius Frömmel. According to a meta-study from 2009, of life sciences scientists in surveys an average of two percent admit that they themselves fake. A third state that they use “tricks” in their publications and also assume that two thirds of their colleagues “cheat”, that is, they improve or - let's call it by name - falsify results.

The costs of scientifically inadequate studies are likely to amount to several hundred billion euros per year worldwide. The consequence of insufficient publication of negative results is also underestimated, a fatal consequence of the scientifically absurd view that only positive results are a success. There is a lot of valuable knowledge here that something does not apply, not generally available to science.

Round numbers incorrectly to get a significant result

An example from social psychology impressively shows that simple fraud also plays a role and that these falsified “scientific findings” can attract a great deal of public attention. Diederik Stapel manipulated more than 50 studies. The "New York Times", the "Neue Zürcher Zeitung", "Spiegel Online" and "Die Zeit" wrote about Stack's experiments. Obviously, this is just the tip of the iceberg. A survey of more than 2,100 psychologists at American universities showed that 23 percent had rounded numbers incorrectly in order for the result to be considered significant. 22 percent stopped collecting data when the expected result was available and 43 percent discarded data after calculating how that data would affect the result.

So science produces not only bad, but even wrong results, also by its own standards, to a considerable extent. Even if the population is not aware of this fact in detail, this fact is nonetheless perceptible. In my opinion, this is the main reason for the loss of confidence in knowledge generated by science.

Forgery-proof protocols and laboratory books

Significant progress in quality assurance can be achieved with a few, comparatively simple measures: Compulsory training for young scientists in scientific work should be standard. The same applies to regular random checks of the statistical significance of publications.

The introduction of forgery-proof test protocols and laboratory books and the possibility of accessing the original data after a publication will discourage people from conjuring back the number of laboratory mice that interfere with the desired result. The obligation to publish negative test results will save you from having to reinvent the wheel again and again. In Germany, the German Research Foundation should spend at least one percent of its budget on quality assurance; that would correspond to 100 to 150 million euros annually.

The awakening of false expectations by science and false expectations of science, which arise from ignorance of the possibilities and limits of science, also make them lose trust. From my understanding of science and decades of experience at the interface between science and politics or society, false expectations in both directions can only be met if science faces the following four challenges.

Science should also be explained in German

First, science has a duty to explain itself to society in a language that people understand. I also expressly mean the German language here. Who, if not the cultural and social sciences, should provide society today with the orientation knowledge that it needs more than ever in a globalized world with drastic changes on a daily basis. Science will only reach the broader population and politics if it is also done in German. Quite apart from the cultural dimension that the German language becomes impoverished if scientific terms are no longer developed in it.

Second: The communication of scientific knowledge must always communicate its limits. The knowledge is limited by the subjective question and the method used. This is not just about science, but also the media as transporters and customers such as politics and business, who easily succumb to the temptation of simple truths and the supposed lack of alternatives.

Third: Scientific knowledge is neutral. As such, scientists cannot make political decisions. In addition to personal vanity, we are also moving in the difficult field of so-called party reports and, last but not least, in the necessary reference to the limits of questions and methods. Scientific findings are fundamental and indispensable for politics, but they only become a political decision through value-based weighting and weighing by politicians. The discussion in the field of tension between economy and ecology proves this impressively.

Science does not lead to the truth

Fourth, science does not lead to truth. Science is subject to a structural and irreversible restriction. We need to realize that science is so successful because it is the best way to get certain knowledge - knowledge - right now. This certain knowledge, however, and this is often forgotten, is relative and as such “certain” only in the awareness of the subjectivity of the question and the limitations of the method. To use Karl Popper: Scientific knowledge is always guesswork.

In this respect, in addition to all justified pride, science also looks good on an undercurrent of humility. In my opinion, the main causes of a decline in trust in knowledge gained through science are largely self-inflicted. It is own goals that endanger the importance of science in this society. Only when science resolutely tackles these problems can it do something for its own position in society. It has to, because without a trustworthy science there will be no progress for society as a whole.

The author is a board member of the Charité Foundation and was Senator for Education and Science in Berlin from 2006 to 2011. His article is based on a text in the volume “From dealing with facts. Answers from the natural, social and human sciences ", Eds. Günter Blamberger, Axel Freimuth, Peter Strohschneider in collaboration with Karena Weduwen, Wilhelm Fink Verlag, Paderborn 2018.

Now new: We give you 4 weeks of Tagesspiegel Plus! To home page