Science Philosophy · 13 October 2011 ·
"I am not going to read your diatribe!"
Paul Feyerabend was one of Popper’s earliest students and followers, but soon turned to criticise Popper’s theories vigorously. Neither of them were shy of words when criticising colleagues’ works, nor would they try to conceal their criticism. Feyerabend recalls the following dialogue:
“‘I am not going to read your diatribe!’ Popper had shouted when he saw my [Paul Feyerabend’s – Antipattern] comments on his diatribe against Bohr. (He calmed down when I told him that many people had complained about my aggressive style and had ascribed it to his influence. ‘Is that so?’ he said, smiling, and walked away.)” (in: Feyerabend 1995, p. 146)
I’ll write a little more on Popper and Feyerabend soon…
Paul Feyerabend (1995), Killing time: the autobiography of Paul Feyerabend, University Of Chicago Press
Politics Science · 15 June 2011 ·
The impact of external funding: Deutsche Bank and the Quantitative Products Laboratory
Both, the German daily TAZ and the weekly Der Spiegel report that the Deutsche Bank has financed lectureships and research projects at Berlin universities, and supported the foundation of the Quantitative Products Laboratory, a research institute in applied financial mathematics. The institute was founded in 2006 and is a jointly operated between the Free University of Berlin and the Humboldt University of Berlin. This in itself isn’t unusual. However, the contract leaked by Peter Grottian, a former lecturer at the Free University, reveals the details and conditions of the financial support by Deutsche Bank, and is surprisingly explicit about the rights of the bank to influence the institutes’s research:
- The institute had to be in close proximity to Deutsche Bank, it moved into the same building as the bank’s “Investment & FinanzCenter” in Alexanderstraße.
- The institute had to display marketing material of Deutsche Bank.
- The institute may not mention the bank in published research without prior permission granted by the bank
- The bank reserves the right to send off members to oversee examinations.
- The bank reserves the right to veto the appointment of professors.
- The bank reserves the right to withhold publication of research if it conflicts with its economic interests, up to a maximum of two years.
All this is bad enough, but the worst is in my view the last point, i.e. the bank’s right to withhold publication, because it is particularly detrimental to the purpose of science as a mechanism of acquiring knowledge. Not only are the outcomes of research withheld, this in itself is only a minor issue. Whether or not a specific study is being published or not does not so much matter for scientific progress. Much worse is that it adds onto an already existing systematic bias of the published literature.
In science there are plenty of such publication biases. For example, journals tend to prefer publishing research that has significant results. Studies that fail to do so are often rejected, on grounds of lacking novelty. Thus, B. Goldacre describes the case of recently published article by D. Bem on precognition, or rather the ability to foresee the future. Whilst the original article was published due to its extraordinary findings, three further studies that replicated the experiment and found no effect were rejected on grounds of lacking novelty. But it is particularly the replication of prior research that is so vital in ensuring that the original findings are not just a statistical blip. If peer-reviewed journals withhold the publication of negative results and failed replications of prior research, they are no longer indicating the current state of scientific research. But I’m digressing.
Bias is already a big problem in scientific publishing, and it is worsened if another layer is left to interfere with it, by publishing research that supports its interests, whilst holding back the results that don’t. It leads to a skewed view of research and inhibits what science is all about: Progress and impartiality. If science is bound to the interests of companies, it will not be able to bring us closer to understanding the world.
Let me make up an example to illustrate my point (I have studied psychology, not financial mathematics, so it’s a bit arbitrary and vague): Let’s say a series of studies at the Quantitative Products Laboratory studied a particular kind of investment. Early research suggests it has high profit margins and the bank starts investing. However, later research reveals this kind of investment is not as safe as initially assumed. In this case the bank might make use of their right to withhold publishing, to disguise their false investments and not scare off current or potential customers. Thus, positive results would be published, the negative ones would get an extra hurdle they would have to take, and a skewed picture would emerge. Instead of leading to greater understanding of an issue, it leads to greater confusion and obfuscation.
Although the problem is not as prominent in areas like financial mathematics, its effects are much more prominent in medical research where drug companies have to make massive upfront investments to develop new drugs. The effects of withholding unfavourable research might even be lethal, if for example side effects of a new treatment are covered up, and only positive outcomes are published. And although most private funding contracts don’t grant explicit rights like the Deutsche Bank, interference often remains on an implicit level, between the funders and the operating research institute.
In the current economic climate many governments seek to reduce costly scientific research and would like to see private investors to step in. But if there is anything we can learn from examples given above, it is how vital independent, state-funded research is.
The contract between the universities and the bank is currently under revision, and both sides indicate that these terms will not remain. Furthermore, Deutsche Bank claims to not have made use of their right to withhold research. Cynics would reply that, given the close affiliation of the research institute to Deutsche Bank, this shouldn’t come as a surprise.
For further reading on publication biases, I can recommend J. Lehrer’s article The Truth Wears Off and D. H. Freedman’s article Lies, Damned Lies, and Medical Science.