Conflicts of Interest
Posted by Thomas Butts, on 28 October 2013
‘‘None of the scientists would comment on the record, for fear that it would affect their funding or that of their postdocs and graduate students’’ Nature, September 2012
‘Is science wrong?’ The Economist, frontcover October 2013
A Disclaimer
I have been very fortunate to come across senior scientists in my career who do not correspond to the stereotype that I put forward below; quite the opposite. I would thus like to make it clear that this piece is not at all aimed at any particular individuals. It is simply a hypothesis as to why science is, or might be, wrong.
A hypothesis
The first quote above comes from a news article that accompanied the publication last year of the findings of the ENCODE project consortium, and was part of Nature’s coverage of the project that included a mini-discussion of the (now seemingly) age-old debate about the merits of funding large consortia vs. investigator-led, hypothesis-driven research.
It came back to me recently as I was pondering a podcast from that most esteemed of journalistic news sources, The Economist. In their 2012 review of the year, they (correctly in my view) called the ENCODE publications (there were 30 papers published across 3 journals: Nature, Genome Research and Genome Biology) ‘‘the most significant findings of 2012 in life science’’. Now, in late 2013, they have run a piece, and indeed a front cover, entitled ‘‘Is science wrong?’’. This new discussion examines not the philosophical underpinnings of the scientific method in its modern incarnation, as might be concluded from the title (and front cover); rather, they bring to light a debate that until now has, I think, been largely brushed under the carpet by the scientific establishment. It is one that the huge army of excellent scientists either having been previously, or likely to be in future, forced out of science have long been airing. Perhaps now that someone other than the disenfranchised have made the point, funding agencies and their government paymasters, not to mention senior scientists, will begin to listen.
The criticism boils down, ultimately, to a single contention: that the modern career structure in science contradicts the very aims of science itself. Journals don’t publish repetitions of previous work, funders won’t fund it, and guess what? Scientists won’t do it. Cue massive surprise and outrage when significant proportions of published work are not repeatable. If you recruit some of the brightest people on the planet whose only way to remain doing what they love is to play a system that pits them against each other in an incredibly rudimentary way, then you cannot act surprised and outraged when they play that system to the detriment of the scientific enterprise.
A significant problem exists within the scientific community, however, that perpetuates this wrong: no one who in a position of power seems prepared to change it. To say that science attracts egos is something of an obvious statement to anyone who has ever met a scientist (or read one of their blogs), let alone been to a scientific conference and seen them talk about their work. The scientific career structure at present does not simply allow self-promotion, it absolutely depends upon it. That this is not in the interests of science as a discipline (and of course the future generations of the world who will depend upon its activity for their technology) should be self-evident to anyone with a brain, let alone the superb education of those at the top of the scientific establishment.
However, if the people who get to the top of science have been rigorously selected for their ability for self-promotion, then those same people are very unlikely to set about systematically re-designing the system for the better. It is rather depressingly reminiscent of a careers session that I heard of recently where some young female PhD students were quizzing a (very) leading light of the scientific world about career planning. ‘Families are not an option’ was the gist of the response from the luminary. She was female. A woman who feels she has had to give up the very notion of family life in order to get where she is surely is less likely to advise young women coming after her that they should be entitled to have both career and family. I may be wrong – I am certainly not a scientific luminary, let alone a woman – but I suspect that I am not.
Science shouldn’t be like this. Surely, we actually do know better. So should funders. Some lone voices have been saying this for years and years. Perhaps now some journalists have cottoned on to it, somebody with power will actually listen. It is, quite literally, not rocket science.
(8 votes)
Loading...
While I completely understand the point of this (much needed) commentary, and I agree for the most part, I have to take some issue with painting with such a broad brush. As a soon-to-be Ph.D., I recently attended a conference and was honored to be presented with a travel award as a result of my having been selected to give a talk. Neither I, nor any of the other speakers or attendees- which included some luminaries from around the world- engaged in self-promotion of the kind to which you seem to refer.
And a question, perhaps from my naïveté: what is the connection between the lack of reproducibility and self-promotion?
Hi,
Thankyou so much for the comment, and my apologies for not saying so until now (responding to referees, email backlog, experiments over Christmas blah blah blah).
You are of course quite right to point out that most scientists, of whatever level, are not solely engaged in self-promotion – I concur that most of us are generally interested in the natural world.
The connection between the lack of reproducability and self-promotion is that the modern uber-competitive career structure discriminates against scientists who replicate others findings, and this is not a good thing.
Finally, the very best of luck with the PhD submission, and if we meet at a conference, we can chew the fat over a beer I hope!
Tom
Reproducibility (or lack thereof) is clearly an important issue, but I’m not sure it’s entirely fair to say that it’s a problem that isn’t being tackled – although clearly we can do more. One interesting recent development is the Reproducibility Initiative (https://www.scienceexchange.com/reproducibility), which has recently been set up that aims to subject papers to validation, and provides an outlet for the publication of these studies.
It remains to be seen how well this works, and whether due credit is given to people who take the time to conduct reproducibility studies, but at least there are now moves in the right direction…
Hi Katherine,
I had no idea about this, so thankyou for making me better informed! I hope that this is evidence of things moving in the right direction..
Hi Thomas,
I think you raise some interesting and important points in this piece, but it seems to me that there are two main issues that are perhaps better treated separately.
(1) The reproducibility issue.
This is certainly a problem, and you are correct in that, under the current system of doing science, it is not possible to correct this. People who are competing intensely for money for new research will hardly manage to obtain funding to do something that has already been done, even if they were inclined to do so. Short of setting aside funding specifically for reproducing data, I don’t see how this can be solved in the long run.
(2) The competitiveness issue
I’m not sure if that is the correct term for it, but I certainly see where you’re coming from. The intense competition for funding (not to mention positions, publications, notice by your peers, etc.) does effectively set up a selection scenario. It is interesting that you pick up the ability to self-promote as something that we could consider “selected for”. I think that’s true, and I also think your point about the luminary telling the questioner that a family wasn’t an option also points out that there is selection for those who will work at their projects to the exclusion of everything else. Something I have noticed myself (in a subjective, non-quantified way!) is that younger PIs in particular, under this enormous pressure, are increasingly not like the professors who taught me, in that they are far less likely to be polymaths. They don’t have time to be interested in anything else, let alone excel at it. Granted, they were older, but even so, if younger scientists don’t even have time to read about something other than science, it’s bad news for science and the human beings doing it. The main driver for people becoming interested in science is curiosity, and I feel it’s a crippling shame (and very short-sighted) that people must effectively crush their curiosity about anything else. As for the self-promotion aspect, I feel that we are selecting for the hugely ambitious – and there is a risk that this will promote people who are prepared to put ambition ahead of principle, something that should not happen in science.
One final point about your questioner asking about families, as this is a particular bugbear of mine. You note that both were female. In my experience, that question has *only* ever been asked by one woman to another, which reveals something very fundamental about the assumptions in our society. Why don’t men stand up in careers talks and ask other men how they balanced their careers with their families? In fact, I challenge you to do this at the next opportunity. I imagine they will be flummoxed. Women ask themselves this question all the time; historically, at least, this has never been an issue for men. They apparently simply *assume* they can have both. Equal parental leave would mitigate some of this, but challenging the assumptions of both genders would be perhaps just as fruitful.
Jo, as ever, a pleasure – I shall indeed ask this at the next career session I go to (or at least try to!).
I also totally agree with you about polymaths, and now I think about it, I think you are onto something with the correlation with age, ironically despite the fact that it is now super-trendy to be ‘interdisciplinary’.
The Economist “Cherry Picked” the Data!
Like several criticisms of science that I have read recently, the Economist attacked scientific methods using a statistical argument that purports to show that scientists routinely make type-one and type-two statistical errors. In the article, the Economist assumed that all statistical data has a p-value of 0.05. Based in this gross assumption, the article calculated the error rate for scientific inquiry. The author should have realized that scientific results often come with p-values of 1X10-9, not just 5X10-2. The article appears to have been intended to show the weakness of scientific method, but clearly demonstrated the bias and lack of sophistication of the writer.
I totally agree about the statistical point. There are a number of other annoying things in the article that underline the point that the author clearly has not ever done any science. Which perhaps suggests that science journalism needs an overhaul of similar proportions to the one that he is suggesting for science itself.