Scientists are riding high at present. Trust in science looks to have risen on the coattails of medical science during the COVID-19 crisis.
But while many international studies and polls show growing public trust in science, another suggests trust in individual scientists will fall.
It may be too early to tell — trust will depend in large part on whether measures like distancing and quarantine do actually work in the long run.
During this science week, which ends Sunday, August 23, some peak bodies have gone hard on the COVID-19 angle. Science and Technology Australia headlines “STEM experts guiding Australia’s COVID-19 success …” who “help the nation and humanity in our hours of greatest need”.
This adds to the framing of the pandemic as war, justifying extreme measures and technocratic solutions. It will play well with those Australians who share that view while excluding others. It is a hostage to the future if measures advocated by public health experts turn out poorly.
The Australian Academy of Science is more balanced. While it had one COVID-related publicity stunt, its overall take on Science Week was diverse. Its stories for the week include marine biology, astrophysics, birds, bees, artificial intelligence and the environment, among others.
This seems wise. Although there are signs of a turnaround in the long-term trend of declining trust in all institutions — a turnaround also enjoyed by politicians and the media — at the same time science is in crisis.
Universities, where most Australian science is done, are facing unprecedented financial pressure due to the collapse in international student numbers. Many young scientists in casual positions are out of work. International collaboration, although partly manageable at a distance, will also suffer.
But problems run deeper than the present crisis.
Bad research infests science faculties. It is not common, but occurs often enough to discredit the entire endeavour. In some disciplines bad research is merely embarrassing. In science, bad research — like discredited studies on vaccine dangers — can cost lives.
Much of science, including medical research, faces a replicability crisis: that is, studies producing results that follow-up experiments cannot duplicate.
This casts doubt on the original findings. Were they perhaps invented for the purposes of obtaining a publication rather than being based in reality?
Both the replication problem and fraudulent articles stem from the same root cause — the pressure on researchers to publish. An article with interesting and exciting results is much more likely to be accepted.
Not surprisingly, academics are therefore tempted to embellish their findings to make them more unusual and noteworthy. It is to the credit of science training that not more researchers do so, given their prospects for pay and advancement depend so much on their publication record.
But it is a real problem for research. There is now a growing movement for creation of journals dedicated specifically for replication studies, repeats of past experiments to prove them.
The pressure to publish is a particular problem for Australia because many researchers are judged on the quality of the journal in which they publish. Quality rankings are subjective and favour “international” (largely American) journals.
Some articles don’t relate to particular countries (distant nebulae, a deep ocean trench, a mathematical proof) and can be published anywhere. But where research is Australian-based it runs a risk of being rejected by an overseas journal.
The confluence of fundamental problems with the COVID-19 moment creates a unique opportunity for a fundamental rethink in the way Australia funds and supports scientific research.
There are many possible options: remove pressure to publish; reward Australian relevance; give replication the same standing as original studies; replace “peer” review by cronies with review by trusted experts; allocate grants randomly among all applications that meet basic quality standards; give more research funding direct to universities instead of through the Australian Research Council and National Health and Medical Research Council (or, vice versa, increase the power of the these bodies to set research priorities); require applications for research grants to be much briefer; make citation of fraudulent research an offence.
There are numerous options, some better, some worse, worth considering.
What is needed now is a mechanism to crystallise discussions already widespread among both scientists and public policy makers.
It would be just as much a mistake to leave this to science as it would be to leave it to government to solve. We need independent and rigorous analysis.
It would make a good Productivity Commission inquiry. If inter-disciplinary rivalry between scientists and economists makes the commission a non-viable route, an independent inquiry would be an alternative way to gather our thoughts on how to solve the crisis.
However it is done, don’t waste the opportunity.
Science delivers huge benefits in science week and any other week. It can be even better if freed from perverse incentives and misdirected institutional pressures.
Stephen, thank you for this article capturing a problem well known to the sciences, and making it of broader interest. I think it’s a timely issue if only because of just how fast preliminary medical innovations are now being hothoused into commercial release.
Yours is the first commentary I have read suggesting that the reproducibility of science should be broadened into public discussion.
I agree in principle, however I’m deeply concerned about the way the role of evidence is taught and used outside STEM, including in public policy, political journalism and those parts of humanities education that typically inform them.
It’s not that I wish to distrust people outside of STEM; it’s just that they seem to be both uncomprehending of and uncommitted to the use of evidence in refuting what they believe, rather than supporting it: an approach well-established in the sciences for centuries now.
I realise that some areas have more discipline than others, and have seen that discipline at work. However, when the sciences have already been hyper-politicised and are now being turned into commercialisable hero-stories, how does a strategic question affecting science funding, productivity and integrity for future generations get safely separated from the populist sound-bite policy-announcements of our modern day? (And haven’t we already seen politically-motivated oversight bodies mooted for Australian sciences?)
The prospective use of the Productivity Commission is a good first thought, but do you feel its other reviews have been safely quarantined from such very issues in recent years? Does the Commission itself?
This sounds like the beginning of an important discussion rather than the middle of one, and your own article suggests that you feel so too.
I hope to see you contribute again.
thanks for a very good comment.
Yes, corruption of evidence in the political sphere, including political journalism, is a problem. There’s even a phrase for it – “policy based evidence” (a play on what used to be a laudable goal, evidence based policy). Much of what is today touted as evidence would be more accurately described as views or opinions. It is so-called evidence gathered to support a preconceived policy position.
On the other hand, there are instances where public policy is being developed on the basis of much more rigorous evidence than used to be applied, including randomised control trials (member for Fenner Andrew Leigh is a big fan of RCTs). There are rays of hope.
Still, there is a huge potential for misunderstanding between politics and science around evidence. I agree it would be worth another followup article.
You rightly comment about the use of evidence in science as the basis for refuting a belief. As you obviously know, but for the benefit of other readers, one of the key concepts in science is falsifiability. If results cannot under any circumstances be able to be disproved (which is not the same thing as those results being wrong!) then they are more religion than science. Science is based fundamentally on testing and scepticism; which unfortunately opens the door to people who want to distort, deny or misuse it.
Some commentators argue that the PC is not as independent of thought as it once was. I feel it’s still quite independent; certainly more so than policy areas of core government departments.
Stephen thank you for your reply.
Though not an economist myself I have been a fan of Dr Leigh’s work from back when he was publishing from the ANU. A lot of people enter politics with clear and good intentions, but he’s one of the few with the intellectual rigour to keep that clear between election cycles. He’s not afraid to math, measure and test, and treats failure as knowledge; federal politics alone could use a dozen more like that of any political persuasion, and if it had them I’d be less worried about entrusting questions of how to improve scientific productivity and integrity to broad public discussion.
You wrote:
While true, it’s now even worse than that, isn’t it? If public commentary isn’t cherry-picking facts, ignoring the relationship between good decisions and sound data, dismissing tolerances and responding to challenge with rhetoric and ad-homimens, isn’t that now seen as a failure of vision and relevance?
We might both describe ‘truth’ in the empirical fashion: a falsifiable claim that can be systematically verified and independently refuted if wrong. But the world of public discussion is now a postmodern one where any quote is a fact (because it’s a record of an utterance), data are somehow not facts, but part of ‘analysis’ which is now synonymous with the publication of rhetorically-ornamented opinion.
In my most idealistic moments, I could imagine the discussion about scientific productivity and integrity as a Town Hall one, and I’d love to visit that world.
I imagine citizen questions like: why haven’t the failure rate and adverse effects of flu shots been publicly discussed? Why isn’t the publication of data rewarded in academia, with interfaces to let ordinary people use it too? Why are complimentary medicines given vast shelf-space in pharmacies alongside medicines that actually work? Why isn’t citizen science given the same social encouragement as sports participation or building playgrounds through a service club?
But I don’t think that’s the public sphere we’re living in, do you? How do you imagine media reporting about a public science integrity discussion? How would news media get the public to engage sensibly with its tinfoil hats off, and who do you trust to report it lucidly?
(I realise you don’t get paid for these chats, so I hope you’ll consider my questions in a subsequent article, and hope to see it published here. Crikey’s pages can be a bit wilder than say The Conversation or The Mandarin, but the discussion is broader, and sometimes better.)
Traditionally, ie the past 40 to 50 years, journals have not published research outcomes that revealed no positive results. Ie, tested a relationship and found it didn’t appear to exist. But this is just as important as finding a relationship.
I’m also a bit more leery of the Productivity Commission, because it has been ‘mainstream’ in its economic theory, ie, more Chicago than anything else. Neo-liberal with a few ‘nudges’ toward the behavioural economics that gave us ‘nudging’ consumers in the direction of conformity to the needs of the powerful.
By contrast, the very unfashionable evolutionary economics has been suggesting for decades the very concentration of economic power that is strangling capitalism today. Don’t wait up for another Keynesian revolution to save capitalism, it won’t happen. What will come next? I’ll be dead before it comes, so no comment.
Not only should we be paying more attention to scientists but we should be studying and applying the science of government, cybernetics.
The lack of independent, non pharmaceutical company benefit, in medical research is a major problem. My ability to trust a vaccine for the corona virus is minimal – too many short cuts and insufficient research of how the virus has affected those who had it. Until Australia sees value in real scientific research funded by independent (either government or philanthropic) means our scientific research institutions and the good scientists trying to work in them will be flawed.
I always love the way that conspiracy nuts attack science using devices that only exist as a result of science.
Imagine if we still had CSL in public hands and the CSIRO were not run by a commercialism obsessed ex “entrepreneur”.