I Bet It's Biased: One Easy Step to Squash Expert Opinions

First Posted: Jun 17, 2013 10:40 PM EDT
Close

Accusations of scientific bias are the new catch-all weapon for anti-science commentators and climate deniers. Why are they effective? New research shows they may exploit part of our psychology.

By James Gilbert

Climate sceptics have won, Martin Wolf lamented in the Financial Times, despite near-universal scientific consensus against them. The sheer longevity of this “debate” indicates deniers attract disproportionate attention – partly due to one of their main lines of attack: scientific bias.

Attacks on scientists' financial and political motivation are increasing. We hear them not only from committed deniers, but also from commentators in mainstream media. Even politicians and US presidential candidates are unafraid to label climate science a “hoax”.

Now, a new study in press at the Quarterly Journal of Experimental Psychology has shown that the public is particularly sensitive to financial bias when placing trust.

Brent Strickland of Yale University and Hugo Mercier of the National Centre for Scientific Research in France asked non-scientists to evaluate scientific studies. Participants decided whether they believed the results of experiments, given what experimenters expected to find, their financial motivation and whether the methods were sound. The results showed that mentioning financial bias reduced people’s belief in findings, even if the methods were flawless.

This is my truth: tell me yours

High profile debates over HIV, passive smoking and, latterly, climate change, have increasingly questioned the social context of science as well as the facts. They gained momentum during the “science wars” of the 1990s which popularized the idea of scientific findings as socially constructed, inseparable from the people producing them. Now, accusations of scientific bias are going viral.

News presenter Megyn Kelly takes pundit Erick Erickson to task for claiming mothers harm kids by pursuing careers. Around midway, Kelly cites a particularly thorough meta-analysis – a study specifically designed to address bias – which shows his claim is false. Erickson responds, not by attempting to address the science, but by saying “experts can be as politically motivated as anyone else”. Then he gives his opinion, as if Kelly had not spoken.

Erickson asks not whether this is good science but instead whose side the scientists are on. His hope is to create a stalemate where, owing to suspect motives, no science is valid. In this vacuum, his opinion is as good as anyone else’s.

Bias is real …

Why are Erickson and others able to get away with this? Unable to challenge scientific consensus with data, deniers can be confident accusing the establishment of bias. No amount of data will shake this claim, because it is impervious to facts.

And scientific bias is indeed corrosive – we have worried about it for many years. Ben Goldacre, a medical doctor and journalist, for example, has been sounding alarm bells for more than a decade. The bigger the incentives to manipulate results, the harder it is to dissociate findings from motives.

… but it doesn’t undermine all of science

The more high profile these arguments become, the more people get used to the possibility of a giant scientific plot fuelled by career-protecting scientists, and the further their trust in scientific findings, however unanimous, is eroded. The problem is exacerbated by the fact that strictly balanced media over-represent minority views, which are typical of many denialists. A BBC review concluded that climate coverage was skewed towards sceptics.

Thus, when a recent study suggested conspiracy theory was associated with climate scepticism, critical bloggers' predominant response was more conspiracy theory. In the right hands, our sensitivity to financial and political bias becomes a potent anti-science weapon.

If scientists are constantly forced into disclaiming their motives rather than explaining results, then Wolf is right: the deniers have won. The debate is no longer about science, but about bias.

What can be done?

Critics are right to be concerned about bias, but within reason. It cannot be used as a wholesale indictment of science. Numerous studies show that bias can be reduced in various ways: increased sample sizes, carefully controlled designs, meta-analyses, pre-trial declarations and commitment to publishing all results, including negative ones, under open-access frameworks.

Academic platforms routinely declare conflicts of interest. Some, like The Conversation and PLOS journals, place declarations very prominently. Possibly, industry-funded studies should be identified as such in their titles, or appear in separate journals.

Unfortunately, denialist accusations of systematic bias rarely rely on specifics that can be challenged. They invoke a fuzzy, boogieman concept of “bias”, such as Erickson’s blanket statement “experts can be as politically motivated as anyone else”. This is as undeniably true as it is undeniably vague. But if it is repeated unchallenged enough times, we stand a real risk of this argument successfully undermining public confidence in science.

Governments are not currently helping: examples include tying funding from national research institutions to economic gain in the UK and in the US, or allowing unlimited conflicts of interest among FDA advisors. Accusations of bias must be challenged at each turn by scientists and communicators alike. The sooner the vast scientific conspiracy is discredited, the sooner we can get back to the science.

James Gilbert does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations. He receives funding from the European Community’s Seventh Framework Programme under a Marie Curie Fellowship.

See Now: NASA's Juno Spacecraft's Rendezvous With Jupiter's Mammoth Cyclone

This article was originally published at The Conversation.

Join the Conversation

Real Time Analytics