The science conundrum

Are we really cut out for this science lark? With our emotions, whims and fancies; it seems unfortunate science is left up to us. People, I mean.

Science is objective, and unbiased. Humans are not. We are good scientists only when we overcome this…which is trickier than it sounds. Bias and subjectivity are very much part of our makeup.

True, we are simply fantastic at gathering data. Building exquisitely sensitive and powerful instruments to poke and prod the mysterious corners of cause and effect. To turn nature into data. The worry is how we turn that data into accurate knowledge.

There has long been a concern that, for example, non-significant results are less likely to be reported by researchers – and that even if they are, journals are reticent to publish them. This is publication bias – it is damaging to the endeavour of science, and to the public trust in that endeavour.

Many a meta-study has suggested this is a real concern. This is why a paper by Professor Johan Hollander from Lund University in Sweden and colleagues Christian Harlos and Tim C. Edgell is worthy of attention. Having conducted the largest meta-study yet of articles related to climate change, Hollander is confident – and indeed relieved – to report that there is no evidence small, statistically non-significant results are being under-reported¹.

While this is good news – in that is suggests the scientific method has worked on arguably one of the most important observations it has yet made – this is very much a paper of two halves. It also highlights anomalies that should give us pause. The most worrying to my mind is the trend for study abstracts, especially in high-impact journals, to well – how best to put this – ‘over-egg the pudding’ when it comes to the significance of the results buried deeper in the paper.

Hollander won’t, quite correctly, go as far as calling them outright lies, but he does suggest “exaggeration” on the part of the authors, editors or reviewers. Part of the cut and thrust, he suggests, of getting a paper published in a high-impact journal – you need a compelling story told at a glance.

This is, and I really can’t stress this enough, not to be taken by the climate change deniers as proof of their ‘cause’. Once again, the study found absolutely no under-reporting of non-significant results. What the paper does highlight, rather brilliantly, is that science is a human activity – with all that entails.

Yet here we are chipping away – and studies like Professor Hollander’s show that, with care and attention, we can slip the mind-forged manacles of our own biases. This is science attempting to correct itself and it makes me very proud – and should make us all proud – to be part of the species which seeks to deepen understanding in this way.

1: Harlos, C., Edgell, T.C. & Hollander, J. Climatic Change (2017) 140: 375.

Phil Prime, Editor