Bad science in the media

Take the case of the infamous arsenic-eating bacteria. In December 2010 no lesser organisation than NASA claimed to have found bacteria living in an isolated lake in California that appeared to feed on arsenic and may actually have incorporated it into their DNA. While not necessarily taking in all the biological subtleties of such a claim, most people would know that arsenic is lethal to all living things. So news of a creature that not only eats it, but possibly incorporates it into their ultimate fabric, must be big news.

The story was reported online through the prestigious journal Science and was picked up by news media all around the world. NASA had found a radically different form of life living right here on planet Earth. Or had it? Almost immediately some microbiologists spotted errors in the methodology and data handling of the original paper. The presence of arsenic was more likely to be a form of contamination of the sample. So they went out and did the decent scientific thing; they tried to replicate the experiments and failed to find the mysterious arsenic-fuelled DNA in the organism now known as GFAJ-1. 

When these replicate test results were released in May last year, they did not get the same level of media coverage as the original erroneous claims. It just wasn’t sexy enough to say that someone made a mistake. It was the scientific equivalent of burying a correction at the bottom of page 26 of a newspaper. I still get enquiries about arsenic eating bugs from people who heard the initial release but not the subsequent correction.

Poor studies are more likely to get picked up by the media if they reflect our prejudices and play on our insecurities. A recent study concluded that men under stress preferred larger women than their less stressed brethren. This touches several societal nerves; sexual attraction, women’s body image, questions of appropriate weight and obesity, men’s sexual selection criteria and how they could be manipulated. Add to this some quasi-scientific explanations based in our evolutionary history and you have a story the media just couldn’t leave alone.

More’s the pity, because the study was deeply flawed in several ways and the conclusions it drew were way beyond the limits of the data collected. The study was conducted on a small group of white male university students, some of whom were subjected to a ‘stressful’ simulated job interview while the others were not. The difference in mean size of the preferred women between the two groups was pretty small and the range of sizes preferred by the two groups overlapped to a large degree. Add to this the complexities of human sexual selection from cultural and other influences and there was nothing meaningful that the study could say on the matter. Yet it still made prominent positions in the media across the globe.

Similarly, scientists have recently had a strong backlash against the media overlaying anthropomorphic interpretations of natural behaviour in other species. Female Laysan Albatrosses are not lesbians just because they co-operate in nest building, egg sitting and rearing of their young, but that didn’t stop the headlines to the contrary. This says more about our societal angst around issues of homosexuality than anything about the life history of these innocent birds. And this is not just a recent problem with the media. It was recently revealed that the biologist on Robert Falcon Scott’s Antarctic expedition was so shocked at strange sexual practices among Adelie Penguins (including necrophilia, homosexual couplings and pack rapes of immature females) that he wrote his notes in Greek, so few could read them and he never published that part of his journals.

Given this litany of poor reporting of bad science or the misreporting of good science, you could be forgiven for not wishing to ever trust any media reports on science again! That is, of course, throwing the baby out with the bath water. There are a lot of good reports of competent science. In fact the bulk of media reports on matters of science do make a pretty good job of explaining and contextualising the original research.

In the end, as with commerce, the caveat should be ‘buyer beware’! If it sounds too good to be true, if it plays to the sensitivities of the audience and if it has not been replicated by other scientists, chances are it’s a beat-up. Yes, science is continually producing amazing and unexpected results but these all need to be read in the context of what is already known.

Good science does not happen in isolation and good science reporting only covers good science.

Dr Paul Willis is the Director of RiAus (Royal Institution of Australia)