As a science journalist, I have been reading academic papers for 30 years. I would guess I’ve read tens of thousands of them looking for new advances to write about or research background information for. Although I’m not a scientist myself, I made it fairly easy to find my way around them.
One lesson I learned is that it can take work to put together the story behind a paper. When I call scientists and just ask them to tell me what they did, they can offer me an exciting narrative about intellectual research. But on the side, we readers have to put the story together for ourselves.
Peer review requirements – which meet the needs of several different experts – can make reading articles even more difficult. Journals can make matters worse by requiring scientists to cut their papers into pieces, some of which are banished to a supplementary file. Reading a newspaper can be like reading a novel and only realize at the end that chapters 14, 30 and 41 have been published separately.
The coronavirus pandemic is now an additional challenge: there are far more articles than anyone could ever read. If you use a tool like Google Scholar, you may be able to refer to some of the articles already cited by other scientists. They can provide the outlines of the last few months of scientific history – isolating the coronavirus, such as sequencing its genome, discovering that it spreads quickly from person to person before symptoms appear. Papers like this are quoted by generations of scientists who are not yet born.
Most won’t, however. When reading a scientific paper, it is important to maintain healthy skepticism. The continuing flood of papers that still need to be peer reviewed – so-called preprints – includes many weak research results and misleading claims. Some are withdrawn by the authors. Many will never make it into a diary. But some of them deserve sensational headlines before burning out in the dark.
In April, for example, a team of Stanford researchers published a preprint claiming that the death rate of Covid-19 was far lower than estimated by other experts. When Andrew Gelman, a statistician from Columbia University, read her preprint, he was so angry that he publicly asked for an apology.