Reading the journals: From new science to science news

May 22, 2013

By Thomas Hayden

When I was a graduate student, I enjoyed Fridays most. Those were the days I'd sit with stacks of science journals and just read for hours, without much direction or purpose other than to immerse. I still read a lot of journal articles when I became a science journalist, but my relationship to them changed dramatically. Where once I meandered, now I scanned with a purpose, and went deeper only when my news-sense started to tingle.

Whether looking for story ideas, writing "paper of the day" science news articles, or compiling background research for an ambitious feature, every science writer needs to be able to dive into the literature quickly, and come out not just with information, but with news, stories, a list of sources to interview, and questions to ask them. Here's a version of the method I teach my science journalism students at Stanford, working off the standard journal article format. But I know it's not the only approach — please share your tips for reading the literature in the comments.

Scan the title and abstract

No big surprise there, but remember that these summaries are for other scientists, not for journalists or the public — you won't find everything you need here. Glean what nouns and verbs you can, though: what was done to whom, how wide or narrow was the scope, and what, if any, wider implications are suggested?

Embrace the introduction

Science is incremental, and the Introduction is designed to place the current study in the context of historical and ongoing work in the field — what it is building on, tearing down, or competing with. It is the science writer's best friend, in other words.

  • Read the first paragraph or two, and you'll know the intellectual context for the current study.
  • Read the next couple and you'll have a short list, from the citations, of who else is active in the field — and a great start on who you should interview.
  • Zero in especially on the last sentence or two, usually preceded by "Here we report" or language to that effect. That's the most direct answer you'll get from the paper to the question, "but what did they actually do?"
  • Keep a running list of unfamiliar terms and acronyms as you read. I jot them down, along with definitions, in the margins for easy reference. You'll need them to understand the paper, and for interviews — and you'll need to figure out how to convey them in actual English for your finished story.

Jump to the last few sentences of the discussion

Just after the authors note "more research is needed," you'll usually find the one moment of speculation allowed in most papers. That's where scientists get to suggest not just what their study contributes to the research enterprise, but what deeper implications it might have, or even how it might be applied. This is as close as the paper will come to answering the question, "So what? Why does any of this even matter?" [Note to science reporters: Your job is to push the researchers to tell you more about this. Their job is to resist.]

Scan the methods section

It's written to allow other researchers to replicate the work, but that's not why you read it. As a writer, you're looking for verbs again — what did the researchers do? Was it a lab study or field research? A clinical study or a meta-analysis of previously published studies? And especially, did the researchers do anything — go to Mars, invent a new technique, wear funny clothes, whatever — that might find a place in your article? Science is a human endeavor, after all, with its full share of frustration and adventure, tedium and triumph. You won't find anything more than a hint of drama in the paper, but the methods often hide just enough to guide your questions in an interview.

Look at the results

And I do mean "look" — often, a quick scan of the tables and charts is enough. You want to make sure the techniques are robust enough and the reported data significant enough to justify that speculation at the end of the discussion. Sure, it's the job of peer review to do that first. But it's your job as a science journalist, too — and can help you pose a good set of questions for your outside sources. Most of the time, it's just a box to check. But sometimes, faulty techniques or weak results actually become the story.

Back to the discussion

Scientists spend a lot of time worrying — and arguing — about caveats, exceptions to trends, and minute details at the boundary between what is known an what is unknown. Well, they should — but that doesn't mean you have to, too. One of the hardest lessons for a scientist-becoming-a-writer to learn is that most of that one-hand, other-hand detail — as important as it is within science — usually falls beyond the scope of all but the most specialized science writing. Writers have to read and understand it all — and keep an eye open for relevant, enlightening or enlivening information — but you'll have a better search pattern for that sifting if you leave the bulk of the discussion for last.

Bonus steps:

  • Author list: Papers can have anywhere from one to dozens of authors. Usually, if you interview the first and last authors listed, you'll have talked to someone who actually did the work (first) and someone who knows its broadest context (last).
  • Submission date: There's always a lag in scientific publishing. Most papers note when the manuscript was first submitted, and often when revised, and/or when approved for publication. Sometimes rapid turnaround times signal hot research. But long turnaround times can be just as interesting — if the publication process took months, the researchers may well have new results in hand that are even more interesting by the time you call and ask. Not asking means you're reporting old news, even on the day a paper is first published.

Thomas Hayden is director of Stanford University's Environmental Communication Master of Arts program. He was a staffer at Newsweek and U.S. News & World Report, co-authored two books, and co-edited The Science Writers' Handbook.

Image credit: Vmenkov [CC BY-SA 3.0 or GFDL], from Wikimedia Commons