By Starr, Douglas
Nieman Reports , Vol. 56, No. 3
Once upon a time science writing was simple: A reporter would read published studies in the scientific literature and write about the latest wonder of research or miracle of medicine.
Things have gotten more complicated since those early days of science journalism. The spread of pollution, the Vietnam war, the Chernobyl meltdown, the Challenger explosion, the emergence of AIDS and antibiotic-resistant bacteria have all revealed a darker, more vulnerable side of science. This is not to say science has gone bad: Our lives have been extended through medical advances and improvements in diet and made more convenient with personal computers and inventions so ubiquitous we take them for granted. Science has become a complex story that can no longer be portrayed as an isolated or idealistic pursuit. What happens in science affects us all and is influenced--even shaped--by money, special interests, and politics. In short, we need to report science as part of the real world.
Given that reality, we teach our graduate students at the Knight Center for Science and Medical Journalism at Boston University to view science in a more interwoven way than it was reported in the old days. Increasingly our discussions focus on context--fleshing out the scientific, economic and social aspects of issues to illuminate their relevance and meaning.
A few years back, students in science journalism courses would be asked to find newsworthy journal articles and "translate" them for the public. Nowadays, we no longer do this exercise since the most skillful journalists act as analysts, not translators. That is not to say that reporters shouldn't follow the literature--with extensive science backgrounds, most of our students already do. But rather than focus their work on a single study, our students use such reports as a point of departure for interviews and other research to reveal the broader currents in the field.
In doing so, they investigate the work on several levels. First, they flesh out the intent of the study--for example, whether it demonstrates correlation or causation, a straightforward distinction that reporters sometimes miss. They determine whether a study's conclusions follow logically from the methods. (You'd be surprised how often they don't.) They challenge the statistics: When a study reports a 50 percent increase in brain cancer among laboratory rats exposed to a certain chemical, the results might sound alarming until the reporter asks about the sample size. If the researcher replies "three," the finding is less significant.
They also ask contextual questions about how this particular study compares with others in the field. What similar studies have come before? How is this one different? And perhaps most importantly: How does this study add to or contradict the existing body of scientific opinion? Such questions situate the work and help separate genuine news from institutional or media spin.
Nowhere is context-setting more important than in medical and nutrition reporting. Readers infer advice from our articles, whether or not we intend it. Plagued by fad diets and simplistic solutions, consumers are often confused by successive articles suggesting conflicting advice. For this reason, it's especially important to spell out the difference between a definitive study and a work in progress and to compare the current work to the broad consensus of scientific opinion.
Two questions are particularly helpful:
1. Are the studies so powerful that readers should change their medication, diet or behavior?
2. What would be the effect of changing those behaviors versus keeping them as they are?
The perils of ignoring such questions became apparent in a cover story last summer in The New York Times Magazine. This bold and contrarian article alleged that the conventional food pyramid we've come to rely on actually has caused the national obesity epidemic, and as a solution resurrected the largely discredited Atkins high-protein diet. …