Science. Communication. Community.
Journalists are turning more and more to data to guide story-telling and spot news trends. What does this influx of data mean for science reporters?
After taking the weekend off in celebration of turning in my Ph.D. dissertation, I arrived in lab this past Monday less than fully motivated. After over a month of intense academic writing and a defense presentation, I had returned to my desk as if nothing had happened. My groggy mind argued as coherently as it could that I just needed a few more days.
A quick glance at my calendar, however, snapped me back to my senses. I was about two months away from official graduation and potential unemployment. So although I had already made attempts at finding science writing internships, it was time for me to get serious about looking for a real job.
An NPR-junkie still semi-delusional from finishing my thesis, I started big, and began searching for openings with my dream employer: New York City’s public radio station, WNYC.
Aware of the tight job market—especially for those looking for steady gigs in the media business—I wasn’t too surprised that WNYC didn’t have an opening for a science reporter. They did, however, want a ‘digital reporter/data journalist.’
Fancying myself as pretty familiar with data, I clicked on through, only to discover that the carefully evaluated 3,000+ data points in my thesis were not going to be enough. The job required substantial statistics knowledge, and experience with statistical programs and computer languages, like R, that I had heard of, but had never actually used. Having contemplated the switch from the ivory tower to the newsroom for a while, I was used to being underqualified on the journalism side. Remarkably, now it appeared that despite my otherwise hefty credentials, I lacked some critical science skills, too.
Humbled, but also intrigued, I began looking into this sub-field of journalism that is just beginning to make its mark on news desks and readers alike. What follows is a whirlwind tour, with an eye toward what it all means for science journalists.
What is data journalism?
Data journalism, sometimes called data-driven journalism, uses data to identify news stories worth telling, and/or uses data in the telling of a story. For instance, examining the salaries of government workers might lead to a story about corruption, or analysis of economic data might suggest that wealth disparities are actually significantly higher than most people suspect. In the second case, the wealth distributions could then be graphed in a meaningful way to begin to change these incorrect perceptions.
Data journalism often, but not always, involves telling the story in part through infographics or interactive applications. These platforms can display far more information than written text can, and if organized properly (see Rebecca Widiss’s earlier post), can tell both general and specific stories. For example, the New York Times’s “Degrees of Debt” series included an interactive graph of student loan debt, which when viewed at the macro scale easily makes the point that both public and private college debt has been dramatically increasing in recent years. Equally powerful is the ability to zoom in on an individual school—say the one you graduated from—and enter your own debt, and see how you compare to your classmates and peers across the country.
In an age dominated by just seconds-long page views, these types of novel data visualizations are eyeball magnets, so they’re popular with news corporations as well. But they’re also incredibly time-intensive and require specific skills, which makes the more complicated projects the domain of large news groups or highly dedicated, passionate bloggers.
Data journalism, however, is not limited to hip data visualizations. Another prominent example is Nate Silver’s blog with the New York Times, FiveThirtyEight, which famously predicted the results of the 2012 U.S. election with remarkably little error. Although Silver primarily tackles politics, he makes forays into sports (see his latest on NCAA brackets) and the arts (Oscar predictions). Silver’s work is somewhat unique for a news outlet because it’s coverage of news before it happens, but really it’s a kind of 21st-century version of news analysis—one dependent on numbers and a deep understanding of statistics—that breaks its own headlines often enough to demonstrate the power of these techniques and keep up a healthy readership.
What does data journalism mean for science reporters?
You’ll note that all of the examples I’ve just mentioned don’t involve science topics. In fact, a lot of data journalism skews toward the social sciences, especially data-rich subjects like economics, or virtually any governmentally funded project, since legislation often requires data reporting.
Of course, science is also data-rich. The UK’s Guardian, which has fast become a leader in data journalism (and visualizations), has used data to cover global cancer rates and climate change, and answer questions like how much sunscreen should be worn to protect against skin cancer. (The latter is a good example of a journalist—much like a scientist—starting with a question, then digging up the data to answer it, and putting it into an attractive, digestible package.)
But for many stories, the best approach remains face-to-face conversations with scientists in laboratories or in the field. One problem preventing fertile crossover is that most science data is privately held until publication, and considerably more esoteric than basic governmental stats—even for trained molecular biologists like myself. Scientists are still the most reliable harbingers of what will be ‘hot’ in the next year, and personal contacts with people working on the most exciting projects are a better ticket to a great story than the most sophisticated statistician wielding the most novel algorithms.
Budding science journalists should nevertheless be aware of this current trend, and it can’t hurt to become familiar with many data journalism tools, especially for making simple, effective graphics (and knowing what to put in words vs. what to keep in an image). For those of us with advanced science degrees that somehow managed to avoid high-level statistics and computer programming, there is still hope: the Internet.
Frankly, I’ve wanted to augment my skills in these areas for some time, and knowing they will likely be put to good use in my future career is enough to convince me that now is the time to bulk up on my stats knowledge and learn basic programming (I’m starting here, here, and here). If my Ph.D. has taught me anything, it’s that learning never really stops.
For more information on data journalism, check out these great resources:
Geoff McGhee’s Stanford video project, ‘Journalism in the Age of Data’
Overview of the Guardian’s approach to Data Journalism