Digging Deep for News: New Approaches to Environmental Reporting

Dr. Jonathan Stray
Overview Project, Columbia University

Although journalists are trained to become rapid, temporary experts in a wide range of topics, researching and reporting stories can sometimes require complex and specialized skills, such as sifting through mountains of data. For the reporters who do this, locating and analyzing data in all its forms (from text or audio to numbers and surveys) and conveying that information in a meaningful way to the public are daunting challenges. These three activities are integral to how Jonathan Stray, an associate adjunct professor at the Columbia University Graduate School of Journalism and project lead on the Overview Project, defined data journalism–a niche specialty within journalism–in his talk during the Metcalf Institute’s 2013 annual public lecture series at the University of Rhode Island Graduate School of Oceanography on June 14, 2013. He further depicted data journalism as a combination of journalism, computer science, statistics, and ethics – skills that are essential to effectively obtain data and report about it in a way that interests news audiences.

Stray pointed out that the data used in reporting sometimes tells a one-size-fits-all story, as in the case of a New York Times story from 2009 showing the overall U.S. unemployment rate (http://www.nytimes.com/interactive/2009/11/06/business/economy/unemployment-lines.html). Stray noted that this rate was a national average across all races, genders, age groups, and education levels, but further digging into the interactive database created by The New York Times staff shows how unemployment may differ significantly between demographic groups.

Stray also stressed the importance of public perception of information, pointing out that when most people think of crime, they think of violent crime, despite the fact that the vast majority of crimes committed in the U.S. are property-related. Data on the occurrence of specific types of crime, therefore, are critical to complete understanding of an issue, and should be carefully considered by journalists when they report a story.

So where should journalists turn to find these data sources? And with a mind-boggling amount of documentation available, what steps should a journalist take to whittle down the information to a manageable size?

These are some of the questions that Stray’s venture, the Overview Project (https://www.overviewproject.org), helps journalists and the general public accomplish. The Overview Project sorts documents into categories, subdividing search results into different subfolders grouped by topic. These folders can be further subdivided or regrouped by different topics or key words. This tool enables journalists to find patterns in large quantities of text data and potentially extract multiple story ideas from the same database. Stray further demonstrated how the Overview Project can be used to assess what happens after a story is published. In other words, by monitoring how news consumers use a news website, it is possible to examine the mechanisms controlling which stories are encountered by a reader, and which of them grab the reader’s attention. It is also possible to observe how information flows by monitoring the frequency of search terms or phrases in the news over time. Looking at data in this way can show how different users are attracted to specific types of news stories.

Stray outlined a few key characteristics and skills of a data journalist. He emphasized that no matter how much data a journalist has at his or her disposal, context is critical to conveying meaning and importance because the data do not speak for themselves. Also, the data journalist must develop the computer science skills necessary to identify the most meaningful presentation of data. In short, Stray argued that a good data journalist is able to comprehend what all the data means and convey that effectively and meaningfully to the public.

Lecture summary prepared by Al Nyack, Ph.D.

Easy Sign Up