Reporting

Was Oakland the nation’s most dangerous city in 2013? Or was it Oakland and Flint? What is a valid distinction, statistically speaking? We show the uses of a “confidence interval.”

A wealth of data is released every week in the United States by organizations of all kinds, from federal, state and local agencies to companies, educational institutions and other nonprofits.

Many large federal datasets contain highly granular statistics that can serve as a launching point for local stories — for example, the location of alternative-fuel stations (graphic at left), campus crimes or payments to doctors by pharmaceutical and medical device companies.

The notion that journalists are math phobic and prone to error when working with numbers is an age-old cliché, embodied in the quip that, for many in the media, the plural of anecdote is data.

The notion that journalism should become more data-driven — and get a little closer to social science — is not a completely new idea. The journalistic sub-field of computer-assisted reporting, embodied in the work of the Investigative Reporters and Editors’ NICAR program, has a long history.

While data journalism can take many forms, mapping is perhaps the most basic and accessible way to create meaning graphically to help audiences understand stories and issues in new ways. It is, of course, one of the oldest forms of data visualization, with a history that stretches back centuries, even millennia.

This post was adapted from VerificationJunkie, a site created by Josh Stearns of the Geraldine R. Dodge Foundation. @jcstearns

Trying to stay informed about the flurry of changes and shifts within the news media business is no easy task. From the decline in advertising and shifts in ad tech to new digital-ethical puzzles and social media strategies, there are many emerging topics — and many outlets and voices, both new and old, digging into the latest trends.