Tell me more about this “data journalismDoes that mean that every journalist should learn to write computer code?
There's little doubt that "data journalism" has become one of the most important subfields of journalism in the past ten years, and even less doubt that this will be a major journalistic growth area in the future. Data journalism might be defined as the application of statistical techniques to the analysis of diverse evidentiary sources such as databases, opinion surveys, and government records, and the subsequent crafting of narratives that stem from this analysis. In other words, data journalism treats data as a kind of journalistic "source," on par with other more traditional journalistic sources like documents, interviews, and direct observations. Practical applications of data journalism actually precede the emergence of the Internet, although one thing that makes today's data journalism unique is the growth of interactivity and the use of open source documents and tools.
Some of the earliest modern applications of data-driven statistical techniques to news reporting can be found in Philip Meyer's work, elaborated in the book Precision Journalism published by Indiana University Press in 1973. In it, Meyer urges his readers to "go beyond the anecdotal" in their practice of journalism, using coverage of the Detroit riots of 1967 as an example. Many of the journalists that attempted to explain the riots relied on traditional reporting techniques to gather their evidence, including "man on the street" interviews and interviews with protest leaders. They also began their stories by largely embracing "common sense wisdom" about why the riots occurred. Meyer, on the other hand, conducted representative surveys of city residents to accompany a series of stories about the state of Detroit in 1967. These surveys revealed that unrest was driven by what sociologists have called feelings of "relative deprivation" and a sense that while life in Detroit had actually improved in the 1960s it had not improved for everyone and had not improved quickly enough for most African Americans relative to other groups. Meyer's findings also showed that the rioters were a specific subgroup and did not reflect the overall attitudes of the area's African American residents.
Computer-Assisted Reporting (or CAR) was a new journalistic technique to emerge from Meyer's work. In the 1980s and 1990s, journalists interested in generating stories from data and using data to shed light on news developments were increasingly using computers, both to access data sets and to carry out the number crunching required to turn this information into narratives. Prominent examples of computer assisted, data-driven reporting included a 1969 Miami Herald analysis that used a computer to uncover patterns in the criminal justice system; a 1972 New York Times story that looked at discrepancies in crime rates reported by the police; and a 1988 Pulitzer Prize-winning investigation called "The Color of Money," which dealt with redlining in middle- class black neighborhoods. In 1989, Investigative Reporters and Editors founded NICAR, the National Institute for Computer-Assisted Reporting. As should be obvious from its name, CAR emphasized the technology that lay behind the data—computers—more than the original concept of precision journalism, which was more philosophical in its argument that journalists should use social scientific technique regardless of the tools needed to do it. And it should be obvious that while many data journalists are indeed computer programmers, not all are even today, and historically very few of them have been.
It's possible, however, that this is changing. Today (and increasingly so in the future) we can expect data journalism to emphasize interactivity (the ability of news consumers themselves to "play" with journalistic data, to personalize it, to visualize it in different ways, and so on) and transparency (making the data sets that lie at the core of data journalism open to analysis by the wider public or by other researchers and journalists). Both of these skill sets are greatly aided by a facility with software languages and computer programming.