/// Data Journalism is improving — fast
Data Journalism is thriving. This the most salient conclusion from the second edition of the DJ Awards organized by the Global Editors Network and sponsored by Google. I was part of a 20 persons jury, chaired by Paul Steiger, founder of Pro Publica. We had to choose among a short list of 72 projects divided into seven categories: data-driven storytelling, investigation, applications (all three for large and small media), and data-journalism section or website.
Here are some quick personal findings.
#1: Data-journalism is a powerful storytelling tool. The Guardian won the Storytelling Big Media category, with this compelling graphic showing the situation of gay rights for each state of the US. It did so by analyzing a range of stats and administrative rules or laws such as hospital visits, adoption, schools or housing. (In half of US states, gays have no clearly stated rights). On that matter, no story could have spoken more loudly.
In a different way, Thomson Reuters collected another prize for its amazing Connecting China project that looks like a visual LinkedIn for the PRC elite. It’s a huge, 18 months endeavor, built on more than 30,000 connections between Chinese power players.
#2: Data-journalism extends well beyond the usual economical/social topics. One DJA 2013 laureate displayed the explanatory power of good data-journalism. The French site Quoi? explored aspects of the art market. In its Art Market for Dummies (available both in French and in English), Quoi? explains who are the most bankable artists (since 2008, it’s Picasso, Warhol, Zhang Da Qian); it also shows why it is a terribly dead-male-dominated business; and it illustrates the rise of Chinese artists. It’s both entertaining and information-rich.
Another French company, WeDoData collected an award for a great app showing the (terrible) state of female/male parity in France, in a smart, user-friendly package commissioned by France Television.
Another great example of clever data journalism expanding to society issues is the Great British Class Calculator presented by the BBC (it won the Data-driven apps category). The project started with a survey of 161,000 persons, conducted with several universities. This helped define seven social classes ranging from the Elite, to Precarious Proleteriat, or more imaginative New Affluent workers or Technical Middle Class.
#3: Tools can be surprisingly simple. In many instances, data-collection and analysis are performed using relatively simple tools such as large Excel or Google Docs spreadsheets (the latter being excellent at scraping data repositories — just google the terms to find tons of resources on how to use those. The Argentina newspaper La Nación, winner of the Data-driven Investigations Big Media category, explained in its DJA filings how it retrieved 33,000 records showing the expenses of senate members by using sets of Excel macro commands.
For its Art Market for Dummies project, the French multimedia journalist Jean Abbiateci explained how he scraped the ArtPrice database (links are mine):
For scraping data, I used Outwit, a amazing Firefox Add-on. This is very useful to convert a pdf file to an Excel file. I used Google to refine, to clean and merge my dataset. I used the Google API Currency Converter for my uniform monetary values. Finally, I used D3.js and Hichcharts.js. I also reused open source code shared by Minnpost and a software developer called Jim Vallandingham.
The projects mentioned above are just examples. A visit to the GEN Data Journalism section is well worth your time. For once, the digital news sector has fostered a healthy, creative segment, one that relies a lot on small agile companies. I find that quite encouraging.
Monday Note – Frederic Filloux