woensdag 27 mei 2015

Are Coders Taking Over

(This article is an edited version of the speech given at the Josh Friedman price ceremony at GIPA, Tbilisi, Georgia, June 9 2015; published at Memeburn  http://memeburn.com/2015/06/has-coding-become-journalisms-most-important-skill/)

Journalism will never be as before. The rise of internet not only changed the business model for media but also the content. Data is the new content. Of course journalism is still about storytelling, but it is changing fast in another direction. Drones- drone journalism- are now almost an accepted tool for reporting. Computers could be instructed to write basic stories: robot journalism. Take for example a story from NY Times Upshot about the best and worst places to grow up in the US and notice how the text changes according to the data selected. 

The latest development is sensor journalism. Buy an Arduino board and a few sensors, write a program for the processor and you are in business. An interesting example is the air quality in San Diego, where students of the University measured the air pollution in different parts of the city. Access to water points in Tanzania or water pollution in streams and dams used as potable water in South Africa are other interesting try outs.  It is undeniable that the influence of technology on journalism especially IT and coding is growing. Where is data journalism going?

To answer this question it is important to focus on the development of data journalism. Data journalism originates from journalists who were using computer based tools of the social sciences to enhance their reporting. Philip Meyer was one of the first to do a survey and use the results in a news story. This line in the development was called CAR-Computer Assisted Reporting. 
The essential characteristics of CAR are:
-        -  Story is central and produced according to journalistic professional standards;
-       -   Stories were investigative;
-        -  Samples were used in surveys to test hypotheses;
-        - Data were not published only the results were used;
-       -   Public was not involved in research and publishing.

So CAR marries social science with journalism. The next development[1] comes from the open source movement and open data movement. For example code4africa and the African Media Initiative (AMI) directed by Justin Arenstein.  This new step brings journalism more under the influence of computer sciences and big data. So the paradigm also changes:
-         - The focus is less on the story but on the data: datasets and analysis are published   together with infographics (for example D3: data driven documents);
-        - The public is involved in getting data or analyzing data;
-        -  Data are complete sets and not samples;
-        -  Use of code and algorithms for analysis, for example Python and R are becoming popular.

Coders taking over 
This means that data journalism is moving away from the social science based investigations and storytelling, towards data collecting, analysis and publishing based on code, algorithms and software form the computer sciences. Are we going to: more IT less journalism, less professional standards and more open network production with public participation? Are the coders taking over  the newsroom?

I don’t think so. Coders are offering their knowledge skills outside the newsroom to journalists. There are lots of services for data journalists available: for example for making graphs, maps or combinations but also for analysis for social networks for example. For example: datawrapper, many eyes, cartodb or Tableau. Many Dutch newspapers are using LocalFocus for a quick visualization of their data. 

Secondly, coders but also former journalists are offering their coding and data journalism skills to the media for enhanced data journalism project. These are all new start-ups, trying to create a market position for data journalism.

Only a small number of coders and developers is working directly in the newsroom, generally of rich media. If these media are regularly do data journalism, it pays off to hire a coder and train them in journalism; at least it is cheaper and easier than training a journalist in computer science.

Pointless Job Title
What is left for the data journalists in the newsroom? You can outsource your data project. Or the journalist dumps the data in the container of a data service and collects the result. Duc Quang Nguyen writes in his article ‘Data Journalist a Pointless Job Title’: “The rise of simple free interactive charting solutions … has considerably democratized data visualization. What used to be performed only by data journalists is now more commonly a standard skill among digital journalists.” 

In the digital era journalism has to re-invent itself. This not done through outsourcing or adding new specialists to the newsroom and running the risk of making data journalism a pointless job title. Journalists should be basically trained in the new data journalism tools in order to do the easier data projects themselves and cooperate with coders and developers in more demanding researches.

[1] For a typology of data journalism: Mark Coddington, Clarifying Journalism’s Quantitative Turn. A typology for evaluating data journalism, computational journalism, and computer-assisted reporting.
In: Digital Journalism Volume 3, Issue 3, 2015
Special Issue:   Journalism in an Era of Big Data: Cases, Concepts, and Critiques

Geen opmerkingen:

Een reactie posten

Opmerking: Alleen leden van deze blog kunnen een reactie posten.