Data Journalism 101: Tips and Tricks

8.04.data1

“Anybody can do data journalism,” said Bahareh Heravi from the National University of Ireland during the workshop “Tips and tricks for data-driven journalism starters” at the International Journalism Festival on 7 April, 2016.

Heravi runs Hacks/Hackers Dublin – an international organization bringing together  journalists (“hacks”) and technologists (“hackers”) to create and establish a new path for news. The workshop covered questions regading the nature of data journalism and the process that should be followed to become a good data journalist, an explanation supported by the presentation of case studies and tools for the job.

What is data journalism?

“There are different definitions about (sic) data journalism, but maybe – said Heravi – the most popular and simple one is ‘Telling stories with data’ or ‘Telling (journalistic) stories with data.” She noted that today’s increased data availability is one of the reasons why data journalism has become increasingly more and more popular.

Using the DIKW model, the panellist explained the hierarchical order multimedia professionals are met with when discussing data: “Data on its own doesn’t have much meaning – it’s just a piece of information with no relationship with any other piece of data, or contens. When we have some relations between our different pieces of data, then we have information,” said Heravi. Using the trade’s 5W1H – what, where, when, who, why and how – journalists achieve the upgrade to the higher ranks of Information and Knowledge. “When we understand the principles between or within our knowledge [of data] then we have Wisdom,” continued the panellist.

According to Heravi, the job of data journalism is to establish a data pattern to construct a in depth story, hence journalists have to “lift data from the data level to as high a level [of potential] as possible,” she explained. To demonstrate the important role data has played in reporting past events, the panellist examined two graphs – Napoleon’s march on Moscow by Charles Joseph Minard, and Florence Nightingale’s coxcomb account of soldiers’ death causes in Crimea.

How to run a data journalism project?

Journalism is a highly competitive trade, and the professionals, both contemporary and future, need to become data savvy. Journalists should not only rely on readily available data: instead, they should be able to locate and collect usable information themselves, followed by a thorough process of cleaning and analysing the data to better frame the story before finally putting ink to paper. Heravi then gave a brief account of international publications that over the past ten years have embraced the resourcefullness of data and embedded it within the traditional journalistic structure: among the most notable are The Guardian, the Financial Times and The Irish Times

Types of data analysis

Media professionals embarking in data journalism have at their disposal not only several tools, but also an array of types of analysys they can adopt when working with data:

i. Temporal data – focused on the ‘when’,  it looks at the temporal distribution of one or more data sets – such as the identification of growth rate, latency of peak times or decay rates; or seeking patterns in time-series data, such as seasonality or bursts;

ii. Geospatial data – uses location information to identify positions, movements trends or patterns over geographical space;

iii. Topical data – uses text to identify major topics, their interrelations and their evolution over time and space;

iv. Network data – identifies connected entities and the relations between them, network properties such as size and density, structure such as clusters and backbones.

Following the analysis type overview, Heravi proceded to categorise the various data tools widely available to data journalists, and provided a list of official and/or trustworthy data providers – such as OECD.Stat and Eurostat – and data portals – like publicdata.eu and Google Public Data.