Data journalism is a term that is used to refer to the increased role that numerical data has in journalism. It brings together computing, technology, statistics and journalism. It is a relatively new term and has become more important as technology and computing has grown. It was just part of investigative journalism before.
The extent to which data journalism is applied is cultural. But it is not just impacted by what is written but how data and statistics are gathered in the first place. This is what this article explains.
Teaching data journalism in developing countries: lessons from ODECA
By Eva Constantaras
Eva Constantaras is a data journalist and trainer who recently wrote the Data Journalism Manual for the UN Development Program. In a special guest post she talks about the background to the manual, her experiences in working with journalists and professors who want to introduce data journalism techniques in developing nations, and why the biggest challenges not technological, but cultural.
Over the last few years, there has been a significant shift in global experiments in data journalism education away from short term activities like boot camps and hackathons to more sustained and sustainable interventions including fellowships and institutes.
There is a growing awareness that the challenge of teaching data journalism in many countries is split straight down the middle between teaching data and teaching journalism — where neither data science nor public interest journalism are particularly common. Open data can be a boon to democracy — but only if there are professionals capable and motivated to transform that data into information for the public.
As Abhi Nemani, a veteran of government open data projects put it succinctly:
“The streets of Heroku are paved with the code of unused dashboards.”
Many attempts to bring data dashboards and citizens together have underestimated the gap between the two.
Bringing data journalism into the classroom
One of the most promising areas of growth has been bringing data journalism into the classroom, to redefine journalism for a new generation.
If data journalism is mainstreamed into journalism programs, that accelerates the process of "data journalism just becoming journalism" as championed by Simon Rogers.
The United Nations Development Program (UNDP) Istanbul Regional Hub has published a Data Journalism Manual on its Open Data in Europe and Central Asia (ODECA) platform in English and Russian to accelerate the adoption of data journalism.
The ODECA initiative supports government representatives, civil society activists, tech activists and citizens that care about and work with open data.
The manual forms the basis of a 500-hour training program being rolled out in several universities and journalism training institutes in the region.
The content is designed to instill the skills and level of data literacy necessary to produce both short data stories on a deadline as well as longer in-depth system stories that explain underlying development challenges facing the region through data and digital storytelling.
The audience for the manual is data journalists seeking to teach skills to others and journalism instructors seeking to grow in the area of data journalism.
Emerging models for data journalism pedagogy
Experiments in universities are being well documented both at an institutional level in reports such as Teaching Data and Computational Journalism by researchers from Columbia Journalism School and Stanford University, and more informally by data journalist academic hybrids such as Dan Nguyen and Jonathan Stray.
These constantly evolving curricula seek to balance the cultivation of critical thinking skills, research methods, software, coding and storytelling in a coherent course that produces students with the mindset and skills to integrate data into their storytelling.
Often, these multidisciplinary programs pull in expertise from across journalism, computer science, mathematics, design and other faculty.
Through trial and error and a whole lot of borrowing from open source examples, I’ve written, refined, rewritten, overhauled and adapted a data journalism curriculum based on experiences with journalists in Kenya, Afghanistan, Sri Lanka, Palestine, Myanmar, Sudan, Armenia, among others.
The goal is simple: a formula to maximize the chances that public interest data journalism takes root in mainstream media in these countries.
Exporting data journalism curricula to developing countries
Many donors, concerned with sustainability, replicability and the 'multiplier effect' have requested a training-of-trainers format for data journalism education, including at a university level.
This overlooks the key factors for the success of such programs in the few universities that actually have them: data journalists who have moved into renowned university journalism programs to lead them.
Launching a data journalism program in a university in a country with few or any data journalists and weak media departments implies teaching data journalism without data journalists.
Imagine a medical school without any doctors on staff, without the risk of such calamitous results.
When the United Nations Development Programme in Europe and Central Asia proposed a 10-day course for university professors from the region, their purpose was not a pipe dream of transforming these universities into data journalism powerhouses but rather to answer three key questions:
- Would the heads of communications and computer science departments buy into the data journalism idea?
- What resources would be required to set up data journalism courses or programs?
- What barriers will they encounter along the way?
The biggest challenge isn’t tech, but critical thinking
After 10 days the participants produced six stories and 12 visualizations but more importantly, the participants provided the answers to all these questions.
Teams of communications and computer science faculty as well as instructors at journalism training institutes from Albania, Belarus, Kyrgyzstan, Moldova, Montenegro, Serbia, Ukraine and Uzbekistan produced data journalism as well as a roadmap for implementing data journalism at the university level.
Data journalism can provide the impetus for pivoting towards public interest reporting
Before the course began, professors identified the biggest challenges facing their journalism students. Surprisingly, the top two had nothing to do with data or technology but critical thinking: the ability to develop public interest angles for their stories and to organise multiple sources of information.
Overwhelmingly, the professors agreed that the biggest value in adding a data journalism program would be to improve the overall quality of news and information in the country, so the course curriculum emphasises an inquiry-based approach to data journalism that prioritises answering important questions over data for the sake of data.
Elira Turdubaeva, Head of Journalism and Mass Communications Department at the American University of Central Asia in Kyrgystan, revealed a maternal healthcare crisis in her story.
It used evidence to demonstrate Kyrgystan’s backslide in maternal mortality compared to the region and the growing divide between safe pregnancy in urban and rural areas.
She said such stories rarely come out in the local media, and she envisions swapping out the traditional reporting model of just writing down what is happening as it happens to meeting citizens’ information needs through data journalism.
Inter-departmental collaboration is key
In countries with robust investigative journalism programs underway, participants saw the value in bringing together computer science faculty and media faculty for joint programs along the US model.
Cornelia Cozonac, Director of the Investigative Journalism Center (CIJM) and lecturer at the Free International University of Moldova, and Georgeta Stepanov, Dean of the Journalism Department at Moldova State University, both teach investigative journalism. During the workshop they harnessed open education data in Moldova to demonstrate that parents were paying unnecessary school fees over and above the official school budget.
They see the value in introducing a semester-long course team taught by investigative journalists and IT specialists. The incentive for students would be their portfolio of data-driven projects that they could use to get jobs in mainstream news organisations.
In Kyrgystan, the communications faculty has already implemented a major course change for its students, swapping out math (Contemporary Math and Math for Life), which were basic requirements, for Theory of Probability and Statistics; Quantitative Methods and Computer Science to lay the foundations for a data journalism program.
A new summer institute to teach journalism and computer science faculty the basics of data journalism through the new curriculum starts this year.
Data access is not as big a problem as anticipated — but technology might be
In countries without access to information laws, professors worried about finding enough data for students to pursue independent projects.
But as the Belarus team discovered, on small topics (in their case, high rates of alcohol consumption in the country) can yield enough subtopics with related data sets to keep an entire class busy for a semester.
Students could investigate everything from liquor taxes and lost productivity due to alcoholism to the demographics of alcohol-related mortality and changes in drinking habits.
Though all would prefer local data sources, in the absence of open data journalism students could begin building skills using international data sets about their countries and regions.
Both Belarus and Moldova identified lack of easy access to computers and technology as a major hurdle to students being able to practice data journalism skills after class.
Professors would need significant training in preparation
Despite an 80-hour training schedule, a complete data journalism training manual and the production of their first stories, the participants universally agreed that more training would be required to prepare themselves and other professors to teach this course.
Ervin Goci, a professor in the Department of Journalism and Communication at the University of Tirana, explained that the idea of data journalism in the country was so new that it would require recruiting younger professors interested in a hands-on approach to teaching to engage in a professional development program.
Several participants proposed a summer program for professors from different departments to come together to both master the training manual and produce their own stories, a program to be launched in the coming months.
Scalable models could be tailored to different universities
Bringing together the heads of communications faculty smoothed over barriers to introducing data journalism into the curriculum, but also brought out all the special accommodations that would be needed for data journalism to be successfully integrated.
In freer media environments, participants were eager to introduce undergraduate courses, develop master’s level programs and create concentrations and continuing education programs.
In more closed environments, faculty suggested a more low profile approach to creating a working group on the topic and running pilot classes on Data for Storytelling to gauge the reception. The ODECA data journalism manual will be the cornerstone to both approaches.