This post is part of the official Global Investigative Journalism Conference 2017 coverage. IJEC organized and covered the academic track, which consisted of papers on investigative and data journalism and discussions of best practices in teaching.
During the Academic Track Lightning Round on Saturday, Nov. 18th, 10 professors and trainers from around the world offered tips and techniques on teaching data journalism.
Data Journalism Lecturer Pinar Dag works at Kadir Has University in Turkey. She started her presentation with advice on how to adjust lectures to the audience. When teaching engineers, for example, an instructor might need a separate dataset more pertinent to that profession.
Dag’s students. Picture: from presentation slides.
Dag’s main tip was to “show, don’t tell.” She said it is better to show students how data stories were made rather than talking about how data reporting is done.
“Explain clearly the relationship between data and journalism with data journalism news,” Dag’s slides said. “Define data journalism with case studies, practice a lot and show more!”
Both for future teaching reference and to show trainees good practice, it is important to make the data collected in class accessible, structure it, and share the cleaning process and analysis on Github, Dag recommended.
Dag said that it is best to start with small datasets so students get comfortable with numbers and to make clear what role the data will play in the stories – whether the data is the basis for the story, the story explains the data, or if the data merely enriches the story. She also said to follow the Data Literacy Matrix.
In conclusion, Dag reminded the audience to stay up-to-date with the latest data tools and to remind students not to get lost in the data but rather be focused on writing remarkable stories. One way to keep them focused, she said, is by showing examples of great data-based investigations.
Reach students through their stomachs
Associate Professor and investigative journalist Jeff Kelly Lowenstein from the U.S. talked about the kind of dataset that is most popular among his students: food inspections.
The Grand Valley State lecturer also stressed the importance of cleaning data and knowing what you are looking for when analyzing it.
Denise Malan, who is a training director at Investigative Reporters and Editors, showed a collection of fun data to use when teaching.
Malan showed sets about the price of eels, satellites, dogs in New York City, soccer scores since 1872 and referred to Kaggle, a data science community also highlighted during Pinar Dag’s presentation.
The datasets Malan referred to in her slides can be found at the National Institute for Computer-Assisted Reporting website:
http://data.nicar.org/training/fundata/
Training as a hike
University of Illinois professor Brant Houston and author of four editions of “Computer-Asssisted Reporting: A Practical Guide,” started his presentation by explaining the usefulness of data in journalism.
Journalists use data to go beyond the anecdotal, to find patterns and trends, to find outliers and to present stories in more comprehensible, compelling and credible ways, Houston explained.
When preparing student training, start with a survey of their skill level, experience and expectations, Houston said. Assess their technical needs, select software based on location, create a realistic syllabus with examples and select data that is relevant to them.
During student training, make sure all the technology has been tested and the room layout fits the training. Start with introductions and with a slow pace. Houston advised to “show first and then instruct.” One way to bring these two aspects together is to start with a test exercise.
When preparing training for professionals, there are different aspects to consider.
Houston, who has been teaching data journalism for 30 years, advised to share proposed programs with the organizers and share the program with the participants beforehand. He also said to ask the participants to do reading – but not to expect all of them to do so, and to consider establishing a newsroom culture during the training.
When it comes to the training itself, there are some points Houston addressed that are applicable for students as well as journalists.
A crucial point is to deal with some journalists’ “fear of data and math,” Houston said. He said consider preparing the students for learning through trial and error. In addition, he suggested starting from a very basic dataset that has three columns and five rows, and to perhaps do data management such as sorting before teaching calculations.
In general, it is important to summarize during classes and to make aspiring data journalists feel confident. Always celebrate achievement and focus on quick early victories, Houston said.
“Training is like going on a hike,” Houston said. “People walk at a different pace, you stop periodically and make sure get everyone together at the same place in learning.”
Get your students to teach you, have fun, use humor and do your own learning every year, Houston said.
As a trainer, it is important to find your style and voice, though you should borrow from the best.
Houston concluded that “teaching is discovery and there are many paths to the palace of wisdom.”
Teaching the value of data
Lailah Ryklief trains, mentors and develops curricula for OpenUp’s Data Literacy Program in South Africa. Her presentation was focused on the importance of getting journalists to appreciate the value of data reporting because it can be slow, challenging work, requires learning new skills and takes journalists out of their comfort zone.
A trainer needs to “create opportunities for trainees to realize the value of data journalism for themselves,” Ryklief said. “This is probably the most important driving factor behind impactful training.”
To help them realize this value, trainers should introduce students to the online data community.
“This community exists largely online, where they exchange their resources, methodologies, ethics, and experiences,” Ryklief said. “Data journalism can thrive from this dedication toward pursuing a culture of openness and collaboration around data.”
For the training to have a lasting impact, trainers should teach journalists how they can bring data into their everyday process, Ryklief advised.
Instruct journalists to start with small datasets and to reflect on a few aspects of their work. Trainees should identify daily habits and see if there is a tool that can make them easier, make sure they are using online resources to their full potential, and see data as a tool to enrich what they are already saying.
Dealing with tough situations during training
Crina Boros is a data journalist at the Thomson Reuters Foundation in London. She also works as a freelance trainer for organizations such as the BBC and Greenpeace. Boros’ presentation focused on what can go wrong during training and how to approach various kinds of students.
Some trainees can be skeptical about computer-assisted reporting. Boros advised to use strong data stories and examples to win them over.
One example, she said, is to show the power of a pivot table to analyze spending data of a political party. Another example Boros cited was an interactive map on electoral constituency in relation to poverty and party dominance.
Boros also shared pointers on working with introverts: use humor, understand their anxieties and reluctance to talk, make them comfortable, fake a newsroom environment, demystify jargon, let them see you fail and solve error messages together.
For shy trainees, Boros said to discriminate them positively, create room for them to contribute, use eye contact and talk in a soft manner.
On the other end of the spectrum are “over-assertive” trainees. Here it is important to avoid eye contact in key moments, not be intimidated by their intellect and talk up the contribution of others, Boros said.
Boros finished her presentation with two slides on “weirdos” and “xenophobes” as types of students.
When confronted with trainees who have outsider or extremist views, teachers should not express political affinities, set clear boundaries, ignore ramblings, make them work in groups, and not feel obligated to answer their calls or correspondence.
When trainees make prejudiced comments, Boros advised to make clear, immediate, polite and firm remarks stating that you do not endorse those comments. A teacher can also point out how data challenges prejudice and indicate that opinion journalism is not a part of the curriculum.
Data analysis challenge
Jennifer LaFleur is data editor at the Investigative Reporting Workshop in Washington D.C. and became the first full-time training director for the National Institute for Computer-Assisted Reporting in 1994. LaFleur, who organized the lightning round sessions, has trained journalists throughout the world while also doing and overseeing investigative stories that extensively use data. Her presentation focused on using two tools, Microsoft Excel and SQL (Structured Query Language), to teach data analysis.
LaFleur’s approach is to create a query challenge.
“Break into small groups and come up with a question to challenge the other groups,” her slides said. “Have groups pick questions and then use those to develop test questions.”
By using excel and SQL basics to solve the query challenges, trainees will learn how to analyze data.
The Excel basics comprise basic math operations such as subtraction, percentage change, sorting, summing, rates, ranking, filtering and pivot tables.
LaFleur recommended using SQLite and teaching the select, where, group by and order by commands.
LaFleur ended with a song explaining SQL basics to the tune of “Doe, a deer.”
“First SELECT the fields you want
FROM the table where they live
WHERE The way to filter down
Now your query’s growing fast
GROUP (BY) to aggregate your fields
ORDER (BY) to run them big to small
JOIN to match your data set
And hit RUN to get results, ho, ho, ho.”