Panama Papers has been hailed as an unprecedented reporting initiative in more than one way. The biggest leak and the largest collaboration in the history of journalism involved more than 370 journalists, who analysed the 11.5 million documents that exposed the offshore holdings of political leaders and prominent public figures.
More than 100 news outlets in some 80 countries worked on the project, coordinated by the International Consortium of Investigative Journalists (ICIJ). The only academic institution to take part was Columbia University's Graduate School of Journalism.
This gave a group of five students on the MSc Journalism (Data Concentration) programme a unique opportunity to be part of a cross-border investigation and practice the skills learned during their degree.
Professor Giannina Segnini, who led and advised the team, told Journalism.co.uk in a recent podcast the students helped ICIJ reporters uncover stories related to shipping and trade issues, by scraping and analysing data from more than 16 databases and cross-referencing it with information from the Panama Papers, while simultaneously working on their individual final projects.
"Students needed to have the capacity to process the data by cross-referencing and cleaning it, and once they managed to do that, they had to find a story, which was quite challenging because there were many other journalists working on similar issues at the same time," she explained.
The Panama Papers interactive database is available online for anyone to browse, alongside many other public datasets that remain unexplored, so Segnini pointed out "there is no limit" for students who want to try their hand at investigative journalism.
"This is just starting and there is a lot out there that hasn't been investigated, so try to challenge yourself first by working on a project that you are passionate about and look for guidance," she advised.
We reached out to two of the students who were part of the team at Columbia University: Emma Paoli, who is now completing an MA in Ethnographic and Documentary Film at University College London (UCL) and works as a freelancer for French newspaper Le Monde, and Jianghanhan Li, who is pursuing the second part of her Columbia degree in Quantitative Methods in the Social Sciences, to find out what they learned from the project and what tips they can offer to student journalists looking to work on similar investigations.
On getting started and finding a story
Paoli, who had worked on a few investigations for Le Monde previously but on a much smaller scale, said data skills are important in a project of this nature, but they are not everything.
"What you need more than data skills is to understand the big picture.
"Before you start looking at the leaks and trying to search specifically for offshore firms, for example, you need to understand what is Mossack Fonseca, what is an offshore firm, what is the legislation attached to it?"
For her final project at Columbia, Paoli had to choose a country that was not being covered by ICIJ reporters, so she picked Uzbekistan, but because she was not familiar with the region and due to the large number of documents available in the database, finding a story took her longer than anticipated.
"Before you pick a story, make sure you will be able to follow through. Do you speak the language of that country, do you know which sources to contact and how to get in touch with them, are you going to translate all the documents you find that are linked to it?
"In my case, Uzbekistan was like a completely black hole in some ways, there weren't that many journalists or researchers looking into it, so this made investigating difficult."
On the skills needed, both journalistic and technical
Li, who had no background in journalism before starting her degree, pointed out research skills are essential in this type of reporting, as "curiosity leads you to figure things out and then go further". However, knowing when to stop digging is as important, because "the story can go anywhere and there's no end to the discovery process".
But depending on the story chosen and what other materials a student might have to use as part of their investigation, specialised skills such as programming or statistical analysis could also come in handy, she added, as was the case for her story on oil trade.
"I used the United Nations Comtrade database, which involves a lot of statistics. So using Tableau and R, I analysed oil trade in Turkey, UAE and Iran across the last two decades, looking at the discrepancies between export and import to show how much oil was lost along the way.
"Because it's a data journalism investigation, you have to use data to unearth interesting stories or anomalies, so for that part, my R and Tableau skills helped a lot."
As the database was structured in a way that made it easy to search for names or keywords to reveal the documents and emails related to a person or a company, most of Li's colleagues only needed to be comfortable with cleaning the data and using spreadsheets to organise the information.
"You really need to be extremely organised and while it may sound a bit obvious, you should set up a spreadsheet from the beginning, with a clear structure, to collect all the information," Paoli said.
"Then you can start finding patterns, trends, or you can start making links between companies."
On the importance of having a human element in the story
Halfway through the process after she had gained access to the Panama Papers documents, Li realised the data itself would not lead to an interesting story unless it was accompanied by a human element. This meant reaching out to experts that could help put the research into context and identify the impact an issue had in a particular region, as well as finding sources to interview for the story later on.
"If you have the chance to talk to an expert who already knows about some of the problems in that area, that could be an easier way to pinpoint the problem, rather than you reading all these materials yourself and finding something on your own.
"For example with my oil story, there was a problem because Turkey wasn't reporting their oil import and export for the last five years, and I found out they had government legislation allowing them to do so, but I was lucky because as I was looking at this, a new story was published explaining these deals.
"I think it's really important to connect the data to something that is happening currently or to an actual person and then you can do the reporting and have a better understanding of what's behind the data and who is involved, and that becomes a story at that point."
On the value of teamwork and collaboration
Both Paoli and Li highlighted the importance of collaboration and teamwork for investigative projects, both with fellow students who can contribute certain skills that would complement each other and by looking for guidance from journalists and academics.
Paoli advised students to "put all their personal ambitions aside" and work collectively, and Li pointed out this type of data journalism work should not be done by just one person, but rather by a group of people who can each bring their strongest skills to the table, whether in data analysis or reporting.
"Don't stop just because you don't have all the skills. You don't need to be the person who knows how to do everything, but you do need to know how to make it happen by asking for help," Li said.
Free daily newsletter
- No coder? No problem. The Newscene launches new tool for building online membership models
- Tip: Eight ways to use spreadsheets for data journalism
- The Economist is using interactive data-driven Stories to widen its Instagram community
- How BBC Good Food used voice search analytics to cook up its first Alexa skill
- Tip: How to dig up information online