Panelist Profiles: Meredith Broussard, Tom Kent and Olga Pierce
On October 15, Meredith Broussard (NYU), Tom Kent (AP) and Olga Pierce (ProPublica) joined us for Tow Tea to discuss “Computational Journalism in Practice.” Below, learn a bit more about each panelists’ background, and their answer to a key question about computational journalism.
Broussard is an assistant professor of the Arthur L. Carter Journalism Institute at New York University and a current Fellow at the Tow Center. Before becoming a journalism professor and investigative journalist, Broussard worked as a software developer at AT&T Bell Labs and the MIT Media Lab.
In her work as a freelancer, Broussard combines her past as a coder with in-depth reporting. Her most recent work, which appeared in The Atlantic, she investigated the correlation between Philadelphia high schools, whose student population generally performed low in standardized tests, and the availability of practice text books for standardized tests at the same school. To substantiate her long-form reporting (‘Why Poor Schools Can’t Win at Standardized Testing?’) with a database-application for journalists on Philadelphia textbook inventories at stackedup.org
Key question: When working with data, which comes first: the dataset or the idea for a story?
“You can do it both ways,” said Broussard. For stories that address systemic or public sector related problems, there is usually an accompanying dataset to work from, whereby journalists can compare policies and laws with the real-life empirical data. When something seems off or weird, that is where stories happen.
Kent is a deputy managing editor and standards editor of the Associated Press and an adjunct faculty member at the Columbia Journalism School. He worked as AP correspondent in Sydney and Brussels, and was bureau chief in Moscow. Among others, his journalistic work also include his role as chief of AP operations in Iran during the Iranian revolution.
As standards editor, he oversees – among others – the management of AP’s computer-written wire stories and financial reports.
Key question: What is the future of AI (artificial intelligence) and machine learning in journalism? Will these new techniques cause further layoffs?
“The bus stops with the journalist,” said Kent: decision-making cannot be relegated to computers. Ultimately, automatizing news production frees time and labor for journalists to work towards more complex investigative projects. The work has to be done, thousands of news releases and financial reports have to written. And journalists need to edit these, too. Susan McGregor added: “at some point you need a person to say: ‘this is objective, this is not objective’”.
AP blog (‘The Definitive Source’): blog.ap.org/contributor/tom-kent
Pierce is the deputy data editor at ProPublica and an alumna of the Stabile Investigative program of the Columbia Journalism School. She specializes in data-driven stories and contributes or leads many of the grand-scale projects undertaken by ProPublica. One of these is the Surgeon Scorecard, an online news-application that allows users to look up individual performance of surgeons in New York based on their patients’ re-admittance rate after complications.
Pierce is currently working on Surgeon Scorecard 2.0.
Key question: What skills should an aspiring journalist acquire to successfully work with data?
Pierce’s advice is to have working knowledge in statistics and practice at least one multi-purpose coding language, such as Python or Java. She also warns to not forget about Excel – a standard and last-resort working environment for many data desks at newsrooms. “You’d be surprised how many people don’t know how to use Excel,” said Pierce. Most importantly, however, a journalist should be able to formulate interesting questions around data.
ProPublica profile: http://www.propublica.org/site/author/olga_pierce
Surgeon Scorecard: https://projects.propublica.org/surgeons/