Producer/Editor: Roselyn Du

Project Assistant: Zoe Hu

(Aaron Mendelson at his KPCC desk in Pasadena, California. He was interviewed by Roselyn Du of Hong Kong Baptist University in February 2019 for his data journalism works.)

Aaron Mendelson can usually be found working on one radio broadcast or another in his office at Pasadena, California. Where most journalists work with verbal content, Mendelson also concerns himself with data analysis. He is thankful to his master’s degree from UC Berkeley Graduate School of Journalism. He said the data skills that he is now using at work for Southern California Public Radio’s data journalism and interactive projects are mostly learnt from there.

Mendelson joined Southern California Public Radio (otherwise known as KPCC) in 2014, and became a member of the station’s investigative team in 2017. As an affiliate station of National Public Radio (NPR), KPCC airs programs produced by NPR and inserts local news within these broadcasts. At KPCC, Mendelson uses data to shed light on a variety of topics, ranging from the influence of outside money in local politics to Los Angeles’ bicycle infrastructure to spiking firearms sales and police militarization. “There are two basic models of reporting routines that we have in KPCC,” he explained. “Often it will be one investigative reporter having an idea, and then they or the editor will approach me. I will investigate what data is available and then we will make a decision based on that. The other way is, I sometimes work on stories on my own with similar process.”

In 2015, KPCC launched a project on police-involved shootings, with the goal of analyzing how often on-duty police officers and sheriff’s deputies in the Los Angeles County fired on civilians. In addition, it analyzed summaries of officer-involved shootings and compiled a database of relevant statistics, making the information available to the public. Just as explained on the project website, this isn’t data you can simply look up.  “Most data on police shootings is scattered, focused on fatal shootings,” Mendelson explained. “The only agency privy to the details of every shooting in the county is the Los Angeles County District Attorney’s Office.” It is a time-consuming task to build this database, but the project has become Mendelson’s favorite at KPCC. “By building this database, we are able to take a much bigger picture of story than just date, time and location. And when data is used well, it will improve the quality of the news story,” Mendelson added.

For any data reporter, access to complete and compelling sources is crucial to producing quality content. Mendelson and his team are fortunate enough to have professional legal support in this regard. “We have a lawyer who works with us and helps us try to get data in some cases,” said Mendelson. “She is not an employee of KPCC, but she does work with us in her area of law.”

While KPCC produces print material in addition to its radio broadcasts, radio users still form the bulk of the station’s audience. Most of these users listen to KPCC using vehicular audio systems, and are likely too busy driving to pay attention to numbers and raw data. So how do editors and reporters deliver data-based information to radio listeners? In fact, they try not to. Mendelson and his colleagues will only pick one or two data-based findings to include in their stories, depending on the story length. “The time constraint can be very strict, partly because we are inserting out programming into the NPR shows. So, when you are choosing what of those things you would report and that would be the luxury to put on the radio.”

Instead of presenting statistics and numbers, Mendelson prefers to draw listeners’ attention to the underlying trends behind the data. One example of this was when Mendelson’s team broadcast an episode called ‘Repeat’ on police shootings. “We were tracking which police officers had shot people most often, and we realized in the shooting report, there was one chief deputy who shot four people in seven months,” said Mendelson. But he and his team didn’t broadcast the data findings immediately. Instead, they filtered out and presented their core findings sans data and only broadcast the data at the fourth or fifth episode, by which listeners had already listened to a couple hours of broadcast. “We can save it until when we feel like we have the most impact,” Mendelson explained.

When asked of his insights into data journalism, Mendelson traced his experiences in the field back to 2013, when he began studying data skills as a graduate student at UC Berkeley. Since his graduation, he has witnessed sweeping changes in data journalism in the U.S. Without a question, tools generate faster than they used to be, from programming to third-party platforms or software, like Python, Tableau, even Microsoft Excel. However, while all these advances in software have led to marked improvement in data visualization, the importance of actually collecting and analyzing data remains undiminished. “The graphics are better, but the work is still the same,” said Mendelson. “You have to get the data from somewhere, and you have to get it and figure out what’s in it.”

While talking about artificial intelligence and its potential dangers to the field of journalism, Mendelson expressed doubts about the likelihood of his job being replaced by machines. “What I would like to see happen is a marriage of automation and human judgement,” he says.

Another phenomenon that Mendelson would like to see in the future is more local data reporting. He wishes reporters who work in broadcast journalism would understand data a little better, and believes that there ought to be more local data reporting standing beyond “The Big East” – The New York Times, Washington Post and Wall Street Journal. “I think it’s worth learning at least a little about data journalism, even if you don’t want to be a data journalist yourself,” he said. “Maybe you want to cover entertainment, but a lot of entertainment is about the Hollywood box office, so numbers are important there. Maybe you want to cover sports, but sport is the most data-rich thing that could possibly be. If you want to cover politics or health – anything – these are not field surveys isolated from numbers.”

This need for data in journalism, he adds, makes it that much more important for journalists to be able to find and use data. “You don’t have to wait for academic researchers to publish a study about what you are interested in. You can do some of that kind of work yourself,” he said. “I think even just knowing how a spreadsheet works and what kind of questions you can answer with the spreadsheet is going to help.”

For more information, you can visit:

1. Officer Involved – A KPCC Investigation into police shootings in Los Angeles County (Series Credits)
2. Unarmed and Dangerous?
3. A Call for Help
4. I See Heroes
5. Black People Shot at Disproportionate Rate
6. How Did We Do This?
7. Explore the Data