Topher Sanders covers racial inequality for ProPublica. He previously reported on education and city government for The Florida Times-Union. Named to the paper's investigative team in 2013, he became the investigative editor in 2014.
How would you describe your job?
I'm a reporter. And like any reporter in the country, it's my job to question systems, officials and industries. I hold those with power accountable for their decisions and I shine light on corruption, failing bureaucracy and harmful practices.
Have data journalism tools enabled you to better pursue stories on the criminal justice system?
Yes. Data tools allow me to dig into systems and to confirm whether an anecdote is simply an outlier or an example of wider concerns.
Your reporting has revealed hundreds of juveniles in the Jacksonville area took plea deals to avoid being charged as adults. Did this reporting begin with data analysis, or did interviews lead you to look into the data surrounding the issue?
Interviews with lawyers led me to request the data, which hadn't been requested before. Once we found the data supported the lawyers' claims, we then sought out juveniles and families willing to tell their stories.
"I attempt to craft my stories by answering all the questions I can, digging as deep as I can and filling all the holes. If you do that, you put the story (whether it has data or not) in the best position to have impact."
How did you go about putting a human face on a story that involved so much data analysis?
Lots of hours sitting in court listening to how judges, prosecutors and defense lawyers dealt with juveniles. I also developed sources that pointed me in the direction of families who said they had been threatened with adult charges. More families said no than said yes, but that's the nature of this kind of work.
What have you learned from your data reporting projects about how to craft a story to have the greatest impact?
Impact is great when it happens, but it is often based on factors beyond the reporter's control. So the pursuit of impact doesn't drive the crafting of my stories. My desire for impact drives the stories I choose to do, but not necessarily how I do them. I attempt to craft my stories by answering all the questions I can, digging as deep as I can and filling all the holes. If you do that, you put the story (whether it has data or not) in the best position to have impact. Clear and easy to understand writing can also put you in a great position for impact. And that's even more important for a data-heavy story. Readers need to easily understand what you've found and why it matters.
What are your go-to tools/programs when working on a story that involves data? Are there any specific databases that you think are useful for journalists who are just getting into data?
Many of the agencies I've been covering aren't super helpful when it comes to providing data. Shocking, I know. So I've used Tabula and Cometdocs a lot to help turn horrible PDFs into spreadsheets I can work with. School data is great for folks new to data. It tends to be very uniform and allows for reporters to do year-to-year analysis across many variables.
How do you go about confirming that what you see in a data set is an accurate analysis of a situation? Do you have other do's and don’ts for journalists looking to expand their data reporting skills?
Data doesn't replace good reporting. I'm working with a data set now that I thought was great initially, but after doing some reporting, I've learned it's filled with errors and misleading information. So the confirmation process for me involves working my sources to learn if what the data appears to be saying is accurate. My only advice around "do's and don'ts" is that reporters should treat data just like any other claim from a source, document or report. If something in a report looks crazy and inaccurate, you wouldn't just run with it. You'd make some calls and bounce it off your good sources before you'd publish it. Treat data the same way. In fact, even if the information looks great, you still have to report and scrutinize your data. That's the only way you'll be lock-solid and have a great night's sleep before you publish.
This interview has been edited for clarity.