« « Exploring Drupal — open source tool of the day

Data Delver: William Hartnett, Palm Beach Post » »

Data Delver: Gregory Korte, Cincinnati Enquirer

I believe all beats would benefit from considering data in their reporting, but some beats demand data analysis on a regular basis.  The nature of CAR is changing — many industry folk have told me that the term “CAR” now encompasses reporters, data analysts and web developers.  It’s certainly a wide field.  One reporter I talked to sees data as key, as he focuses on the analysis.  All tools are a means to reporting for Gregory Korte, an enterprise and investigative reporter at the Cincinnati Enquirer.

This profile of Korte is a part of my continuing series I’m calling “Data Delvers,” where I pass on summaries, quotes and audio clips from conversations with journalists using technology to find, analyze and convey data-driven stories and/or projects to the modern audience.

The path to CAR

When he was a student at Ohio University, he didn’t see technology as influencing reporting. While he started to take a computer programming class– he thinks it was Fortran –he didn’t complete it.

In the fall of 1993, he interned at the Washington Bureau of the Cleveland Plain Dealer, working with an aviation reporter who was using FoxPro to analyse data on reel-to-reel tapes.  “It looked time-consuming and intimidating,” he said.  He didn’t think of using computers to write stories for a few years, working at the Lorain Morning Journal, and then the Akron Beacon Journal.

In the late ’90’s, an editor taught the Akron reporters about Excel, and Korte used it to analyze budgets on his city hall beat.

His first major CAR story came in 2004, three years after he moved to the Cincinnati Enquirer.  Korte used a relational database to reveal more than 2,000 landlords in the local county claiming tax deductions on multiple rental properties, which is illegal. The NICAR conference met in the same city his newsroom was in that year. He didn’t go.

By the end of 2005, other database reporters had left, and Korte found he was the new expert by default. He got into mapping, and started maintaining databases both online for the public and in-house for backgrounding reporters — these initiatives were part of a response to a Gannett plan focusing on the restructuring of newsrooms and emphasizing the importance of data.


CAR requires flexibility and self-learning

Nowadays, he works as a reporter on breaking news, investigative pieces, and online database projects. And currently, he’s spending a lot of time on Census and stimulus work.

How does Korte strive to avoid data ghettos, non-contextualized posting of information?  By posting data sets and mapping by location, and allowing users to compare data sets by filtering the map.  This is a project called the CinciNavigator.

Korte, like many data reporters, has had the flexibility to work from home or the office during his career.  He said he likes how much he is able to get done at home.  “If I’m in the office, I tend to feel guilty about dead ends or taking the time to learn something.”  As he was picking up new skills, he said he knows that it took him two or three hours to do something that now takes him five minutes.  “I always feel like there’s something I haven’t learned, you don’t even know what you don’t know.”

His latest project is continuing to get into web development.  He’s been exploring web frameworks and Django since 2007, when he built CincyPolitics on his own time, which maps election results.

He credits NICAR conferenes and support from the Investigative Reporters and Editors organization with speeding up his learning.  He has participated in boot camps, and recommends the hands-on panels at the conferences.


Census 2010: The story that demands CAR

Korte’s been looking deeply into the census, the king of all stories for data reporters.  Data collection begins on April 1 of this year, and the data will be released on Dec. 31.  As a sample story, Korte’s looking closely at the possibility of Ohio losing two Congressional seats, because of an expected dip on population.  For that reason, he said he will be closely waching the county population count.

But that’s just the beginning.  “It’s a long, slow trickle of information,” Korte said.

One of the big changes from Census 2000 is that there used to be an embargo, so reporters would have a few days to analyze data before writing their stories.  Now, that embargo has been removed, and the reporters get the information when the public does, making it more difficult to write in-depth stories quickly, and making interactives much more difficult to produce in a timely fashion.


All reporters should have fundamental data skills

While not everyone needs to be a specialist, Korte said, “Everybody should know Excel or a spreadsheet, and how it’s organized, how to do counting, totaling and averaging.”

“Don’t be afraid of data,” said Korte.  To him, that means being willing to use data sets, but also approaching the numbers as you woud any source.  “Think about the data that is available, talk to the people who have it, think about what information you want.”

Using programming for analysis and web scraping is a great way to take your skills further, he said.  “The value of programming is that it can speed up the more repetitive tasks that you were doing manually, and it allows you to do what others cannot.”

“Most reporters see a new database, and it’s just too tempting– they’re only doing the story because it’s low-hanging fruit, begging to be analyzed.  The issue in news is to find the right data.”


Working as part of a team

It’s through collaboration between beat reporters and data specialists that the best stories can emerge, Korte said. He cited a package called “Gas Price Secrets Revealed” (PDFs of print articles here, data app here listing current gas prices and other info).  He partnered with James Pilcher.  While Korte focused on the database analysis, Pilcher had more experience on that beat, and he was better able to find people to talk to and quote in the story.

There are still ways project workflow can be improved, Korte said.  “I wish there were more discussion at the beginning of a project, every map starts to look alike after a while.”

Korte said this might be helped if people made the extra effort to not rush to presentation, but think ideas through before jumping into packages.

Here’s how he suggests approaching a data project: “Take a deep breath, and don’t rush to anything within an hour.  Make sense of the data.  A data dump just leads to a superficial story and presentation. Find the story, then support it.  Stories last forever.”

Sounds like good advice across the board to help us create journalism of substance.

« « Exploring Drupal — open source tool of the day

Data Delver: William Hartnett, Palm Beach Post » »
  • http://gregorykorte.com Gregory Korte

    One minor clarification, in case I somehow misspoke or my this is misinterpreted: I don’t want anyone to think I’m claiming credit for CinciNavigator. That’s really Andy Crain’s masterpiece. At most I contributed about 2 percent, helping to conceive it early on (though the idea really came from Adrian Holovaty’s Chicagocrime.org), and point Andy toward some public record data sources.

    Other than that, I think you did a good job making me sound like I have a coherent philosophy of computer-assisted reporting. Thanks for asking.

    [Reply]

  • Pingback: Journalism students: Please don’t do this : Gregory Korte()