What Does a 'Data Delivery Editor' Do?

Apparently there are a lot of black bear sightings in Virginia. If you live in that state, you might want to know where and when those bears were spotted. That's why Matt Chittum, of The Roanoke Times, is building a database and accompanying map.  His title at the newspaper is "data delivery editor."

This bear map will join a rapidly expanding set of databases collected in the newspaper site's DataSphere, which launched last Thursday. The DataSphere, modeled after the DataUniverse at the Asbury Park Press and RocDocs at the Rochester Democrat & Chronicle, collects in one place a couple of dozen databases with information important to the local audience. (Learn more about DataUniverse.)

The Roanoke Times -- known for its pioneering efforts in multimedia and interactivity -- hadn't put as much thought into how it handled databases until a dust-up earlier this year involving a decision to publish concealed-weapons permits on the site.

As you might predict, many, many people were unhappy," Roanoke Times editor Carole Tarrant explained in an e-mail interview.

"I didn't want that controversy to ruin any exploration with databases. I didn't want us to forever shy from running them because we had one really bad experience. I wanted to show they could be used for the good -- for good journalism -- if you carefully weighed the value of the information you were presenting."

So Tarrant created the new staff position of data delivery editor and set about building new databases, linking to existing databases on the Times site, and linking to outside databases of government and community information.

"The data delivery editor wasn't an additional position -- we had to make a choice to not do something else," Tarrant said. That's not an easy decision when your newsroom is shrinking. Chittum explains his role and what he's working on in the Datablog.

One piece of the DataSphere, The Beamer File, allows users to follow the 21-year career of Virginia Tech football coach Frank Beamer. You can sort by season, month, opponent, the opponent's conference and games that were televised.

Other than a couple of maps, there isn't much visual representation of data so far in the DataSphere. The New York Times is showing what's possible in this regard: a graph of a presidential debate transcript and a map of presidential campaign visits. Also, Ben Fry (who literally is writing the book on visualizing data) built a cool Major League Baseball salary analyzer to show which teams have the most cost-effective payroll. A slider bar on the graph enables the viewer to see how the results changed throughout the season.

Chittum has some good ideas for how he'd like to improve the DataSphere with visualization tools. One tool would convert data to tag clouds, bar charts, pie charts and fever charts. (For the uninitiated, a tag cloud or data cloud is a weighted list. Here' an example showing world population weighted by font size.)

"I'm especially fascinated by tag clouds, partly because I haven't seen them on mainstream newspaper sites, but they can offer quite effective quick-read presentation," Chittum wrote in an e-mail. "One idea I have in mind is a tag cloud of the most common drugs people overdose on. Maybe the causes of death of for all Virginians as a tag cloud."

Pretty innovative stuff for a newspaper. That's what can happen when you create staff positions such as "data delivery editor."

http://poynter.org/column.asp?id=31&aid=132248