July 25th, 2013 Huge Digital Strategy: Data Research and Visualization

On Thursday, July 25 2013, OLC attended Huge Digital Strategy’s event, Data Research and Visualization featuring Amanda Cox of The New York Times, Tye Rattenbury of Facebook, and Chris Whong of Socrata. The event was moderated by a representative from Huge.
Mod: Can you introduce yourselves?
Amanda Cox: I make graphics for The New York Times. I’ve been there for about seven years now.
Ty Rattenbury: I’m at Facebook as a data scientist. I used to work at Intel. 
Chris Whong: I work at Socrata and do open-source data. I’m in data source analytics.
Mod: What’s your thoughts on data popularity?
CW: Basically, the whole ecosystem around data is supporting visualization. There’s a 
Github-kind of a culture and the resources, they’re right in front of you.
AC: I think making good visualizations targeting the general populations, the people making it, is around 200. 
TR: Data is a particular kind of craft. It certainly has reached a certain type of crowd. There is a lot of what looks like data visualization out there. You see increasing use of data to start arguments and it’s increasing in general discourse. You also have stuff like moneyball, which has become popular in the public.
Mod: Can you speak to visualizations? How you got there?
CW: I started with GIS. It was a form of data visualization. Just being able to work with those kinds of data really helped to democratize data.
Mod: Can you talk about open-access data?
CW: Getting it in the hands of others is half the battle. Putting it in a spreadsheet is part of the work. It needs to be further democratized.
AC: There’s not an either or distinction for us. We actually need to play well with others. Instead of having a map and story, we have a map on top of a story. It’s not any different than crop returns. We call it data journalism.
Mod: I guess you have editorial control over the story?
AC: Whatever makes sense.... We have journalists with great stories and data scientists with great data and we work to combine the two.
Mod: What about you, Ty?
TR: One of the ways we used data was to track what people were doing on their smartphones. We have a visualization that tells less of a story and more of an observation and insight on what’s happening. Visualization became the point to connect the two. It was a way to democratize the value of data.
Mod: Can you speak to the value of data you work with?
TR: We work with second-by-second data—that adds up to millions of data points per user. At Facebook, you’re talking about trillions and trillions of brackets per day. The work is really hard. We spend a lot of time generating plots to get a sense of directional datasets. It’s an interesting interplay of interpretation of what’s going to happen.
Mod: What are some skills that people need to understand data visualization?
AC: I thought I would be able to show uncertainty as a strength. It’s actually understanding where the numbers come from. You need to be comfortable with looking at lines.
TR: The ability to interrogate data—the key is to understand regenerative data. You’re comfortable knowing—I think it’s conceptually knowing that data might now tell a story.
CW: I think knowing to look at process is important.
Mod: Can you define the role of a data scientist?
TR: It depends on the context of where the work is being done. It usually deals with computational data. Also, what’s the frame of work that you’re going into? Data scientist need to encapsulate all of that.