There is probably no government data used more by journalists -- and non-journalists -- than the trove of population and demographic information collected by the U.S. Census Bureau. But while the bureau has kept improving its tools for online data access, it's still a challenge for someone not well-versed in the workings of the census to find the most useful information -- let alone identify ideas for a journalistic story.
So when my colleagues and I at the Medill School of Journalism were thinking about interesting data sets that we might make more useful for journalists, the Census was an obvious choice. It seemed like just the right focus for a new, experimental class focused on developing "tools for journalists" and enrolling a mix of journalism and computer science students.
The class -- "Collaborative Innovation in Media and Technology" -- just wrapped up last week, with five interdisciplinary student teams presenting prototypes of tools journalists could use to make Census data more valuable. All of the tools are interesting, and I will likely write more about them in the future, but for today, I want to highlight one of them: American Visualizer.
American Visualizer, now in a functional "alpha" form, is worth the attention because it was the most fully realized of the tools created in the class, and because it was developed by a team including Andrew Paley, the fifth "programmer-journalist" attending Medill on a Knight News Challenge scholarship program intended to bring skilled programmers and Web developers into journalism.
Andrew, along with journalism master's student Monica Orbe and computer science student Daniel Kim (and with guidance from Medill Prof. Owen Youngman and Northwestern computer science professors Kristian Hammond and Larry Birnbaum) developed American Visualizer to make it easier to identify interesting stories from the Census.
The site uses information from American FactFinder, the online query tool developed by the Census Bureau to provide public access to its data. American FactFinder, though, is a "data labyrinth," the students said. And even if a user can find his or her way through the labyrinth, the data is delivered in tabular form. Rendering the data graphically -- often the best way to understand its significance -- requires importing the data into a spreadsheet or other software and then creating a graph.
"This tool instantaneously translates data into meaningful information -- from unintuitive and overwhelming collections of American FactFinder tables into immediate, concise and engaging visualizations," Andrew says. "And it does this on demand for whatever geographic region the user wishes. It also allows for the comparison of two regional datasets."
In its current form, American Visualizer makes 10 different datasets available -- five for general visualizations and an additional five for comparison visualizations. Here are some suggestions for seeing its utility (best viewed with the latest version of the Firefox browser):
- From the opening screen, enter a city and state and a type of data you're interested in (housing, population by age, population by race, population by gender and population by level of education). Click "Create" to see a graph of this data. (Note: for a big city, the search results can be a bit slow, since at this point American Visualizer aggregates data from multiple zipcodes.)
- To see other types of visualizations, click on the "Advanced" button in the lower right. Here you can extract data for individual zip codes, compare cities to one another and compare zip codes as well. You can display the data based on raw numbers (for instance, the number of owner-occupied vs. rental units) or based on percentages of the whole. For the comparison of cities and zipcodes, there are additional data sets available: Labor force, mean commute time, median household income, and population below the poverty level.
Technical details: American Visualizer takes advantage of the Open Flash Chart open source visualization library. Beyond that, the underlying architecture is built on standard and widely available LAMP stack server technologies--mainly PHP and MySQL.
Of course, since this is just an "alpha" release, there are many improvements and enhancements that Andrew and his team want to make: speedier query results; additional data types; user-generated customization of fonts, colors and layout; the ability to embed the visualizations, and a mobile app that would generate data based on the user's geolocation.
"Data alone can tell stories. The problem is that data-only stories can be hard to read," Andrew says. "But pictures, as the saying goes, are worth a thousand words."
This was the third interdisciplinary class Medill faculty members have co-taught with Hammond and Birnbaum -- and the first to focus on tools for journalists. These collaborative classes are conducted under the auspices of the Medill-McCormick Center for Innovation in Technology, Media and Journalism.
The first collaboration, last spring, served as a capstone class for the Medill master's students who participated. In that class, student teams created working prototypes of five products combining journalism and technology. One of them, dubbed "StatsMonkey," which writes baseball game stories from box scores and play-by-play information, has attracted a fair amount of attention. One of the team members who developed StatsMonkey was Nick Allen, one of the Knight "programmer-journalists." Asst. Prof. Jeremy Gilbert and I served as the Medill faculty for this class.
The second collaboration, taught by Gilbert, Hammond and Birnbaum, took place in the 2010 winter quarter and enrolled undergraduate students from the two schools. I haven't written about it here because none of the Knight scholars were involved.
Andrew enrolled at Medill last September and has one quarter of graduate study left at Medill, which he'll complete this fall. This summer, he will be working on News21, a multimedia reporting project involving journalism schools from around the country. (Also working on News21 will be Manya Gupta, the 4th Knight "programmer-journalist" scholarship winner.)
Among our scholarship winners, Andrew is somewhat unusual in that he actually studied journalism before -- as an undergraduate at Saint Michael's College in Vermont. Before coming to Medill, Andrew was a musician and a Web developer, most recently for LongTail Video, best known for its open-source media player.
Learn some more about Andrew in this Q&A:
Tell us about your background.
I was born in New Haven, Connecticut, and spent my childhood split between Madison, Wisconsin, and the hills around Burlington, Vermont.
After high school, I went off to Boston to study new media at Emerson College, but the program was in its infancy then -- and I was already becoming versed in web design/development -- so I switched gears/schools.
I ended up back in Vermont at Saint Michael's College, pursuing a journalism degree and a concentration in fine arts. While there, I co-founded, designed and helped launch the first online publication at the school and was a finalist in an international competition to re-imagine Internet browsers. I graduated in 2006, but I hadn't been on campus since 2004, finishing through a protracted series of independent studies that I arranged with key advisors.
My departure from the college campus was due to my other life in music. I spent most of 2004 through 2007 recording and touring the continent (and, eventually, Western Europe in 2009) with my band and through other projects. I continue to write, record and play with a couple of projects.
After many years of itinerant life, I settled temporarily in Brooklyn in 2008 and took a job as lead designer and web developer at LongTail Video. I'd been doing regular freelance and volunteer design/development work for a wide array of companies, bands, non-profits and politics-related entities throughout touring, and the timing worked out.
And then I came back to journalism.
How did you get interested in journalism?
I've always been a political junkie and a writer. It was a natural fit. After a few years away from it I came back because information is a powerful and potentially overwhelming thing, and I'd like to play some part in figuring out how to parse the glut of it growing online into something meaningful and useful. That's really going to be the key going forward -- not just information access, but information clarity and context. Beyond that, I think that the media has been failing us (and our local, national and global debates) for many years, and I'm hoping to be involved in changing that.
What do you think are some interesting career opportunities for people who blend journalism and technology knowledge?
There are all the (newly) traditional places that a tech-oriented person could show up in the newsroom: web producer, database spelunker, interactive designer, etc. But that's an incomplete portrait of the possibilities.
In the same way that creative development and information design have upended much of the old world media -- from Napster to to Bittorrent to YouTube to Twitter to hundreds of other innovations big and small -- I think news is next. And in many cases, the new media barons at the helm of all this innovation came out of literally nowhere. They were 20- and 30-somethings with big ideas and enough development prowess to get them done. That's where the real opportunity for tech-minded people who have a passion for news and information lies -- creative innovation (with both existing tools and those yet to be created).
News is ripe for this kind of directed reinvention, and I think it's already starting to happening with many of the open government and "sunlight" initiatives taking hold online (not to mention the ways those other innovations -- say, YouTube and Twitter -- have altered the way news happens). Pushing forward will require developers to build the new platforms and re-imagine existing ones and journalists to make them meaningful. I would imagine that those who will do this most effectively will be the ones who understand both journalism and technology.