Click the the image to read our entire series. Public domain image via DARPA
Gone are the days where journalists could boast about not being math people.
While the image of an investigative journalist sifting through reams of printed spreadsheets and pouring over documents has endured, the tools have changed.
The rise of data journalism – which is often tied to the meteoric rise of analytical news ventures – has pushed journalism educators to innovate. With employers increasingly demanding data skills, schools are paving the way for data education programs. And because the toolkit and best practices for data journalism continue to evolve, many programs are tackling the emerging field by empowering students to aptly use data in their reporting, instead of dedicating classes to learning specific technologies.
A growing trend
The idea of using numbers to report the news is not new. Computer-Assisted Reporting (CAR), or simply using computers to gather and analyze data for stories, has its roots in the early 1990s, when journalists relied on database programs like Microsoft Access and Excel to store and organize data.
But the genesis of digital journalism has created new possibilities to analyze large data sets, draw compelling stories out of public data sets and visualize changes over time using a new set of tools. The rise of data reporting has been driven in part by the large catalog of programs that are available open source, where a product’s source code is made available through a free license for contributions by developers. The availability of these tools exemplifies the ethos espoused by many leading journalists and developers working with data, which emphasizes transparency and collaboration in making resources open to the public.
The National Institute for Computer-Assisted Reporting (NICAR) has been educating journalists since 1989. Since then, David Herzog, the academic advisor to the group, said there’s been an “explosion” in the variety of tools available to journalists, from online mapping platforms and data parsing tools to web scraping software automated to condense unstructured data sets online into a form that can be readily stored and analyzed.
Herzog, an associate professor at the University of Missouri, also said there’s been a substantial uptick in journalists attending NICAR trainings, and more employers investing in data skills by financing staffers’ training.
While teaching jobs typically don’t perfectly follow the same trajectory as the data journalism boom, Meredith Broussard, an assistant professor at Temple University, said the trend of universities beginning to hire more instructors focused on data courses will intensify in coming years.
Different approaches to integrating data education
Just as professional journalists continue to test and develop new tools to support data reporting, journalism schools are implementing a number of approaches to prepare students for the rapidly changing landscape.
At Temple, Broussard said there has been a move to integrate data skills at every level of the journalism curriculum. Students are first introduced to data journalism in the school’s required introductory class and dive deeper into the principles of CAR – like advanced database searching and using documents as sources – in a required research class. After mastering the basics, students work together to craft stories, each taking a specialized role on three-person teams, similar to how national newsrooms churn out data-driven stories.
Broussard’s class worked on Stackedup.org, a tool to help parents determine whether Philadelphia schools have the books they need.
But Broussard says her courses are less about teaching technology and more about arming students with the ability to pick the right tool for the job.
“[Technology] is well enough integrated into our culture and corporate life where knowing how to use a computer doesn’t set you apart anymore,” she said. “It’s important to be a good journalist and be able to use whatever tools you need to tell an impactful story.”
Striking the right balance
At the University of Wisconsin-Madison, assistant professor Chris Wells is one member of a faculty team launching the journalism school’s first foray into a course dedicated specifically to computational approaches to communication, with an upper-level data journalism class first offered for fall 2014.
Wells said he looks to strike a balance between the technical and the conceptual. He plans to begin by teaching students basic numeracy and statistics before covering design elements and visualization later in the course to support the reporting students already know how to do.
“We want to bring the perspective of data visualization to a place where our students are already strong – telling compelling stories,” he said.
Down the road, he expects data analysis to become better integrated across the school’s curriculum as more faculty turn their focus to digital storytelling and see the possibilities from this inaugural course.
While some schools tackle data projects in individual courses, the Lede Program at Columbia University and the Knight Lab at Northwestern University occupy a unique space between computer science and journalism education with a larger currcular approach. While Knight focuses more on pairing students with faculty and staff to develop open source, web-native tools for journalists and news organizations, the Lede offers certification programs to post-undergraduate students.
The Knight Lab at Northwestern University creates apps, prototypes and proof-of-concept pieces, often driven by data and scraping. (Screenshot)
“All data is bad data”
After 15 years in newsrooms, Herzog is well acquainted with journalists who say their strength is working with people, not numbers. But in a world awash in data, he said journalists who do not have the skills to evaluate data are missing out on telling more well-rounded, interactive web stories.
Because most newly minted journalism school graduates are not entering the job search with advanced data skills, Herzog’s students – trained to use online mapping, geographic information system programs and mySQL, an open source database, to name a few – have a competitive edge. And students know it: Herzogs’s courses are always enrolled to capacity, with students turned away for lack of space.
“All data is bad data until you become good readers of data and [are] comfortable using it for the purpose you’re trying to do,” he said. “I want to prepare students to be nimble in the workplace, learn how to learn new tools and work their way through problems.”
Toward a decentralized model
While national leaders for data education – NICAR, the Lede and Knight among them – have proved successful ventures for both instruction and entrepreneurship, a low barrier for entry makes data education widely accessible.
Wells said although UW-Madison’s program was made possible through an Educational Innovation grant from the university, a wide range of free, open source data tools means the tools for instruction are not resource-intensive.
Without the expense of software licensing, the limiting factor for journalism schools looking to launch data programs could instead be the availability of faculty members equipped to teach these courses, as Broussard said a “critical mass” of journalists able to teach data journalism is still developing.
Journalists have also surged ahead to develop more decentralized tools for reporters to get data training, including building training guides available for free online.
“The hiring in journalism is in data journalism. There’s no doubt about it, people with those skills are being snapped up,” Wells said. “Students coming out with those fundamental skills…are [finding them] powerful assets hitting the job market.”
Katherine Krueger is a recent journalism graduate from the University of Wisconsin-Madison and former editor-in-chief of The Badger Herald. You can find her tweeting frequently at @kath_krueger and blogging at katherinekrueger.com. She is hoping to continue her journalism career in New York.
