Executive Summary

Over the past century, journalism schools have developed solid foundations for teaching shoe-leather reporting techniques. Hundreds of universities teach how to interview, how to develop sources, how to cover a beat, and how to write a breaking news story, a feature, a sports dispatch, or an investigative piece.

But the practice of data journalism has been largely left out of the mainstream of journalism education, even as the field’s relatively small core of devotees has honed it into a powerful and dynamic area of practice. For decades, data journalists have competed for the profession’s highest prizes and secured positions of distinction within the most competitive news organizations, yet our research has found that relatively few journalism schools offer courses in this area, let alone a concentration, even as these schools have expanded instruction in presentation-focused digital skills.

The authors of this report believe that all journalism schools must broaden their curricula to emphasize data and computational practices as foundational skills. To place data journalism in the core of journalism education will mark a crucial advance in what schools can offer their students. Journalists who understand data and computation can more effectively do their jobs in a world ever more reliant on complicated streams of information.

Beyond teaching, too few journalism schools support faculty research into tools and techniques of data-driven reporting, despite rich opportunities for developing theories and applications that may change journalistic practice. Journalism schools that embrace research in their missions can transform themselves into innovation hubs, introducing new tools and techniques to the profession and across their universities, instead of merely preparing students to enter the field.

This report offers a snapshot of the state of data journalism education in the United States and outlines models for both integrating the use of data journalism into existing academic programs and establishing new degrees that specialize in data-driven and computational reporting practices. While we focus on the state of education in one country, we hope that the results may also be useful internationally.

But first, a definition. When we say “data journalism,” we mean using data for the journalistic purpose of finding and telling stories in the public interest. This may take many forms: to analyze data and convey that analysis in written form, to verify data found in reports, to visualize data, or to build news apps that help readers to explore data themselves. This field also encompasses the use of computation—algorithms, machine learning, and emerging technologies—to more effectively mine both structured and unstructured information to find and tell stories. The ability to use, understand, and critique data amounts to a crucial literacy that may be applied in nearly every area of journalistic practice.

We interviewed more than 50 journalists, educators, and students, and we evaluated more than 100 journalism programs across the nation. This report features a chapter detailing quantitative findings, such as the number of U.S. journalism programs offering classes in data, computation, and related tech skills. We also include a chapter of qualitative findings in which our interviews and classroom observations offer some color and texture to this picture of the present state of data journalism education and its potential.

Among our findings:

• Many journalism programs offer few courses in data journalism, and nearly half offer no classes at all.
• The classes offered are largely introductory, and the need is still largely for the basics, such as knowing how to use a spreadsheet, understand descriptive statistics, negotiate for data, and clean a messy data set and then “interview” it to find a story.
• The field offers a few foundational textbooks, but beyond that lacks a broad and strong core of literature to help teach both the history and practice of data journalism.
• Many journalism programs do not have a faculty member skilled in data journalism. Hiring professional journalists as adjuncts may pose many challenges, one of which is that job openings outnumber qualified applicants.
• Graduates with data journalism skills are better equipped to succeed, our interviews show. Faced with a decision to hire an entry-level reporter with no data skills or one who knows how to use a spreadsheet or query a database, the data skills provide a key edge.

Among our recommendations:

• Journalism schools can collaborate across the university to meet the burgeoning need for instruction in data and computation but should be wary of trying to outsource too much—while understanding how to do math, statistics, or computer programming is an important component, data journalism is much more than that.
• Journalism programs can integrate alternative teaching methods to help fill the gaps in their own faculty. Examples include cooperative teaching among different university departments, online courses, and independent tutorial packs.
• Journalism programs can choose among several models of instruction, all of which begin with a key component: at least one required class in analyzing data for stories—what historically has been termed computer-assisted reporting (CAR).
• Journalism schools that embrace both teaching and research into data journalism methods will be poised to fundamentally improve the way future journalists will inquire into matters of public interest and communicate with their audiences.

Following our findings, this report outlines several model curricula and general recommendations. We offer a model for a core, required course in data journalism. Then we suggest ways of introducing data and computation into existing journalism classes such as Ethics and Global Reporting. Next comes a set of full model curricula for degrees and concentrations in data and computational journalism. Finally, we address a range of institutional concerns on matters ranging from finding teachers to providing technological infrastructure.

Our objective is not to replace or diminish shoe-leather reporting in journalism instruction, but to augment it with data-driven and computational techniques. This report is meant to describe the state of data journalism education, to underline the urgency of incorporating these skills to equip the next generation of reporters, and to offer guidelines for moving journalism schools in this direction.