Data Power Conference

Picture of a conference workshop

<< Back to Programme

Panel Session 1c): Data Journalism (Chair: Eddy-Borges Rey)


Empirical Passions, Empirical Power: The Long History of Data Journalism

C.W. Anderson, College of Staten Island (CUNY)

Today data journalism is a hot topic and the use of journalistically inclined data visualization appears to be on the rise. According recent overviews of the field (Howard 2014), academic historiography (Parasie & Dagiral, 2013), and self-talk by the founders of new data journalism projects (Silver 2014), this new form of quantitative reporting rescues journalism from its empirical backwater and brings reporting closer to an ideal if popularized form of social science. Journalism and social science are fusing into the strange hybrid of data journalism, it is claimed. This paper takes a look back at the strange pre-history of data journalism, and in doing so it attempts to shed light on our present era of journalistic hybridity. The paper draws on new materialist theory (Coole and Frost 2010), science and technology studies (Anderson and deMaeyer 2014), and recent calls for passion and affect to be more widely integrated into the media production studies agenda (ie, Deuze and Witschge 2014).

Specifically, this paper examines how the fuzzy boundary line between journalism and social science was erected by telling the story of one important (but by now largely forgotten) news magazine, The Survey Graphic (1921-1952). This paper argues that the Survey Graphic embodies both the apex and the exhaustion of three important Progressive Era tendencies: the muckraking tradition, the naïve empiricism of the social surveyors, and the “problem-oriented” wing of the new profession of sociology. The paper further argues that the Survey Graphic represented the final flowering of this casually hybrid journalistic tradition. The second half of the paper briefly places the empirical culture of the Survey Graphic into dialog with other, later journalistic decelopments-- precision journalism, data journalism, and computational journalism-- in an attempt to shed light on today’s drive towards a robust journalistic empiricism.

Anderson, C. and DeMaeyer J. (2014). “Introduction: Objects of Journalism.” Journalism: Theory, Practice, Criticism.

Coole, D. and S. Frost (2010) (eds). New Materialism: Onotology, Agency, Politics. Raleigh, NC: Duke Univeristy Press

Deuze, M. and Witschge, T. (2014). “Passion, Politics and Play in Journalism Start-Ups.” Presentation at Social Media and Public Space Conference, Amsterdam, NL. June 20, 2014

Howard, A. (2014). The Art and Science of Data Driven Journalism. Tow Center, Columbia University. Online at

Parasie S. and E. Dagiral (2013). “Data-driven journalism and the public good: ‘Computer-assisted-reporters’ and ‘programmer-journalists in Chicago.” New Media and Society. 15(6): 853-871

Silver, N. (2014). “What the Fox Knows.” Online at, retrieved June 29, 2014.


Remediation isn’t the Remedy: Social Media Bias and Broken Promises of Data Representativeness

Jonas Andersson Schwarz, MKV, Sodertorn University

Based on a recent quali-quantitative study of social networking sites (SNSs) I explore the intersection of conventional mass media and social media in order to address a number of urgent problems relating to visibility, accountability, and the power to influence.

In earlier work, I have critically engaged with the concept of “social big data” (Bolin & Andersson Schwarz, 2015). Through new empirical findings, I will focus on the problem of front-end perception versus back-end access. Our findings suggest that, contrary to popular belief, despite their metrological character, SNSs make for capricious conditions regarding estimations of quantities, unfavorable to representativeness—particularly in their front-end uses/appropriations. Despite best intentions among professional communicators, a categorical (even polarizing) logic is introduced when estimations are executed in flawed, uncritical ways—e.g. when journalists rely on front-end access in order to make real-time estimations of popular opinions.

When conventional mass media actors let “social media” serve as a representative of an imagined “general public”, this is rarely based on comprehensive oversight, or statistically tenable analysis. Rather, claims that “social media” engagement of various kinds is “trending” are routinely based on snap judgments affected by bias, limited oversight, and “filter bubbles.”

While the “social media logics” of e.g. Twitter and Facebook are distinct from the established “mass media logics,” I argue that certain effects and processes are catalyzed when these two logics interact—sometimes in highly problematic ways. Existing societal discord about ways of knowing and discerning the world around us risks becoming amplified rather than remedied.


Narrating Networks of Power: Narrative Structures of Network Analysis for Journalism

Liliana Bounergu, University of Amsterdam, Jonathan Gray, Royal Holloway, University of London and Digital Methods Initiative, University of Amsterdam, and Tommaso Venturini, SciencesPo Medialab

In an era of Big Data, networks have become the core diagram of our age. As popular books on the topic contend, the concept of networks has become central to many fields of human inquiry and is said to revolutionise everything from medicine to markets to military intelligence.

In the context of media and journalism, using data to map networks is praised for its potential to expose the workings of power, be it financial or political. The work of the artist Mark Lombardi, as well as power mapping projects such as They Rule, Muckety, Little Sis, Poderopedia and the Organized Crime and Corruption Reporting Project’s Visual Investigative Scenarios have opened up journalistic imagination about how network analysis and mapping might be used in the service of journalism.

While journalists have been experimenting with network analysis and mapping to discover and tell stories with data for decades, the breakthrough moment of this analytical and storytelling device in journalism has yet to come.

Journalists have been reluctant to embrace network analysis and visualisation, and not without good reason. While network analysis can be an effective exploratory tool, in order to be used as narrative tools networks have to be embedded in a rich conceptual framework to generate meaning.

In this article, we propose a possible framework to breathe meaning into networks, a vocabulary of narrative functions that network analysis can play, based on the popular social research approach of ‘issue mapping’, and on examples of use of network analysis and mapping techniques in journalism. Developed at the crossroads between Science and Technology Studies and Internet Studies, issue mapping operationalizes concepts from Actor-Network Theory (ANT) in order to study the state of public issues.

The resulting classification of narrative structures of network analysis in journalism and issue mapping will provide an opportunity to reflect on the potential and limitations of network analysis for mapping power in the context of journalism, as well as on how essential aspects of journalistic epistemology – such as notions of time, space and narrative – are being reconfigured by this set of technologies, practises and concepts.


Quantifying journalism - A Critical Study of Big Data Within Journalism Practice

Raul Ferrer Conill, Karlstad University

The irruption of digital journalism introduced several opportunities and challenges to journalism. Roughly two decades after the introduction of the internet, big data has started to transform the way we understand information and how to use it. The quantification of visitors, readers, and users’ interactions has become the de facto analytic tool for digital newspapers analysis. Accordingly, robot journalism and new storytelling techniques, such as gamification, have started to use and apply the data in order to create a personalized news experience, to suggest specific content, and to enhance interpersonal interactions within the system.

But what happens when big data is targeted to the journalists themselves? How is the quantification of journalistic output received by journalists when the data is used to assess their own quality? This paper aims to answer these questions by looking at the case of the sports news website Bleacher Report. B/R turns journalists into users by awarding them with points according to their writing career statistics regarding their contribution to the site. Number of reads, number of comments, number of lead stories, and other metrics keep adding points defining each author’s reputation level. This quantification becomes an important factor to assess the journalist capacities.

When data is used to turn work into play and quantity into quality the values and norms upon which traditional journalism is built seem to be under threat. This case study provides the room for a critical discussion on the potential use of big data through game mechanics targeting news-workers.