Beyond Neutral Numbers: Who Controls the Evidence in Great Data Journalism?
NUS NMC5364 Assignment 1: Panel Presentation & Facilitation (25%) | Presenter: Aven Ng | Instructor: Dr. Shan Wu
Executive Summary & Core Thesis
Data journalism relocates human judgment; it never removes it. Credibility cannot rest upon numerical polish or technological spectacle; it must be earned through auditable, accountable, and publicly challengeable evidence.
Working Distinctions (Tong & Zuo 2021)
Decouple three concepts: Accuracy (factual correctness of calculations), Procedural Objectivity (systematic, auditable documentation open to correction), and Value-Free Neutrality (the epistemological impossibility of category-free knowledge). Rejecting neutrality is NOT discarding verification.
Key Readings Synthesized
- Young et al. (2018): In 26 Canadian award finalists, 96% had interactivity, but only 15% allowed open user exploration, and 69% lacked downloadable data.
- Wu (2021): In Singapore's COVID-19 dormitory clusters, state access structures conditioned data gathering and classification.
- Loosen et al. (2020): In 225 DJA projects, 68.4% used official data, but 52% expressed criticism or calls for public intervention.
- Appelgren (2018): Interactive dashboards frequently present an "illusion of interactivity" that channels readers along paternalistic corridors.
- Arias Robles et al. (2023): Among 62 Spanish stories, 98% named sources, but only 24% provided methodology sections and only 11% offered downloadable data.
7 Critical Questions for Class Facilitation
- If numbers are socially constructed categories, does rejecting neutrality damage public trust, or offer a more robust foundation?
- Under what specific conditions does routine institutional data dependence cross into editorial capture?
- Can an interactive data visualization ever provide genuine exploratory autonomy without sacrificing narrative coherence?
- During a health crisis, should newsrooms publish incomplete official dashboards immediately, or withhold until field audits can verify omissions?
- When is withholding raw data an ethical obligation (protecting vulnerable subjects) rather than a transparency failure?
- If we replace objectivity with "public interest" (Tong & Zuo 2021), who gets to define what constitutes the public interest?
- Does the industry's obsession with complex custom graphics marginalize simpler tools that offer superior civic utility?