Data Journalism Explainer
Data journalism uses structured information as evidence for reporting. The data may come from government records, scientific measurements, budgets, court files, election results, sensors, surveys, or information collected by a newsroom. A spreadsheet is not a story by itself. Journalists begin with a question, learn how the data was produced, and decide whether the records can answer it. They interview people, read documentation, and compare the numbers with events on the ground so that patterns do not become detached from human experience.
Much of the work is preparation. Real datasets contain missing values, inconsistent spellings, duplicate rows, changed definitions, and dates or locations stored in incompatible formats. Reporters clean those problems while preserving the original files and documenting each decision. They study metadata and data dictionaries to understand what every field means. Before calculating totals or rates, they check the unit of analysis: a row might represent a person, an incident, an application, or a payment. Confusing those units can produce a persuasive but false conclusion.
Analysis can be as simple as sorting and counting or as advanced as statistical modeling and geographic analysis. Journalists look for trends, differences, outliers, and relationships, then test whether the result survives basic checks. They may compare several sources, reproduce calculations, ask subject-matter experts to review methods, and contact the agency that created the records. Correlation does not prove causation, and a large dataset can still be incomplete or biased. Good reporting states those limits and avoids claiming more precision than the source supports.
Visualization helps both investigation and explanation. Charts and maps can reveal patterns that are difficult to see in rows, but design choices affect interpretation. Axes, scales, categories, colors, and geographic boundaries should make the evidence clearer rather than more dramatic. Reproducible notes, published methodology, downloadable data when lawful, and links to source documents let others examine the work. The final story combines quantitative findings with interviews and context, using data to strengthen accountability rather than replace reporting.
Publication does not end the data process. Newsrooms should preserve the source files, analysis steps, and version history so findings can be checked or updated when agencies revise their records. Corrections need to explain whether the problem came from the original data, the newsroom’s calculation, or the way a result was described. Reporters may also build searchable tools that let communities examine local details, but those tools require privacy review and maintenance. The strongest data journalism remains accountable after launch, not only persuasive on publication day.
Public agencies publish many datasets through open-data portals, reports, and records requests. Journalists also obtain court files, corporate disclosures, research databases, leaked documents, and information gathered through surveys or scraping when lawful and ethical.
Provenance matters. A reporter records where a file came from, when it was downloaded, who created it, and what period it covers. A polished portal does not guarantee accurate records, so the source's collection process and incentives need scrutiny.
Cleaning is not cosmetic. Decisions about duplicates, categories, missing entries, and name matching can change a result. Journalists keep untouched originals, work on copies, and record transformations so another person can reproduce the analysis.
Unexpected values often become reporting leads. A sudden gap may mark a policy change; an impossible date may reveal an export error; a cluster of unmatched names may show inconsistent recordkeeping. Reporters investigate rather than silently forcing every anomaly into a tidy table.
Responsible stories distinguish exact counts from estimates and explain margins of error, missing jurisdictions, suppressed values, or changing definitions. They describe important assumptions in ordinary language and avoid visual precision that the source cannot justify.
Editors may publish a methodology note, code, and selected source files so readers can inspect the process. Privacy, contracts, or source protection can limit disclosure, but the newsroom can still explain its methods and the checks used to support the findings.
Explore more "Explainers"
Discover additional explainers across politics, science, business, technology, and other fields. Each explainer breaks down a complex idea into clear, everyday language—helping you better understand how major concepts, systems, and debates shape the world around us.