DATA4207 · Data Analysis in the Social Sciences
Data Journalism
Week 11 examines how data analysis reshapes journalism — from forecasting to newsroom investigations — and the pipeline of sourcing data, plotting descriptives first, then modelling to account for confounders, and visualising for a general audience. Students run a non-assessable data-journalism group project that rehearses the full analyse-and-communicate loop. The emphasis is clear, honest charts for readers, not experts — a communication skill the assessed report also rewards.
What this chapter covers
- 01Data journalism = telling stories to a generalist audience supported by numbers (journalism + statistics + computing + design)
- 02The pipeline: data collection → analysis (patterns) → visualisation (often interactive)
- 03Storytelling with data: no jargon, a clear narrative with human interest, highly effective visualisations
- 04Descriptive-then-model workflow: plot descriptives first, then model to account for confounders
- 05Reporting vs data journalism: putting quantitative analysis at the centre, with rigour about causation and observational limits
- 06Assembling a panel: rolling averages, lags, and converting counts to rates per capita
- 07Communicating complex ideas to a non-technical audience; exemplars (BBC, Guardian Data)
- 08Non-assessable COVID-19 group project rehearsing the full loop
Building a reader-facing data story
- +1Source and clean the panel: join the data sources on a location key, convert raw counts to rates per 100,000, and add a rolling average (e.g. 14-day) and any lags. Expect roughly 80% of the effort to be joining and cleaning, not modelling.
- +1Descriptives before models: plot the trends and distributions first to understand the data and spot confounders, then fit a model only if it sharpens the story and accounts for those confounders.
- +1Visualise for a general audience: no jargon, a clear narrative with human interest and significance, and one highly effective (ideally interactive) chart rather than many dense expert plots.
- +1Keep the chart honest: label axes, avoid distorted scales, and communicate uncertainty. The reader should grasp the point in seconds and not be misled — accessibility and honesty are the marks of good data journalism.
Key terms
- Data journalism
- Journalism that puts data analysis at the centre of the workflow — telling stories to a generalist audience supported by numbers, blending journalism, statistics, computing and design, and often interactive.
- Descriptive-then-model workflow
- The discipline of exploring and plotting the data (descriptives) before fitting any model, so you understand distributions and confounders first and model only to sharpen the story.
- Rate per capita
- A count converted to a comparable rate by dividing by population and scaling (e.g. new cases per 100,000). Essential for fairly comparing areas of different size.
- Rolling average
- A smoothed series (e.g. a 14-day mean) that removes short-term noise so a reader can see the underlying trend; computed in R with tools like zoo::rollmean().
- Storytelling with data
- Making data-driven insight accessible and engaging through three moves: no jargon, a clear narrative with human interest, and highly effective visualisations.
- Audience-facing visualisation
- A chart designed for a non-technical reader — clear, honest, quick to read — rather than a dense expert plot. Honesty (labelled axes, undistorted scales, stated uncertainty) comes before cleverness.
Data Journalism FAQ
How is data journalism different from ordinary reporting?
Data journalism puts quantitative analysis at the centre of the story rather than using data as decorative support. It demands greater rigour — attention to confounders, causation and the limits of observational data — and leans heavily on visualisation, often interactive. The skill is knowing what data is relevant, where to find it and how to examine it, then communicating the finding clearly.
Why plot descriptives before modelling?
Because the descriptive picture tells you what the data actually looks like — its distributions, outliers and likely confounders — before you commit to a model. Jumping straight to a regression risks modelling artefacts you would have caught with a plot, and for a general audience a clear descriptive chart often communicates more than a coefficient table.
Is the Week 11 project assessed?
The data-journalism group project this week is non-assessable — it exists to rehearse the full analyse-and-communicate loop end to end. Even though it does not carry marks, the communication skills transfer directly to the assessed group projects and the 50% individual report, where clear, honest visuals are rewarded. Confirm what counts on Canvas.
Can AI help me with data journalism in DATA4207?
Yes, as a study aid. Sia can explain the data-journalism pipeline, the descriptive-then-model workflow, and how to make a chart accessible and honest, and it can give feedback on your narrative and visual choices. It teaches the craft and checks your reasoning; it does not do graded work, and University of Sydney academic-integrity rules apply — confirm on Canvas whether AI is permitted for a task.
Assessment move
Use this week to sharpen communication, which pays off across every assessment. Take a real dataset, assemble a clean panel (rates per capita, a rolling average, a lag or two), and practise the descriptive-then-model order: plot first, model only if it helps. Then build one reader-facing chart and write a short, jargon-free narrative with a human angle, checking the chart is honest — labelled axes, undistorted scale, stated uncertainty. Study a couple of exemplars (BBC, Guardian Data) for how they make numbers accessible. The Week 11 project is non-assessable, so treat it as low-pressure rehearsal for the visuals and clarity the assessed report rewards, and confirm assessment details on Canvas.
Working through Data Journalism in DATA4207? Sia is AskSia’s AI Statistics tutor — ask any DATA4207 Data Journalism question and get a clear, step-by-step explanation grounded in how DATA4207 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
