Communicating with Data

Analysis that no one understands changes nothing. The final, easily skipped step of data work is communication: taking a result you trust and making it clear, honest, and usable to people who did not do the analysis and will not read your working. A sound finding buried in a confusing chart is, for practical purposes, no finding at all.

Communication is also where honesty is won or lost. The same numbers can be framed to enlighten or to mislead, often without a single false figure. A truncated axis, a cluttered design, or a cherry-picked window can bend a true dataset into a false impression. Communicating well means choosing forms that let the data speak plainly.

Match the chart to the question

The right chart follows from what you are trying to show, not from what looks impressive. Comparing categories calls for a bar chart; showing change over time suits a line; revealing the relationship between two variables suits a scatter plot; showing a distribution suits a histogram. Pie charts strain once there are more than a few slices. Ask what comparison the reader must make, then pick the form that makes that comparison effortless.

Honesty in framing

A chart makes an argument, and small choices change what it argues. Starting a bar chart's axis above zero can exaggerate a tiny difference into a dramatic one. Choosing an unusual time window can hide a real trend or invent a false one. Using inconsistent scales, or area to stand for a single number, distorts proportion. Honest framing means letting the reader see the data on a fair scale, disclosing the axis and the source, and resisting the design that merely flatters your conclusion.

Cut the chartjunk, tell the story

The statistician Edward Tufte coined the term chartjunk for decoration that adds no information: heavy gridlines, three-dimensional effects, needless colour, clip-art. Every such element competes with the data for the reader's attention. Strip them away so the signal stands out. Then give the chart a job: a clear title stating the takeaway, labels where the eye lands, and one message per view. A good data story leads the audience from question to evidence to what they should do next.

Imagine reporting that a support team cut average response time from 10 hours to 9 hours. On a bar chart whose axis starts at 8.5 hours, the second bar looks about half the height of the first, suggesting a stunning collapse. On an honest axis starting at zero, the change is a modest, truthful step down. Same data, opposite impression. The honest version might pair the bar with a single sentence, 'response time fell 10 percent this quarter', so the reader leaves with an accurate takeaway rather than a manufactured shock.

Simplicity can also mislead when it hides necessary context. A clean line showing rising sales, stripped of the fact that a competitor collapsed or that prices doubled, tells a tidy but incomplete story. Honest communication is not merely minimal; it is minimal plus the context a fair reader needs to interpret the number. Removing clutter is a virtue only when what remains still tells the truth. The aim is clarity without distortion, neither decoration nor careless omission.

The Swedish physician and statistician Hans Rosling (1948–2017) became famous for making dry statistics vivid. In his 2006 TED talk, often titled 'The best stats you've ever seen', he used animated bubble charts to show how countries' life expectancy and income changed over two centuries, dismantling the tidy myth of a fixed divide between so-called developed and developing worlds. The animation let the audience watch nations move across the chart, so time and change became visible rather than abstract.

Rosling co-founded the Gapminder Foundation in 2005 with Ola Rosling and Anna Rosling Ronnlund, and their Trendalyzer visualisation software was acquired by Google in 2007. His approach was not to dumb data down but to design it up: pick the comparison that overturns a misconception, animate it so the pattern is unmistakable, and narrate what it means. His 2018 book Factfulness carried the same conviction, that clear presentation of real data can correct even widely held false beliefs.

• 'A fancier chart communicates better.' Decoration usually competes with the data; plainer designs read faster and truer.
• 'If every number is accurate, the chart is honest.' Framing, axes, and time windows can mislead using entirely true figures.
• 'Communication is separate from the real analysis.' A finding no one understands or acts on has no effect; presentation is part of the work.
• 'Simpler is always more honest.' Stripping necessary context can mislead as badly as clutter does.

You are handed a chart with a truncated axis, three-dimensional bars, and a vague title. Rebuild it into an honest, clear version and state the one-sentence takeaway it should deliver.

Think Like a Maester: A chart is an argument, so make it an honest one by showing the data on a fair scale and letting the shape, not the decoration, do the talking.

1. How do you decide which chart type fits a given question?
2. How can a chart mislead while using only accurate numbers?
3. What did Edward Tufte mean by chartjunk, and why remove it?
4. How did Hans Rosling make statistics vivid without distorting them?
5. Why can an over-simplified chart be as misleading as a cluttered one?

Communicating with data turns a trustworthy analysis into something an audience can understand and act on. Choose the chart that makes the necessary comparison effortless, frame it on a fair scale, and remove the chartjunk that competes with the signal. Guard honesty at every step, because true numbers can still be arranged into a false impression. Hans Rosling's work shows the upside: clear, well-designed data can overturn misconceptions and move people. The goal is always clarity without distortion, a plain and truthful path from question to evidence to action.

• Misleading Statistics
• Becoming Data-Literate