Information design

03 / Visualise data

03 / VISUALISE DATA

|

Key question

Which representation matches the statement?

Charts reveal patterns that remain hidden in tables. Which visualisation is most suitable depends on what you want to show: a comparison, a trend, a proportion, or a correlation.

Learning objective: You can choose the appropriate chart type for a statement and reduce a visualisation to its core message.

First the statement, then the diagram.

3.1 / From Question to Graphic

Each graphic answers a question.

A visualisation does not begin with the tool, but with the question: What should the viewers understand? From this follows the key message, and from the message follows the chart type.

It is helpful to have a headline that conveys the main finding rather than just the topic: 'Rents rise faster than wages' instead of 'Rents and wages 2015–2025'.

A title is a statement, not a label.

3.2 / Chart types

The message defines the form.

The most common statements and suitable forms:

COMPARISON

Bar

Values of different categories side by side.

DEVELOPMENT

Line

Change over time.

SHARE

Stacked bar

Parts of a whole, or a pie chart if there are few parts.

DISTRIBUTION

Histogram

How values are distributed across a range.

CONTEXT

Scatter plot

Whether two variables are related.

LOCATION

Map

Values by geographic location.

Pie charts are only easy to read with a few slices, as people are worse at comparing angles and areas than lengths.

We compare lengths better than areas.

Try it out
Chart Finder

Choose what you want to show. The tool recommends suitable chart types and highlights common mistakes.

What do you want to show?

I want to compare values of different things.

Recommended

Bar chart

Lengths on a common axis compare most precisely. Sort the bars by size.

Caution

Pie charts and 3D bars: areas and perspective distort the comparison.

3.3 / Reduce

Show less, say more.

Edward Tufte coined the term data-ink ratio: the proportion of ink used to actually display data should be as high as possible. Superfluous grid lines, borders, 3D effects, and decorations (chartjunk) are merely a distraction.

Reducing does not mean hiding information. Highlighting through colour or labelling draws the eye to what supports the message.

If it doesn't show anything, it can go.

Exercise

Visualise a statement

Task: Visualise a small dataset, for example your screen time over a week.

STEP 1

Collect the data and formulate a question about it.

STEP 2

Summarise the key message in one sentence.

STEP 3

Select a type using the chart finder and sketch two variants.

STEP 4

Reduce the better variant and give it a descriptive title.

Evaluation: Do others understand your statement without an explanation?