The FT Visual Vocabulary is at the core of a newsroom-wide training
session aimed at improving chart literacy. This learning resource is
inspired by the Graphic Continuum by Jon Schwabish and Severino Ribecca.
This is not an attempt to teach everyone how to make charts, but how to
recognise the opportunities to use them effectively alongside words.
Read the Chart Doctor feature column for full background on why we made
this: Simple techniques for bridging the graphics language gap
For D3 templates for producing many of these chart types in FT style, see
our Visual Vocabulary repo.
Related reading
The full content of the poster, along with links to related material,
including research and examples of best practice. This is a work in
progress.
General
-
National Geographic: Taking data visualisation from eye candy to
efficiency
-
William S. Cleveland and Robert McGill: Graphical Perception: Theory,
Experimentation, and Application to the Development of Graphical
Methods
-
Hadley Wickham: A Layered Grammar of Graphics
-
Tracey L. Weissgerber et al: Beyond Bar and Line Graphs: Time for a
New Data Presentation Paradigm
-
Numeroteca: Uses and abuses of data visualisations in mass
media
-
Andy Cotgreave: The inevitability of data visualization
criticism
-
Alberto Cairo: "Our reader" won't understand something as complicated
as that!
-
Alberto Cairo: Visualization's expanding vocabulary
Deviation
Emphasise variations (+/-) from a fixed reference point. Typically the
reference point is zero but it can also be a target or a long-term
average. Can also be used to show sentiment (positive/neutral/negative).
Example FT uses: Trade surplus/deficit, climate change
Diverging bar
A simple standard bar chart that can handle both negative and positive
magnitude values.
-
Chart Doctor: How the FT explained Brexit
Diverging stacked bar
Perfect for presenting survey results which involve sentiment (eg
disagree/neutral/agree).
Spine chart
Splits a single value into 2 contrasting components (eg
Male/Female)
Surplus/deficit filled line
The shaded area of these charts allows a balance to be shown – either
against a baseline or between two series.
Correlation
Show the relationship between two or more variables. Be mindful that,
unless you tell them otherwise, many readers will assume the relationships
you show them to be causal (i.e. one causes the other). Example FT uses:
Inflation & unemployment, income & life expectancy
-
Chart Doctor: The German election and the trouble with
correlation
Scatterplot
The standard way to show the relationship between two continuous
variables, each of which has its own axis.
-
Chart Doctor: The storytelling genius of unveiling truths through
charts
-
Maarten Lambrechts: 7 reasons you should use dot graphs
-
Tim Brock: Too Big Data: Coping with Overplotting
-
Sara Kehaulani Goo: The art and science of the scatterplot
-
Chart Doctor: The storytelling genius of unveiling truths through
charts
- Examples: FT
Line + Column
A good way of showing the relationship between an amount (columns) and a
rate (line)
-
Data Revelations: Be Careful with Dual Axis Charts
-
DataHero: The Do’s and Don’ts of Dual Axis Charts
-
Harvard Business Review: Beware Spurious Correlations
Connected scatterplot
Usually used to show how the relationship between two variables has
changed over time.
-
Robert Kosara: The Connected Scatterplot for Presenting Paired Time
Series
-
Data Revelations: Be Careful with Dual Axis Charts
-
Examples: Washington Post
Bubble
Like a scatterplot, but adds additional detail by sizing the circles
according to a third variable
-
Chart Doctor: The storytelling genius of unveiling truths through
charts
- Examples: FT
XY heatmap
A good way of showing the patterns between 2 categories of data, less
good at showing fine differences in amounts.
-
Chart Doctor: Use fewer maps to illustrate data better
Ranking
Use where an item’s position in an ordered list is more important than
its absolute or relative value. Don’t be afraid to highlight the points of
interest. Example FT uses: Wealth, deprivation, league tables,
constituency election results
Ordered bar
Standard bar charts display the ranks of values much more easily when
sorted into order
Ordered column
See above.
Ordered proportional symbol
Use when there are big variations between values and/or seeing fine
differences between data is not so important.
Dot strip plot
Dots placed in order on a strip are a space-efficient method of laying
out ranks across multiple categories.
Slope
Perfect for showing how ranks have changed over time or vary between
categories.
Lollipop chart
Lollipops draw more attention to the data value than standard bar/column
and can also show rank and value effectively.
Distribution
Show values in a dataset and how often they occur. The shape (or ‘skew’)
of a distribution can be a memorable way of highlighting the lack of
uniformity or equality in the data. Example FT uses: Income distribution,
population (age/sex) distribution
-
Joey Cherdarchuk: Visualising distributions
Histogram
The standard way to show a statistical distribution - keep the gaps
between columns small to highlight the ‘shape’ of the data
-
Aran Lunzer and Amelia McNamara: Exploring histograms
Boxplot
Summarise multiple distributions by showing the median (centre) and range
of the data
Violin plot
Similar to a box plot but more effective with complex distributions (data
that cannot be summarised with simple average).
Population pyramid
A standard way for showing the age and sex breakdown of a population
distribution; effectively, back to back histograms.
Dot strip plot
Good for showing individual values in a distribution, can be a problem
when too many dots have the same value.
Dot plot
A simple way of showing the change or range (min/max) of data across
multiple categories.
Barcode plot
Like dot strip plots, good for displaying all the data in a table,they
work best when highlighting individual values.
-
Maarten Lambrechts: Interactive strip plots for visualizing
demographics
Cumulative curve
A good way of showing how unequal a distribution is: y axis is always
cumulative frequency, x axis is always a measure.
Change over Time
Give emphasis to changing trends. These can be short (intra-day)
movements or extended series traversing decades or centuries. Choosing the
correct time period is important to provide suitable context for the
reader. Example FT uses: Share price movements, economic time series
-
Flowing Data: 11 Ways to Visualize Changes Over Time – A Guide
Line
-
The standard way to show a changing time series. If data are
irregular, consider markers to represent data points
-
Chart Doctor: A chart’s ability to mislead is off the scale
-
Office for National Statistics: Does the axis have to start at zero?
(Part 1 – line charts)
-
Quartz: It's OK not to start your y-axis at zero
-
Vox: Shut up about the y-axis. It should't always start at zeroEmily
Schuch: How to Make a Line Chart that Doesn't Lie
Column
Columns work well for showing change over time - but usually best with
only one series of data at a time.
-
Chart Doctor: A chart’s ability to mislead is off the scale
-
Office for National Statistics: Does the axis have to start at zero?
(Part 2 – bar charts)
Line + column
A good way of showing the relationship over time between an amount
(columns) and a rate (line)
Stock price
Usually focused on day-to-day activity, these charts show opening/closing
and hi/low points of each day
Slope
Good for showing changing data as long as the data can be simplified into
2 or 3 points without missing a key part of story
Area chart
Use with care – these are good at showing changes to total, but seeing
change in components can be very difficult
Fan chart (projection)
Use to show the uncertainty in future projections - usually this grows
the further forward to projection
Connected scatterplot
A good way of showing changing data for two variables whenever there is a
relatively clear pattern of progression.
Calendar heatmap
A great way of showing temporal patterns (daily, weekly, monthly) – at
the expense of showing precision in quantity.
Priestley timeline
Great when date and duration are key elements of the story in the
data.
-
Chart Doctor: Communicating with data: Timelines
- Examples: FT
Circle timeline
Good for showing discrete values of varying size across multiple
categories (eg earthquakes by contintent).
Seismogram
Another alternative to the circle timeline for showing series where there
are big variations in the data.
Part-to-whole
Show how a single entity can bebroken down into its component elements.
If the reader’s interest issolely in the size of the components,consider a
magnitude-type chartinstead. Example FT uses: Fiscal budgets, company
structures,national election results
-
Flowing Data: 9 Ways to Visualize Proportions – A Guide
Stacked column
A simple way of showing part-to-whole relationships but can be difficult
to read with more than a few components.
-
Robert Kosara: Stacked bars are the worst
Proportional stacked bar
A good way of showing the size and proportion of data at the same time –
as long as the data are not too complicated.
-
Chart Doctor: How to apply Marimekko to data
Pie
A common way of showing part-to-whole data – but be aware that it’s
difficult to accurately compare the size of the segments.
-
Robert Kosara: Ye olde pie chart debate
-
Robert Kosara: Pie Charts – Unloved, Unstudied, and
Misunderstood
-
Robert Kosara: An Illustrated Tour of the Pie Chart Study
Results
-
David Robinson: How to replace a pie chart
-
Office for National Statistics: The humble pie chart: part 1
-
Office for National Statistics: The humble pie chart: part 2
-
Ian Spence: No humble pie: The origins and usage of a statistical
chart
-
Jeff Clark: In defense of pie charts
-
Stephen Few: Save the Pies for Dessert
Donut
Similar to a pie chart – but the centre can be a good way of making space
to include more information about the data (eg. total)
Treemap
Use for hierarchical part-to-whole relationships; can be difficult to
read when there are many small segments.
Voronoi
A way of turning points into areas – any point within each area is closer
to the central point than any other centroid.
Arc
A hemicycle, often used for visualising political results in
parliaments.
Gridplot
Good for showing % information, they work best when used on whole numbers
and work well in multiple layout form.
Venn
Generally only used for schematic representation
Waterfall
Can be useful for showing part-to-whole relationships where some of the
components are negative.
Magnitude
Show size comparisons. These can berelative (just being able to
seelarger/bigger) or absolute (need tosee fine differences). Usually
theseshow a ‘counted’ number (for example, barrels, dollars or people)
rather thana calculated rate or per cent. Example FT uses: Commodity
production, marketcapitalisation
Column
The standard way to compare the size of things. Must always start at 0 on
the axis
Bar
See above. Good when the data are not time series and labels have long
category names.
Paired column
As per standard column but allows for multiple series. Can become tricky
to read with more than 2 series.
Paired bar
See above.
Proportional stacked bar
A good way of showing the size and proportion of data at the same time –
as long as the data are not too complicated.
-
Chart Doctor: How to apply Marimekko to data
Proportional symbol
Use when there are big variations between values and/or seeing fine
differences between data is not so important.
Isotype (pictogram)
Excellent solution in some instances – use only with whole numbers (do
not slice off an arm to represent a decimal).
Lollipop chart
Lollipop charts draw more attention to the data value than standard
bar/column – does not HAVE to start at zero (but preferable).
Radar chart
A space-efficient way of showing value pf multiple variables– but make
sure they are organised in a way that makes sense to reader.
Parallel coordinates
An alternative to radar charts – again, the arrngement of the variables
is important. Usually benefits from highlighting values.
Spatial
Used only when precise locations orgeographical patterns in data aremore
important to the reader thananything else. Example FT uses: Locator maps,
population density,natural resource locations, naturaldisaster
risk/impact, catchment areas, variation in election results
-
Chart Doctor: Use fewer maps to illustrate data better
-
Matthew Ericson: When Maps Shouldn’t Be Maps
-
Mapbox: 7 data visualization techniques for location
Basic choropleth (rate/ratio)
The standard approach for putting data on a map – should always be rates
rather than totals and use a sensible base geography
-
Vox: The bad map we see every presidential election
-
Vox: This “bad” election map? It’s not so bad.
-
UX•Blog: Telling the truth
Proportional symbol (count/magnitde)
Use for totals rather than rates – be wary that small differences in data
will be hard to see.
-
Stephen Few: What Can’t Be Built with Bricks?
Flow map
For showing unambiguous movement across a map.
Contour map
For showing areas of equal value on a map. Can use deviation colour
schemes for showing +/- values
Equalised cartogram
Converting each unit on a map to a regular and equally-sized shape – good
for representing voting regions with equal value.
-
Chart Doctor: How the FT explained Brexit
-
5W Blog: The power of cartograms and creating them easily
Scaled cartogram (value)
Stretching and shrinking a map so that each area is sized according to a
particular value.
-
Chart Doctor: The search for a better US election map
-
5W Blog: The power of cartograms and creating them easily
-
Vox: The bad map we see every presidential election
Dot density
Used to show the location of individual events/locations – make sure to
annotate any patterns the reader should see.
-
Chart Doctor: The search for a better US election map
Heat map
Grid-based data values mapped with an intensity colour scale. As
choropleth map – but not snapped to an admin/political unit.
-
5W Blog: The power of cartograms and creating them easily
Flow
Show the reader volumes or intensity of movement between two or more
states or conditions. These might belogical sequences or geographical
locations. Example FT uses: Movement of funds, trade, migrants, lawsuits,
information; relationship graphs.
-
RJ Andrews: Picturing the Great Migration
Sankey (aka river plot)
Shows changes in flows from one condition to at least one other; good for
tracing the eventual outcome of a complex process.
-
Chart Doctor: Data visualisation: it is not all about
technology
Waterfall
Designed to show the sequencing of data through a flow process, typically
budgets. Can include +/- components.
Chord
A complex but powerful diagram which can illustrate 2-way flows (and net
winner) in a matrix.
Network
Used for showing the strength and inter-connectedness of relationships of
varying types.
Todo:
Uncertainty
-
Scientific American: Visualising uncertain weather
-
Oli Hawkins: Animating uncertainty
Animation
-
Chart Doctor: The storytelling genius of unveiling truths through
charts
-
Evan Sinar: Use Animation to Supercharge Data Visualization
Interactivity
-
Chart Doctor: Why the FT creates so few clickable graphics
-
Gregor Aisch: In defense of interactive graphics
-
Zan Armstrong: Why choose? Scrollytelling and steppers)
Map projections
Colour