Start free trial
Trending New Genres Authors
EnglishEnglish
EspañolSpanish
简体中文Chinese
繁體中文Chinese (Traditional)
FrançaisFrench
DeutschGerman
日本語Japanese
PortuguêsPortuguese
ItalianoItalian
한국어Korean
РусскийRussian
NederlandsDutch
العربيةArabic
PolskiPolish
हिन्दीHindi
Tiếng ViệtVietnamese
SvenskaSwedish
ΕλληνικάGreek
TürkçeTurkish
ไทยThai
ČeštinaCzech
RomânăRomanian
MagyarHungarian
УкраїнськаUkrainian
IndonesiaIndonesian
DanskDanish
SuomiFinnish
БългарскиBulgarian
עבריתHebrew
NorskNorwegian
HrvatskiCroatian
CatalàCatalan
SlovenčinaSlovak
LietuviųLithuanian
SlovenščinaSlovenian
СрпскиSerbian
EestiEstonian
LatviešuLatvian
فارسیPersian
മലയാളംMalayalam
தமிழ்Tamil
اردوUrdu
Searching...
SoBrief
Data Visualisation

Data Visualisation

A Handbook for Data Driven Design
Turn messy data into honest visuals that answer real questions without misleading readers.
by Andy Kirk 2016 368 pages
4.07
190 ratings
Listen
Immersive
V2.3
Amazon Kindle Audible
Try Full Access for 3 Days
Unlock listening & more!
Continue

Key Takeaways

Design charts to open understanding, not dictate what viewers conclude

Understanding has stages. Kirk defines data visualisation as giving data a visual form so people can understand it. Viewers first perceive what is shown, then interpret what it means, then decide what it means to them. A chart can help with the first two steps, but the last depends heavily on the viewer’s knowledge, interests, and circumstances. The designer can influence the result, not control it.

One chart, different reactions. A chart of Lionel Messi’s Barcelona seasons shows goals rising to roughly 73 in one season. A football fan can recognise why that is remarkable; someone unfamiliar with the sport may read the values but miss their significance. A viewer who is uninterested may not engage at all. The realistic aim is to make understanding possible, not guarantee agreement or action.

Analysis

What's striking is the distinction between making information available and ensuring it changes someone's mind. That distinction is useful well beyond chart design: public health communication, policy reports, and educational materials all depend on what an audience already knows and cares about. The three-stage model also discourages a common design fantasy, that clarity alone produces action. A chart can explain a risk without giving a viewer the power, resources, or motivation to respond. Its effectiveness should therefore be judged against its actual purpose, not against an imagined universal reaction.

Use a repeatable workflow to make better choices under real constraints

Work in stages. Kirk’s process moves from defining a brief, to acquiring and understanding data, to deciding what to show, and finally to designing the presentation. The sequence reduces a sprawling task into manageable decisions. A chart choice made before clarifying the audience or the question is likely to solve the wrong problem.

Treat sequence as a guide. The process is not a rigid recipe. New information can send a visualiser back to revise the question, gather more data, or change a design. Time, skills, software, and stakeholder demands all limit what is feasible. Kirk compares the pursuit of a perfect solution to a fantasy: a responsible designer aims for the best fit under actual conditions. Sketches, notes, backups, feedback, and careful checking help keep the work moving without mistaking speed for good judgment.

Analysis

The workflow resembles project management in fields such as architecture: clarify the brief before committing resources to construction, then revise as the site reveals its realities. Its strength is not that it eliminates uncertainty, but that it makes the reasons for decisions easier to inspect. The caution is that a process can become bureaucratic if every small chart receives the same paperwork as a public-facing investigation. The useful principle is proportionality: keep the decision sequence, but scale the formality to the stakes and complexity.

Begin with a question, then design for the people who need its answer

Find the origin. A project should start with curiosity: a question, an expressed need, or a promising dataset that might reveal something. Then establish its purpose: whether the result should inform, explain, support exploration, or encourage a response. These choices depend on who will use the work, what they already know, and where and how they will encounter it.

Fit the experience. Kirk’s purpose map helps designers consider how much interpretation to supply and how much to leave to viewers. A newsroom graphic may foreground a finding; an interactive tool may let specialists investigate many angles; a display for an expert audience may need little explanation. The map also distinguishes precise reading from a broader, more emotional impression. Neither is automatically superior. A metaphorical display of wealth inequality, for instance, can convey disproportion without enabling precise value comparisons.

Analysis

This emphasis on audience and task connects naturally to human-centred design, which begins by asking what people are trying to accomplish in a particular setting. It also exposes a tension: designers may know what viewers need before viewers can articulate it, especially when the audience lacks subject expertise. Audience research is valuable, but simply asking people what they want can privilege familiar formats and immediate preferences. Good briefing balances expressed needs with evidence about the task, the subject, and the consequences of misunderstanding.

Interrogate the data before trusting the story it appears to tell

Inspect the raw material. Data rarely arrives ready to chart. It may contain missing records, inconsistent dates, duplicate entries, or unclear categories. Kirk distinguishes detailed, record-level tables from cross-tabulated summaries: the detailed form preserves more options for analysis. He also urges visualisers to identify whether values are text, unordered categories, ordered categories, or measurements, because those differences affect what comparisons are valid.

Ask what it represents. A dataset is not the phenomenon itself. Check who collected it, how it was gathered, which cases are absent, and what transformations or assumptions are involved. Social media posts about an election, for example, cannot stand in for the views of an entire country. Keep unaltered copies, record changes, and disclose important limitations. These habits protect both analytical accuracy and the audience’s ability to judge the result.

Analysis

The book’s insistence on provenance treats data work as an epistemic problem, not just a cleaning chore. A dataset is produced through choices about what to count, whom to include, and how to encode what happened. Those choices can reproduce institutional blind spots even when every calculation is correct. This is especially important when data is used to describe people or communities. Transparency cannot repair an unrepresentative source, but it gives readers grounds to assess its limits rather than mistaking a neat table for a complete account.

Explore widely, because an empty result can still be the finding

Look before deciding. Exploratory data analysis, or EDA, combines visual inspection and statistical methods to discover patterns and test questions. Kirk recommends both targeted investigation and open-ended examination. The first follows a specific question; the second leaves room for an unexpected pattern to redirect the inquiry.

Let surprises earn their meaning. A study of marathon finishing times found noticeable bunching around three-, four-, and five-hour marks. Runners appeared to cluster near familiar time goals, a pattern easier to notice in a visual display than in a simple summary statistic. But a striking pattern is not automatically an explanation. Research the subject and check whether the apparent relationship holds up. Even a lack of change, a blank region, or a missing value may matter if context explains why it is unexpected. If the data does not support the original idea, change the question rather than forcing the evidence.

Analysis

EDA makes an important space for discovery between theory and communication. Visual patterns can suggest questions that a preplanned statistical test would never have raised. That freedom also creates a risk: the more comparisons an analyst tries, the easier it is to find accidental patterns and mistake them for robust findings. The book’s pairing of visual inspection with statistical checking is therefore crucial. Serendipity is a useful engine for generating hypotheses, but it is not, by itself, evidence that a discovered pattern is stable or causal.

Choose the angle, boundaries, and emphasis before choosing the chart

Decide what to show. Editorial thinking is the act of choosing among the legitimate views a dataset could support. It has three practical questions: what angle of analysis is relevant, what belongs inside the frame, and which elements deserve emphasis. These decisions determine the evidence viewers see, not merely its appearance.

Select with discipline. A New York Times article about Peyton Manning’s touchdown record used several views: the history of passing records, the careers of earlier record holders, and performance by age. Each addressed a different part of the question of whether the record could be surpassed. The framing also limited which players and dates appeared. Adding every available comparison would not necessarily make the story stronger. A good selection gives enough context to support the main question without burying it in interesting but unnecessary detail.

Analysis

The photography analogy is powerful because it makes selection visible: every frame includes some things and excludes others. In data work, however, framing can carry political consequences. A chosen time window, threshold, or denominator may change what appears normal or exceptional. The answer is not to pretend that a view from nowhere exists. It is to make consequential choices defensible, explain exclusions where they matter, and consider whether a second view would reveal context that the main frame suppresses.

Match visual encoding to the question, not to the software menu

Choose how values appear. Visual encoding means assigning data to visible properties such as position, length, area, colour, or connection. Different choices support different tasks. A bar’s length is usually easier to compare precisely than the area of a circle, so bars suit close quantitative comparisons while bubbles often communicate only relative scale.

Preserve the truth of the scale. A bar chart should begin at zero because bar length represents the full quantity; truncating its axis exaggerates differences. Circle sizes must be calculated by area, not by simply doubling the diameter when a value doubles. Three-dimensional styling can obscure the quantities rather than clarify them. A map may also distort geographic size, so the projection should fit the comparison being made. The best chart is not the most unusual one, but the one whose visual properties support the intended reading without misleading viewers.

Analysis

The underlying lesson is that charts are not neutral containers. Their visual grammar makes some comparisons easy and others difficult, so choosing a form is a substantive analytical decision. Research on graphical perception helps explain why position and length are often effective for precise comparisons, while area and colour are less exact. Still, maximum precision is not always the goal. A broad overview can be more useful than exact values when the audience needs a quick sense of scale. The key is to match perceptual precision to the decision at hand.

Add interaction only when it gives viewers useful control

Make each action worthwhile. Interactive features can filter data, reveal detail, change views, or guide users through a sequence. They are especially valuable when one screen cannot hold the necessary depth or when different users have different questions. Tooltips can expose exact values without crowding a chart; filters can let users isolate a group.

Do not confuse activity with understanding. A complicated control panel may create work without creating insight. For before-and-after comparisons, viewers may understand change better when both states appear together than when they must remember an animation. Kirk also stresses that every control must function as promised, work in its intended format, and be understandable to its audience. A static graphic can be the more accessible solution when interaction adds no real analytical or communicative value.

Analysis

This is a useful application of the idea of affordance: a control should make clear what action it enables and what consequence follows. Interaction can distribute complexity across user-chosen views, but it can also shift the designer’s explanatory burden onto the viewer. That trade can be appropriate for expert users, but less so for a broad audience seeking a quick answer. Accessibility therefore includes more than the number of options. It includes whether users can discover, operate, and recover from those options without losing the central thread.

Use colour and layout to clarify meaning, not manufacture importance

Give colour a job. Colour can distinguish categories, show ordered values, or direct attention. Different hues work well for separate categories; changes in lightness usually better express a low-to-high scale. A diverging scale can distinguish values on either side of a meaningful midpoint. Avoid a rainbow palette when its shifting hues do not communicate a clear order.

Build a readable hierarchy. Kirk notes that red and green can be difficult to distinguish for viewers with common colour-vision deficiencies, so colour should not be the only carrier of meaning. Use labels, shapes, or patterns as backup. In composition, size, position, and empty space also guide attention. Place related charts near each other when direct comparison matters, and choose scale intervals that make sense to readers. If a coloured or enlarged feature looks important, it should have a reason to be important.

Analysis

Colour and layout function like rhetoric: they tell viewers what deserves attention before any caption is read. That makes restraint an ethical choice as well as an aesthetic one. Redundant cues, such as pairing colour with labels or shape, can preserve meaning across different viewing conditions and abilities. Yet no palette is culturally universal, and even familiar conventions can mislead when applied inconsistently. A design system can create useful consistency, but each chart still needs to be checked against the data, the audience, and the specific meaning assigned to its visual choices.

Build visual literacy by practicing both making and reading charts

Learn from both sides. Viewers can approach a chart by identifying its purpose and data source, learning what its visual encodings mean, then asking what the pattern implies and whether it matters to them. Unfamiliar charts may take practice, but the effort should be rewarded with understanding. Creators, meanwhile, need more than design skill: the work draws on data analysis, research, project management, communication, technical development, and editorial judgment.

Improve through review. Kirk groups these capabilities into seven roles, arguing that a team can cover them collectively even when no single person can master them all. Before release, test accuracy, usability, and unnecessary clutter. Afterward, reflect on what worked, what was constrained, and what should change. Clicks and page views can indicate reach, but they do not prove that people understood the work.

Analysis

The strongest implication is that visualization literacy is reciprocal. A creator who understands how charts are read can design with fewer hidden assumptions, while a practiced viewer can ask better questions about sources, scales, and omissions. This resembles scientific literacy: neither side needs specialist mastery, but both benefit from knowing how claims are constructed and evaluated. The book’s interdisciplinary model also challenges the lone-genius stereotype. Collaboration can improve a project, though only if teams share responsibility for accuracy and meaning rather than treating design, analysis, and communication as isolated handoffs.

Analysis

The book’s central contribution is its insistence that visualization is a chain of judgments, not a styling exercise. A chart is the visible endpoint of choices about a question, a dataset, an audience, a frame, and a visual language. That perspective usefully relocates quality from novelty to fit: whether the graphic is honest about its evidence, usable in context, and composed with care.

Its strongest ethical insight is that truthfulness alone does not guarantee trust. A technically defensible axis, colour scheme, or selection rule can still invite a false impression. Trust therefore depends on both the integrity of the underlying work and the likely interpretation of its form. This creates a productive obligation to anticipate confusion, disclose assumptions, and inspect how visual emphasis changes perceived significance.

The framework also makes a valuable distinction between exploration and communication. Exploratory displays help analysts discover patterns; communicative displays must decide which patterns deserve attention and how to make their meaning available. Confusing these purposes often produces cluttered, unguided graphics. Yet the boundary is porous: exploratory work generates hypotheses, while communication can prompt new questions. Strong practice keeps the two modes connected without pretending they are identical.

A limitation is that the book’s broad repertoire can make the craft seem more settled than it is. Perceptual principles help choose among encodings, but they cannot settle disputes about what to measure, whose experience counts, or which uncertainty matters. Nor does visual accessibility guarantee institutional accessibility: a clear chart cannot compensate for inaccessible data or an audience excluded from decision-making.

The most durable reading is thus both practical and critical. Use process to make choices explicit, use perceptual knowledge to make comparisons legible, and treat every visual display as an argument whose evidence and omissions deserve scrutiny. A good visualization does not eliminate interpretation. It gives interpretation a more reliable place to begin.

Summary in 30 Seconds
Effective visualization begins with a question, audience, and purpose, then tests the data for gaps, bias, and faulty categories. Exploratory analysis should welcome unexpected patterns and null results. Choose visual encodings for the comparison required: bar length supports precision, while area can mislead. Use zero baselines, restrained color, clear hierarchy, and interaction only when it adds useful control.
Contains spoilers

Last updated:

Report Issue

Review Summary

4.07 out of 5
Average of 190 ratings from Goodreads and Amazon.

Data Visualisation by Andy Kirk receives generally positive reviews, praised for its comprehensive coverage of the visualisation process, design principles, and an impressive central reference section featuring 49 chart types. Readers highlight its value as both a learning resource and ongoing handbook. Comparisons to Storytelling with Data and How Charts Lie position it as more technical and wide-ranging. Some critics note excessive text and limited graphs for a visualisation book, but most recommend it for practitioners and data professionals seeking depth over simplicity.

Your rating:
4.73
9 ratings
Want to read the full book?

Glossary

Purpose map

Maps experience and tone

A planning aid for thinking about the kind of viewing experience a visualization should offer and the tone in which it should present data. It helps a designer consider how much explanation to provide, how much exploration to enable, and whether the audience needs precise readings or a broader impression.

Editorial thinking

Selects what the data shows

The process of choosing which legitimate view of a dataset to present. It involves selecting an analytical angle, deciding what data to include or exclude, and determining which elements deserve emphasis. These choices shape the meaning viewers encounter and guide later decisions about charts, annotations, colour, and layout.

Exploratory data analysis

Finds patterns and questions

A way of learning about data through a combination of statistical analysis, visual displays, and subject research. It can test a focused question or allow unexpected patterns to emerge. Its findings help analysts understand what the data may support, but apparent patterns require contextual interpretation and, where appropriate, statistical checking.

TNOIR

Classifies common data types

A mnemonic used in the book to group data as Textual, Nominal, Ordinal, Interval, or Ratio. Textual data consists of unstructured words; nominal data names categories; ordinal data has ordered categories; interval and ratio data are quantitative, with ratio values having a meaningful zero. These distinctions affect valid analysis and visual representation.

CHRTS

Groups chart purposes

A taxonomy for organizing chart types by their main analytical task: Categorical, Hierarchical, Relational, Temporal, or Spatial. It is intended as a practical way to connect the question being asked with a suitable family of visual forms, rather than as a complete catalogue of every possible chart.

Visual encoding

Maps values to visible properties

The assignment of data values to visual marks and attributes. Marks include points, lines, and areas; attributes include position, size, and colour. A chart combines these choices so that viewers can perceive the records and relationships represented. Different encodings vary in how accurately and easily people can compare values.

LATCH

Organizes information five ways

A mnemonic for possible ways to sort information: by Location, Alphabet, Time, Category, or Hierarchy. The appropriate ordering depends on what makes values easiest to find or compare. Alphabetical order can aid lookup, while chronological or quantitative ordering may better reveal change or rank.

Seven hats of data visualisation

Profiles complementary project capabilities

A framework describing the varied capabilities involved in visualization work through seven roles: project manager, communicator, scientist, data analyst, journalist, designer, and technologist. The roles highlight that good work combines coordination, communication, methodological care, data handling, inquiry, design, and implementation, often across a team rather than within one person.

Three stages of understanding

Separates reading from meaning

A model of how viewers engage with a visualization. Perceiving means identifying what the display shows; interpreting means connecting visible values and relationships to subject-matter meaning; comprehending means deciding what that information means personally. Designers can support these stages, but audience knowledge and circumstances shape the outcome.

About the Author

Andy Kirk is a multifaceted creative whose work spans both data visualisation and fiction writing. His debut novel, Thumbs, Beasts & Thieves, marks the first instalment in a trilogy set in the world of Blessaeb, featuring a character named Fee. Kirk's background in crafting compelling narratives for major international brands — spanning products from chocolate biscuits to beer — has honed his storytelling instincts. Inspired by his travels, observations of nature, history, and everyday life, Kirk channels a boundless curiosity and a head full of eclectic knowledge into his imaginative, fantastical stories.

Download PDF

To save this Data Visualisation summary for later, download the free PDF. You can print it out, or read offline at your convenience.
Download PDF
File size: 0.27 MB     Pages: 17

Download EPUB

To read this Data Visualisation summary on your e-reader device or app, download the free EPUB. The .epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.
Download EPUB
File size: 1.26 MB     Pages: 20
Want to read the full book?
Follow
Listen
Home
Swipe
Library
Get App
Try Full Access for 3 Days
Listen, bookmark, and more
Compare Features Free Pro
📖 Read Summaries
Read unlimited summaries. Free users get 3 per month
🎧 Listen to Summaries
Listen to unlimited summaries in 40 languages
—
❤️ Unlimited Bookmarks
Free users are limited to 4
—
📜 Unlimited History
Free users are limited to 4
—
📥 Unlimited Downloads
Free users are limited to 1
—
Risk-Free Timeline
Today: Get Instant Access
Listen to full summaries of 26,000+ books. That's 12,000+ hours of audio!
Day 2: Trial Reminder
We'll send you a notification that your trial is ending soon.
Day 3: Your subscription begins
You'll be charged on Nov 24,
cancel anytime before.
Consume 2.8× More Books
2.8× more books Listening Reading
Our users love us
600,000+ readers
Trustpilot Rating
TrustPilot
4.6 Excellent
This site is a total game-changer. I've been flying through book summaries like never before. Highly, highly recommend.
— Dave G
Worth my money and time, and really well made. I've never seen this quality of summaries on other websites. Very helpful!
— Em
Highly recommended!! Fantastic service. Perfect for those that want a little more than a teaser but not all the intricate details of a full audio book.
— Greg M
Save 62%
Yearly
$119.88 $44.99/year/yr
$3.75/mo
Monthly
$9.99/mo
Start a 3-Day Free Trial
3 days free, then $44.99/year. Cancel anytime.
Unlock a world of fiction & nonfiction books
26,000+ books for the price of 2 books
Read any book in 10 minutes
Discover new books like Tinder
Request any book if it's not summarized
Read more books than anyone you know
#1 app for book lovers
Lifelike & immersive summaries
30-day money-back guarantee
Download summaries in EPUBs or PDFs
Cancel anytime in a few clicks

Why one summary?

Scanner
Find a barcode to scan

⭐
⭐
⭐
We have a special gift for you
Open
38% off
DISCOUNT FOR YOU
⭐
⭐
⭐
$79.99
$49.99/year
Only $4.16 per month
Continue
2 taps to start, super easy to cancel
Settings
General
Widget
Loading...