Better Data Visualizations
April 27, 2021

How a wider repertoire of charts gives Better Data Visualizations

By Paul Laughlin

The latest book to join the pantheon of my recommended Data Viz texts is “Better Data Visualizationsfrom Jon Schwabish. This one was keenly awaited. I’d heard Jon’s updates on progress for months beforehand. He was even a guest on our podcast while he was still working on it.

So, I’m delighted to say that it has proven well worth the wait. It also nicely fills a gap, such that it complements those I have recommended previously. It is also clearly laid out and has a pleasing symmetry. I recommend it highly. Especially for analysts seeking to expand the repertoire of chart types that they use to visualize data.

Jon is the founder of the data visualization and presentation skills firm, PolicyViz. He’s also a Senior Fellow at the Urban Institute, a nonprofit research institution in Washington DC. Jon’s a generous blogger, podcast host, and producer of data viz education tools. So, I was not surprised to find this book packed with practical examples and advice analysts can apply at work.

Structuring a Data Visualisation education

As a trainer of Data Visualisation myself, it is always interesting to see how others approach the subject. Each person has their own unique style. Cole Nussbaumer Knaflic focuses on storytelling in a business context and decluttering your graphs to focus the eyes on the key messages. Andy Kirk focuses on both a thorough grounding in theory and application of best practice through each step of a consistent workflow (as well as real depth as a reference work).

Perhaps not surprisingly given Jon’s role, he structures his book to provide a simple overview of key principles needed and then spends most of his time expanding our repertoire of charts. He uses the 3 parts of this book to clearly structure his material:

  • In Part One he summarises visual processing, guidelines for chart design, and how understanding your audience should guide your choices.
  • Part Two reviews a wealth of different chart types under 11 groupings, including most of those highlighted in his famous Graphic Continuum visual summary.
  • Finally, Part Three covers Data Visualisation Style Guides and example makeovers (case studies in improving charts).

There are also two bumper appendices that provide advice on data viz tools and plenty of CPD resources for your further study.

So, let’s review in a bit more detail how each of those parts can help you improve your Data Visualisation craft.

Part One: Principles as groundwork for the rest of this book

After a nicely personal introduction (so you can relate to Jon’s own learning journey) he manages to condense a surprising amount of important theory into only 53 pages. These include research on perceptual rankings (most effective visual channels), Gestalt visual perception principles, Anscombe’s quartet, and preattentive processing examples.

Building on this, Jon artfuly simplifies a range of best practice design advice into 5 overarching principles:

  1. Show the Data (prioritisation, clarity on insight/message and drawing focus to the key message)
  2. Reduce the Clutter (worked examples of how to do that, with many more later in the book)
  3. Integrate the graphics and text (from the need for active headings to appropriate use of annotation)
  4. Avoid the Spaghetti Chart (from use of Small Multiples to recognising how changing chart type can help)
  5. Start with Grey (selective use of colour to draw the eye to the most important data in meaningful context)

The final chapter of this part is entitled Form and Function. As well as covering that design principle, this chapter considers the multiple need for visualizing data. It is a helpful reminder that although most of this book focuses on static explanatory charts, data viz has a clear role to play for both exploratory work and increasingly through interactive visualizations. This latter point is also considered during the book from time to time.

Part Two: More Chart Types than you can shake a stick at

Ok, if you are reading this blog you probably have at least a passing interest in data visualisation. You may think you already have a wide knowledge of chart types. Well, I guarantee you will find at least one example you’ve didn’t know before in this section. It is the longest collection of different chart/graph types that I have so far seen in print. A great opportunity to expand your graphicacy and to be prompted to test new types on your audience.

This part is broken down into 10 different types of data, showing potential chart types that could be appropriate. Plus, Jon has added a chapter on better table design, for when that level of detail is appropriate to visualise.

Here is what I mean by the wealth of examples that Jon shows, including a supporting explanation and design advice for each chart type. For each of these different types of data types he includes this many examples:

  • Categorical data/comparing categories (17 different types)
  • Time series data (14 different types)
  • Distributions of data (9 different types)
  • Geospatial/Geographical data (5 different map types, with multiple variants of each)
  • Relationship data (8 different types)
  • Part-to-Whole data (5 different types)
  • Qualitative data (9 different visualisations, a topic often missed in other texts)

I mentioned the final chapter in this section covers tables. This is very helpful with examples showing how you can transform an impenetrable table into an effective visualization. This includes two walkthroughs of data table redesigns and Jon’s 10 guidelines for better tables. Given how often analysts (and other technical teams) still need to include tables of data this is an important aspect to not overlook.

Part three: Finishing in style

Like an expert author, Jon closes this book by weaving together themes he has introduced and been playing with throughout. The first of those is a topic that Jon has been championing on his blog and podcast for some time. The need for Data Visualisation style guides in organizations. But, rather than have this as a block of theory at the end of the book, he has been showcasing style guides throughout. At this stage in reading, you recall that each of those different chapters on chart types introduced a consistent style guide for each chapter. So, you have already seen how this can provide a consistent visual impact and aid comparison.

In this chapter, Jon deconstructs the anatomy of a chart and then walks through how a style guide can ensure consistency and professional presentation for each element. He covers colour palettes (building on the colour theory embedded earlier in the book). Jon also covers style guidance for Fonts, Chart Types, Exporting Images, Accessibility and Diversity. So, this is so much more than just a template PowerPoint deck. I heartily recommend both reading this chapter and putting in the work to develop your own in-house Data Viz Style Guide.

In line with the practical style of this whole book, Jon’s closing chapter is on the topic of redesigns or makeovers. He includes eight examples of improving on data visualisations or tables through applying the principles and chart options from earlier in the book. In this way, this closing chapter is also a pleasing weaving together of earlier threads. We are reintroduced to the power of some earlier praised chart types, like dot plots, scatter charts, line charts, and heatmaps It really helps ground in practical examples the thinking needed to improve.

How will you develop Better Data Visualisations?

Well first, I recommend that you buy this book. Plus, if you are interested in a particular chart type that you have not used before, it is well worth checking out the related episode in Jon’s #OneChartAtATime YouTube series.

But, data visualization is a craft rather than a subject learned by rote. So the real learning happens in your own practice. I hope this book encourages you to experiment more and to give time to revisit and hone your charts.

One of the other things to praise in this book is Jon’s honesty and humility in making clear where there is no one right answer. In many examples, he brings to life how much it is the audience, insight, and your own aesthetic that will guide which chart type you use. His principles can help you avoid pitfalls, but you also need to get your eye in. I hope this beautiful and enjoyable book encourages you to do just that as you expand the repertoire of chart types in your palette.