I have a confession to make: my past is paved with chart-making sins, including some capital ones (yes, 3D pie charts, too). But years ago I saw the light in Edward Tufte’s The Visual Display of Quantitative Information
Every God Has His Moses: Edward Tufte and Stephen Few
Some time after that first revelation, I stumbled on Stephen Few’s Show Me the Numbers
Or should I say one and a half?
Edward Tufte and Stephen Few are often cited together, as if they were a single entity. For many of us, simple mortals, Stephen Few is some kind of translator of God’s voice. Given Few’s background, that wouldn’t be completely inappropriate…
For some time that’s how I looked at Few’s work on charts and data visualization. But I was wrong. They do share similar views about basic data visualization principles. And they seem to share the same level of stubbornness, too. But there is a major difference.
Tufte, the Artist vs. Few, the Engineer
Tufte is an artist. His data visualization principles derive from Ludwig Mies van der Rohe’s minimalism, and in that sense, he approaches charts from an aesthetic point of view. His charts are as beautiful as a chart can be, if you happen to like the aesthetic minimalism.
I don’t know how and when Few became aware of the need for better data visualization. But he embraced Tufte’s principles not because he is an aesthete like Tufte, but because he values efficiency and those principles happen to improve it.
Stephen Few would never title a book “Beautiful Evidence”. He doesn’t mind to use Excel to create his chart examples, while Tufte needs full control of details like kerning (and he uses a designer’s tool, Adobe’s Illustrator).
On the other hand, Tufte would never write a book about dashboards (Beautiful Dashboards? brrrr…). From an actionable, business visualization point of view, Tufte is The Visual Display… Almost everything else is beautiful, yes, and perfect for the coffee table.
And while Tufte escaped Flatland for good, Few still keeps both feet firmly on the ground, discussing BI tools, pie charts or irregular time series (and I don’t think his new book
The Need for a New Business Visualization Model: the Emotional Link
Both approaches are very consistent and they give you a set of guidelines that you can apply to all your charts and adopt as a general framework.
What I am not comfortable with is their positivist attitude, specially in Few. Because Tufte’s charts are aesthetically pleasing, we can derive some emotion from that. In Few’s case, his charts are purely functional.
I still don’t know where to draw the line between purely rational/functional visualizations and the eye-candy. Let’s see this pattern:
Boy meets girl, boy gets girl, boy loses girl, boy gets girl back.
Do you feel emotionally overwhelmed? No? Do you even care about the story? Do you even care about the boy and the girl? Let’s try again:
John fell in love with Anna the moment she spilled coffee on his shirt.
This sounds much more interesting. Add three more sentences and you’ll complete the boy-meets-girl pattern. Both versions share the same pattern, but the second one adds some (perhaps irrelevant) detail and creates an emotional link between the audience and the characters.
You need that in data visualization, too. You don’t have to cry because you chart shows a market share drop in Alaska, but you must connect with the reality behind the chart and the data.
The Need for a New Business Visualization Model: Interaction
Jacques Bertin says that knowledge is built by the user when interacting with the chart. Why interaction (and animation) is absent from Tufte’s and Few’s books is something I don’t really understand.
Although I respect Tufte and Few, I feel that there are pieces missing in their theories. We can borrow some pieces from Bertin’s work (and Tukey’s?) and that will surely help, but the real issue here is to find the balance between the need to correctly (bureaucratically?) display the data and the emotional response that helps to keep the audience interested.
Back to you, a very simple question: what are Tufte and/or few missing? What pieces do we need for a XXI century visualization?
You need Tableau.
I’ll give it a try one of these days.
@marc. Tableau is great, but it still just a tool after all. Although Tableau tries to constrain your choices when building visualizations it’s still possible to create bad ones. I think Jorge’s question is about things larger than just tools.
To his point, a good visualization must have impact on the end user in order to create a connection to the data. If the user doesn’t have a connection to the underlying data it’s useless to them. Think about working with tutorial or sample data (e.g. Tableau’s Superstore data) vs. the data that’s most important to you – the data of your business or your study or area of interest.
Visualizations are a crutch for the non-analyst (and even the analyst) to explore the data and their relationships through relatively simple means.
Every dataset has at least one “shape” or “story” to it and in that context visualizations can go down two basic paths – they can show the specific shape that the analyst (e.g. author) thinks is most important OR they can allow the user to explore the many different shapes actually contained within the data.
Static displays, generally speaking, are good at showing a single shape, exploring a single story contained within the data – you can get around this somewhat by using things like small multiples, but still static displays are you usually about a single story/shape within the data – they do not foster exploratory analysis and tend to assume that work has already been done by the analyst and ergo the most important shape is presented.
Can and do we need new types of visualization? Probably, but we’re constrained there on two sides (which are important from an analytical perspective). On the one hand we are constrained by our biology (what kinds of shapes can we gauge accurately). On the other hand we’re constrained by technology – representing dimensions beyond 2 is relatively difficult and not particularly intuitive with our current platforms.
So I think that’s what needed most currently is better tools for exploring data. In other words, the interaction game needs to be stepped up. Where Tufte and Few do a great job of defining the usability of static graphics what are the usability rules for interactive data displays? Allowing users to explore the data lets them find their own story and potentially means that they become more easily connected to the data (and its implications) because it is their own personal discovery. Not to mention that they may find shapes that the analyst never dreamed.
One other thought here, most visualizations don’t do a great job of defining the validity of any particular shape (relationship) – is there a strong or weak correlation? Just because there is a correlation doesn’t mean there’s causality (e.g. not necessarily actionable). So, I think visualization needs to start helping people to understand the validity of any particular shape or story, otherwise we will tend to assume that just because we found a shape that it’s true and important.
After speaking to Stephen, I have a very good feeling about the new book. I think it will bridge the gap between theory and practicality in a very simple and easy to understand way.
“…the real issue here is to find the balance between the need to correctly (bureaucratically?) display the data and the emotional response that helps to keep the audience interested.”
The way I look at it is there is a best practice method to data visualization and a sub optimal approach. Sometimes, depending on the audience, the designer needs to enter the gray area between the two. The real challenge is to train the audience to see why one approach is better. I think the presenter of the data must create in a way that solicits the emotional response specific to their audience. And I don’t think the answer is in software either.
Seriously, we aren’t trying to save the world or walk on water, we are simply trying to present data in a way that enlightens. I think S. Few understands that there is elegance in simplicity and functionality.
@Jorge – thanks! In web usability they use “paper prototyping” as a pre-development test of concepts – not knowing Bertin (I must admit I’m not a scholastically trained infoviz person) it sounds similar except for the intent.
And I agree that the tool should not force it’s logic on you but then you’re dependent on the data informing the logic – which can be just as confusing! Especially without some notion of a known framework (often referred to as a mental matrix or mental model I believe) to play within.
Here’s an ancillary question for you all: Is it possible to effectively democratize information visualization (make a platform that anyone can use correctly with any data for effective exploration of the data) or is it inherently artisan (meaning that only a studied, gifted person can create uniquely effective visualizations that are, by nature, proprietary to the data involved)?
@ Clint: Excellent comment, thanks. “Design for interaction” seems to be absent from the thoughts of many data visualization experts, and that really puzzles me. Excel doesn’t help, either – creating a dynamic chart it’s not easy for a beginner. Tools account for a small part of the problem, specially when they force upon you a logic that is not yours – that’s what Tufte’s essay on Powerpoint is about. But Bertin was discussing dynamic charts in the 60’s, and he was actually doing it with sheets of paper…
@Tony: In a business environment people are paid to pay attention, and I will never lower my standards just to keep the audience entertained, but pure functionality can become counterproductive. Also, often people don’t realize that they don’t know how to manager their data, how to prioritize, filter, etc. (basically Ben Schneiderman’s Visual Information-Seeking Mantra). Even worse, they believe that that’s just a low level issue that can be solved with some Excel formulas.
Nice post. Thought provoking.
I had the pleasure of attending day-long seminars from Tufte and Few within months of each other last spring.
I left the Tufte seminar inspired, with his great books, and thinking as I look around me at the ~500 people that had paid $300 to hear him, “this is a good gig.” Despite that notion, I left thinking that I might have been in the presence of God.
I left the Few seminar also inspired but with with practical advice that I could apply immediately. I will never use a pie chart or 3D graphics without remembering his seminar. I have new insight about what graphic tools work best for a given application. I would equate Few’s role to that of a good pastor who helped me see the light. If he was leading the Church of Practical InfoGraphics, I would likely be sitting in the front row, “born again”, and tithing.
Thanks for the post.
Don
@Jorge @Clint: Great title, great ideas.
I think we can get a lot of mileage out of these biblical analogies.
What we need is opinionated tools. This is something Tableau (apparently) does well and Microsoft Excel is famously bad at–we can’t get rid of pyramid charts even in Excel 2007 guys? Really?
Theoretically, solid metadata could allow a tool to be correctly opinionated and show us the right data. However, that metadata never exists in practice or only exists transiently in the heads of experienced analysts as they are exploring the data.
@Clint: I agree completely that interactions/animation are a blind spot for Few and Tufte. Visualizing significance is underexplored too and is really, really important in a business context.
Jorge,
This is Moses. I appreciate both your interest and critique of my work. I am, of course, always happy to have my work compared to Tufte’s, who in many respects will always be the master. I’d like to respond to a few of your points in an attempt to clarify and further the discussion.
You mentioned that you don’t know when I became aware of the need for better information visualization. As is true of many others as well, I experienced an epiphany while attending one of Tufte’s one-day seminars many years ago. I had been working in the field of business intelligence for many years already, but hadn’t ever thought much about the way we present information. Because of that experience, the seeds of my current work were planted. Contrary to your assumption that I didn’t embrace Tufte’s principles because I was an aesthete like him, my immediate response focused as much on the beauty of his work as its efficiency. I don’t draw a line between beauty and usefulness. To me, there is great beauty in anything that combines elegance and effectiveness. I would love to write a book someday with “beauty” or “beautiful” in the title. Doing so would feed my soul, but for now I’m focusing on the practical skills that most people need—people who would run away in terror if they thought the production of “beauty” was expected of them.
You mentioned that I create my charts using Excel while Tufte insists on Adobe Illustrator, an artist’s tool. My choice to use Excel to create the charts that appear in “Show Me the Numbers” was made for practical reasons, to make the lessons that I teach in the book accessible to a broad audience, even those who rely exclusively on Excel—a vast majority. In truth, I transferred every one of those charts into Adobe Illustrator to improve their resolution and on occasion touch them up a bit before placing them in the book so the images would appear both practical and beautiful in print. Illustrator was also the tool that I used for all of my designs in “Information Dashboard Design.” When no software can produce charts as I want them to look, just like Tufte, I rely on Illustrator.
“Tufte, the Artist vs. Few, the Engineer”? I propose a slight change to characterize the relationship of our work more accurately. “Tufte, the Artist and Few, the Designer.” By nature, I view the world through the eyes of a designer and solve problems primarily from this perspective as well. I appreciate the work of fine engineers and worked as a software engineer myself for several years, but the role didn’t suit me.
“Why interaction (and animation) is absent from Tufte’s and Few’s books is something I don’t really understand.” I can easily explain why this is true of my two existing books. This is because both of my books are about data presentation, not data sense-making. Most of the quantitative messages that people need to communicate to others do not require interaction with the display other than looking and thinking. My new book, due to be released in June, deals with data sense-making, and therefore addresses interaction extensively, as do many of the articles that I’ve written. Interaction drives the process of data analysis.
“The issue here is to find a balance between the need to correctly display the data and the emotional response that helps keep the audience interested.” I don’t believe that tension exists between correct display and emotional response. I believe that if you present information that people need and care about in the clearest and most accurate manner possible, only then can people experience the truth—including the emotional truth—that lives in the data. I welcome anything in a visual display that communicates the story truthfully and with all the emotional impact it deserves.
Take care my friend,
Steve
Steve,
glad to see you (virtually) out and about. I find the distinction between artist and designer interesting…because of the way I have worked with them in my online experience.
In any case, on to why that’s interesting. 1) it’s a very subtle distinction and if someone is not “in the know” they may not appreciate it and even those in the know may understand the differences … differently. For me, when we’re talking about artist and designer; the artist takes the role of blue sky creator – the creative force behind the vision of a website for instance – while the designer takes the role of executing that vision in the ‘real world’ – something that will actually function within a boundary of constraints that the artist is not usually worried about.
By way of another analogy, if the artist is the architect, the designer is the construction (could be a civil engineer right?) contractor on hand to bring the architect’s vision to bear (is that an analogy I lifted from you? I know I read it some where). So would I tend to think designer = visual engineer in Jorge’s analogy.
Don’t you just love semantics? :~)
-Clint
Great post. And, “Jorge, This is Moses.” what a response!
It’s impossible for me to come up with an idea that would make the best visualizations even better; I’m not publishing books or writing whitepapers on the topic. But visualization as a whole would also be better if bad examples were less abundant.
Excel is by far the most accessible tool and is responsible for spreading bad practice. Chris’ point about removing pyramid charts and products like BonaVista’s Chart Tamer approach the problem by making the application better. Though Microsoft doesn’t get everything right, they collect wide information on feature use (their Customer Experience Improvement Program) and I bet there are real people out there who would be upset if their beloved Pie of Pie chart disappeared.
Books like Show Me the Numbers and examples like Jorge’s Demographic Dashboard approach the problem by educating the application’s users. I’ve been to Tufte’s one day course and loved it. But Tableau, Illustrator, seminars, or books on theory before practice are probably too much of an undertaking for most – investment is a barrier to education.
Few’s Visual Communication whitepaper was what first got me hooked on the subject because it explained theory with a focus on business data. This blog and the others on the subject do the same. Tufte the Artist is inspiring, but most of us make charts as part of our professions, not for the aesthetics. Anyone can make better visualizations with readily available tools and quick lessons from the experts.
Better tools and interaction sound great for making the best better, but spreading digestible bits of education will make the biggest difference for the widest audience.
Great discussion. I think a good distinction between an Artist and a Designer was given by Alan Cooper in his book About Face 3:
“The goal of the artist is to produce an observable artifact that provokes an aesthetic response. Art is a means of self-expression on topics of emotional or intellectual concern to the artist and, sometimes, to society at large.”
“Designers, on the other hand, create artifacts for people other than themselves. Whereas the concern of contemporary artists is primarily self-expression, visual designers are concerned with clear communication.”
I greatly admire the works of both Tufte and Few. Each of their contributions to this field have been enormous. I have been fortunate enough to attend both of their workshops. Both were excellent but both were radically different.
Tufte’s was like a day at the theater of data visualization. It was informative, inspiring, and beautiful. And when it was over, I left feeling energized.
Stephen’s, on the other hand, was more like a three day data visualization and analysis boot camp. It was a non-stop mix of class room work, breakout sessions where you had to complete assignments with a work group, and Q&A sessions, but most of all, it was filled with practical information you could immediately use in the business world.
Clint,
Thanks for pointing out that the terms “engineer” and “designer” mean different things to different people. Designers, as I define the term, determine how something should look and function to perform a particular task. Engineers figure out how to make things look and function as designed. Engineers implement designs, using them as a blueprint, transforming them into something real. In the realm of software, although some people can do both, the roles are distinct. Designers use images and words to describe how a product should look and function and then engineers write code to make it function as described. In my work, I create the blueprints; I rarely build the house.
Steve
For me the essence of Jorge’s original blog is that we need to do more storytelling and be less factual. Just think about it: to whose presentation would you rather go the statistician or the storyteller? There has to be a balance off course. Somebody who brilliantly combines the facts with a story is Jonathan Jarvis (see http://vimeo.com/jonathanjarvis) in his depictions of the Credit Crunch and the Stimulus Package. Like Stephen Few he takes a design view to present information but he adds more, into something he calls Mediators (media+information). From his website I quote
“New Mediators are practitioners who combine methods from design, journalism, and narrative analysis. The result is designed transparency — information that is not only made available, but accessible, relevant and beautiful. ”
He creates small movies and in them breaks down complex information into images. I wonder if this approach could be used to create dashboards that are neither static nor interactive but story telling. To me thinking along these lines would be the way forward in dashboard development. Instead of presenting only the facts you tell your users what they are actually looking at and why it is important. After having done that you could still give them options to further investigate the data. I believe that this approach has some advantages: it helps explain things better, it is more interesting to look at and it will help to memorise important facts.
Anton Roodhuijzen
Novice to all this… but in answer to Jorge’s question about what Tufte is missing, I first thought “taking a position: on why the data matters, or it’s consequence”, from a business perspective. Then I thought about Charles Joseph Minard’s “War of 1812” graphic that Prof. Tufte presented in his seminar. To me, what makes that most powerful is the way the consequence sinks in to the viewer. It is not forced, no diatribe, it is itself a statement. If it had been explained, or bulleted it would totally diminish the impact. Tufte says it is the first anti-war graphic, while not even mentioning Napoleon.
Ever heard about Mr. Hichert from Switzerland? He developed a concept called “SUCCESS”. The main issue is to tell a story. Visualization help telling it, but if there is nothing to tell nothing will help you. He trained thousands of business people during the last ten year in Germany, Austria and Switzerland. If you call Stephen Few an engineer, Mr. Hichert is a “double-engineer” – his concept is very well structured and he developed rules that makes quantitative business documents very easy to read. Scaling is one of his vavorite topic. So e.g. the Swiss Post used a scale in their official anual report (something like 1centimeter is 10 Million CHF). It like it very much!
Much fun reading Mr. Hicherts articles an examples http://www.hichert.com
Have fun
Joerg
Joerg: I know Mr. Hichert’s work and I know that some of the readers here are Hichert’s die-hard fans. Actually, his office was kind enough to send me a copy of his poster (it’s hanging on my wall). Unfortunately, his work is not known as it should in the English-speaking visualization community. I’d like to help change that.