Getting Acquainted with a Data Set

The other day my team was just kicking off a new business intelligence dashboard project and the client’s data modeler and DBA led us through a through of the data landscape. We were knee deep in entity relationship diagrams, data dictionaries and all kinds of data “stuff” that made my eyes glaze over.

I subsequently discovered Stephen Few’s fabulous article entitled “Exploratory Vistas: Ways to Become Acquainted with a Data Set for the First Time” and I thought to myself, “OMG, I should have kicked off the meeting with a joint review of Stephen’s approach to getting to know the data!”.

Here is the link to the pdf: Stephen Few: Exploratory Vistas

The intro:

When you lay eyes on an unfamiliar territory for the first time, it always makes sense to get an overview before venturing into the thick of it. An unfamiliar data set is like an unknown land. Unless you have unlimited time to wander (who does?) and don’t mind getting lost, it helps to study a map before starting out. If no map exists, then you should head for high ground to get the lay of the land.

Analytical journeys are quite different from vacationing in Italy or France. Leisure travel involves a series of destinations with the hope of enjoying ourselves along the way. We want to add sweet memories to our lives of the great meals, beautiful sites, and interesting people we meet along the way. When we embark on an analytical journey, however, we don’t pre-plan all our destinations and we’re not just collecting memories. The goal of the analytical journey is a thorough understanding of the territory, often to solve specific problems. We approach the journey as if we’re preparing to become tour guides, able to explain each site in a way that ranges from its history to predictions about its future. We must learn to navigate like a native.

I really like the part about using treemaps. Here’s a look at a Panopticon treemap:

Be sure to check his pdf for more!

The Dashboard Spy

Stephen Few Judges Infographics Contest

Business intelligence visualization expert Stephen Few was one of the judges of an international competition for journalistic infographics. He and 9 other brilliant judges spent over 3 days poring over the merits of 1300 examples. They awarded 7 gold medals, 25 silver medals and 70 bronze medals for infographic excellence. Alas, they did not produce any infographic detailing the results nor showcasing the winners. (Now, THAT’S a great idea for an infographic!

Take a look at the brainpower of this panel of judges:

Stephen Few judging infographics

Stephen’s blog post, Malofiej 19 in Review, details his experience in depth, so be sure to visit.

He reveals that the top prizes include a Best of Show for the Print category done by National Geographic for “Gulf of Mexico: A Geography of Offshore Oil”.

Here’s a peek at the winner:

Gulf of Mexico Geography of Offshore Oil

And here’s my favorite, the Best of Show for the Online category: New York Times for “How Mariano Rivera Dominates Hitters“. Click that link to watch the amazing video.

mariano rivera pitching infographic

Be sure to visit Stephen’s blog for the other winners.

Hubert Lee
The Dashboard Spy

Taking Data Chart Minimalism Too Far

Show the data and nothing but the data, right? We’ve all seen the trend towards minimalism in our charts and graphs. Sometimes this results in business dashboards populated by mysterious and tiny scratchings that only a true data visualization expert can understand.

In the recent issue of his Visual Business Intelligence Newsletter: Sometimes We Must Raise Our Voices, Stephen Few explains what he calls a “rare disagreement” with the principles of graph design as espoused by Edward Tufte (See The Visual Display of Quantitative Information
).

Few begins with an explanation of the title of this issue:

When we create a graph, we design it to tell a story. To do this, we must fi rst fi gure out what the story is. Next, we must make sure that the story is presented simply, clearly, and accurately, and that the most important parts will demand the most attention. When we communicate verbally, there are times when we need to raise our voices to emphasize important points. Similarly, when we communicate graphically, we must find ways to make the important parts stand out visually.

He then tells the importance of Edward Tufte’s work on his own career path.

My original thinking about graph design was formed almost entirely by the work of Edward Tufte. I owe him not only for the formative development of my knowledge, but also for inspiring me to pursue this line of work in the first place. I left his one-day seminar over 10 years ago with my mind ablaze and my heart beginning to nourish the kernel of an idea that eventually grew into my current profession. Even after many years of working in the field of data visualization, which has involved a great deal of experience and study that has expanded my expertise into many areas that Tufte hasn’t specifically addressed, I have only on rare occasions discovered reasons to disagree with any of his principles. The topic that I’m addressing in this article, however, deals with one of those rare disagreements.

Now, isn’t that an interesting setup? Steve’s writing is always superb and the reason why I subscribe to his newsletter. Visit the Perceptual Edge site for details.

Getting back to the disagreement with Tufte, Few has some issues with carrying the idea of minimalism too far. He explains Tufte’s concept of keeping the Data-Ink Ratio high and minimizing Non-Data ink on a chart.

Data Ink Ratio

There are 2 graph design principles closely related to the Data Ink Ratio concept.

Erase non-data ink (within reason)
Erase redundant data-ink (within reason)

Few is in complete agreement with the first rule, but has problems with the second rule about erasing redundant data-ink.

He presents a series of charts based on Tufte examples where he shows the danger of taking the minimalist approach too far. Each successive chart shows an increasing level of minimalism (is that me speaking in an extremely backassward way or what?)

Ed tufte chart

Tutfe Few Graphs

Tufte minimal chart

Stephen Few demonstrates Tufte Data Ink Concept

and here is an alternative by Stephen Few:

Clear data visualization by Stephen Few

Read his post for his explanations of why Tufte’s charts may take data-ink redundancy removal too far.

Tags: Data Visualization, Stephen Few, Edward Tufte

Dashboards