Plowing the Dark

Here's a description of an economics simulation made visible in a virtual reality room:

A crimson comet, at ten o'clock, just above the horizon, paints an upturn in third-quarter commodities. A rose of starbursts means stubborn unemployment.
Hidden relations spill out, suddenly obvious, from a twist of the tabular data. Tendencies float like lanterns across the face of a summer's night.
...
In this room of open prediction, facts flash like a headland light. The search flares burst around you where you stand, lost in an informational fantasia: tangled graphical dances of devaluation, industrial upheaval, protective tariffs, striking shipbuilders, the G7, Paraguay, Kabul. The sweep of the digital – now beyond its inventors' collective ability to index – falls back, cowed by the sprawl of the runaway analog. Five billion parallel processors, each a world economy, update, revise, negate one another, capsize the simulation, pumping their dissatisfied gross national product beyond the reach of number.

(From Plowing the Dark by Richard Powers)

This is from a book I'm reading in which the author imagines the power of visualization which may soon be available through further advances in computing technology. The beauty of this is that it is a NOVEL, not a dry technology forecast, or glossy marketing brochure. The power of storytelling triggers our emotional and imaginative response, allowing us to connect with this future as a reality. It reminds me of Neal Stephenson's Snow Crash, in which people interacted in the online "metaverse" via "avatar" personas – a vision of Second Life ("SecondLife is a 3D online digital world imagined, created, & owned by its residents"). While Plowing the Dark is strictly in prose (no graphics), it paints a vivid picture of the power of visual representation to extend our awareness and understanding. I'm not far enough into it to know whether he explores the potential of the technology to further our wisdom, or explores its downsides as well as its benefits. I suspect he examines both sides, as there was already an example of one programmer developing a black eye from a "collision" with a virtual branch!

This book illustrates the principle behind scenario building. A vivid, multi-sensory, plausible scenario causes us to feel the impact of a situation, motivating us to take steps to bring it about, in the case of a desirable outcome, or prevent its occurrence, in the case of an undesirable future.

By the way, you don't have to be a computer geek to enjoy this book. The technical details are described at a high level as an artist character seeks to understand her role in the project, and to contribute to it, and to explain the excitement and dedication (or obsession!) of her fellow researchers.

The bottom line: Whether through story, or scenarios, or graphical representation, visualization is extremely powerful for an individual or group to create a desired outcome, whether that is a new product, a solution to a tough problem, or a better world.

Ram Dass

I saw Ram Dass speak this weekend.* He was not much more than a famous name to me before, but I wanted to take advantage of the opportunity to see for myself why he was famous, and whether he had anything to share that I might want to learn.

His manner of speaking was unusual. Most speakers rush to fill every second with words. His rate of delivery was glacially slow, maybe 10 words per minute sometimes. I wondered at first if this were due to some age- or disease-related issue, or deliberate. I wondered if he was just making this up as he went along, and waiting for the words to come to him. I had time in between words to wonder about a lot of things, and to think about the actual words and their meaning, and why he was giving each one so much importance. I think it was a way of letting us experience living in the now, living each moment, really hearing each word. He never said, "Just live in the now", though he wrote a book called Be Here Now, which I assume conveys that message. He created a space for us to actually experience it. This is what the very best teachers do: design an experience so students can create their own learning. This is what I aspire to do in my teaching and facilitation. There were other master teachers at the event, who used their manner of presentation (body language, rate of delivery, repetition, emphasis) to deliver their message more powerfully than through their words alone. I have a lot to learn.

* See my post on HonuaOla, below.

HonuaOla

This weekend I attended an event called HonuaOla*, subtitled "Earth, life, and all things sacred". When I tried to describe it to my husband and friends before I went, I had a hard time, even though I'd read some publicity on it. My description sounded vague and new-agey, and didn't explain why it seemed compelling to me. But after reflecting on it afterward, and reading more about it on their website, I felt they'd been remarkably successful in conveying a single message through many voices and media, and that that was part of their message. We are diverse, and each have to craft our own journey through life, take action, celebrate life, and support others in their journeys. This sounds cliche, but experiencing it was profound.

The organizers requested that each attendee fill out a survey. I did, but felt my brief response was superficial and didn't capture the depth of the event's impact on me. This post addresses my need for a more thoughtful response.

The event was very well organized. The presenters and performers were excellent, and it was professionally run. The location, the Maui campus of Kamehameha Schools, was stunning. I liked the main stage format, with its variety and liveliness, and the separate panel discussions, which went into more depth, but again from multiple perspectives.

On the main stage, short performances and lectures alternated, a format which worked well. I've been to similar events, such as Bumbershoot in Seattle, where the genres are segregated. This may be due to the vastly different scale of the two events, but for HonuaOla created serendipitous reinforcement of their theme, as, looking through the schedule, one might have identified some particular draws, and then found interesting parallels or new ideas while sitting through a couple performances they might not otherwise have chosen.

One aspect that didn't feel as well integrated, though, was the "non-profit row". Various non-profit groups had tables set up in the shade of buildings surrounding the quadrangle between the stage and the panel discussion building. It felt like a type of ghetto, segregated: one had to have a reason to go there, and then felt trapped. If they were trying to reach people who weren't already familiar with or sympathetic to them, they needed a way to interact with them, to draw them in. Maybe each could have a slot on the speaking/performance schedule, or just a teaser in the announcements between acts, like when a puppy was featured from the Humane Society. Another way to improve the already excellent event would be to offer more ways for the attendees to interact with the presenters, like associated short workshops or seminars, to make connections with each other and the presenter, and put their learning into action – which was one of the main messages of the event.

* In the Hawaiian language, Honua = earth, Ola = life

Testing text & image placement

There's no interesting info here today, I'm just testing!



Text that I want to be to the right of the first image.




Text that I want to be below the first image but above the second one.



Text that I want to the right of the second image.




Text that I want following the second image.

An example of analytical design in action

A few days ago I made some graphs of data my husband and I have kept for the past three years, of the number of days we've windsurfed. What you see below is what I came up with. (Click on any image to see a larger version. Some are PDFs.)

Yes, keeping track of this kind of detail this might seem pretty anal. We're engineers. But these records help us check our instincts when we comment that this month seems better, or worse, than the same month in past years. However, the original charts were becoming unwieldy. We'd posted them on our refrigerator, and they were fading, and took up too much space.

I'm writing about this here because I thought that walking through my thinking, and the process I followed, would be helpful in illustrating the practical aspects of the principles I've been studying for the past few years. Anyone who needs to analyze and report data, whether for a high school science report, a professional journal article, or a business presentation, faces a similar process. In any activity, we encounter choices. Knowing the principles relative to the activity allows us to make informed choices. In this case, the choices should lead to a graphical product that its intended audience finds easy to use.

Analytical design

What do I mean by the term I used in this post's title, analytical design? It is the phrase used by Edward Tufte to describe graphics created to convey information to others. Other terms used for it are information design and information architecture.

In a series of four books, Edward Tufte has compiled examples of good and bad analytical design, and extracted principles that guide it. I won't try to summarize them all here, but will illustrate some of them as I walk through this example. They guided the choices I made in creating my design.

What is the thinking task?

The first question to ask yourself before starting an analytical design project is "What is the thinking task this display is supposed to help with?" Put yourself in the viewer's head and imagine what questions s/he would be trying to answer when looking at your finished design. In this case, I knew exactly what the viewer would be looking for, since I am that viewer. The questions I would be asking of this data are things like:

• How many sailing days did we have last March (say)?
• How does that compare with other Marches?
• Can we expect to sail more in April than in March?
• If our friends ask when is the best time to come to sail, what should we advise them?

This is the thinking task that my information display will address.

Consider design options

The original charts, by the density of the colored-in days, provide general "at-a-glance" answers, and our monthly tallies provide the numerical values, but with three years of data we're probably about at the limit of what we're willing to wade through. A more concise summary would be nice. A numerical table

is compact, but hard to comprehend at a glance. A graph would communicate comparisons more quickly. I sketched out three different ways of grouping the data in graphs before settling on one. (I also recognized that in summarizing, I lose some data from the original charts: I won't be able to answer questions such as "What was the longest run of consecutive sailing days that K had?". If this kind of question were important, I would make different choices and come up with a different design.)

Refine the design

Then I used Excel's graphing options to create graphs for each year. My job should be done now, right? Well, not exactly. The Excel charts are quick and easy, but require some tweaking to be most useful. The design should facilitate the thinking task. The default graphs

are rather garish and overdone, calling attention to insignificant details while obscuring more important ones. Comparing the default design

with my variation,

consider these questions:

• Which one is easier to read?
• Which looks more professional?
• Which would you rather present as your work?

I hope you agree that my design is an improvement over the default. (And if not, perhaps you could let me know why.)

One of Tufte's principles is to let your data speak. Another is to maximize the data-ink ratio. In other words, don't let non-data elements of the graph, like the background or the grid, overwhelm the data. With these principles in mind, I made the following changes to get from the default design to my final version:

• Remove gray background - improves contrast & readability
• Remove outline around plot area - unnecessary non-data ink
  
• Change colors of data sets to be more restful (easier to look at) and convey consistent coding (blues for J, purples for K)
• Remove outlines around vertical data bars - unnecessary
  
• Change symbols for data points - clearer
  
• Add value labels for data bars and year-end totals - informative
  
• Standardize vertical scale ranges between years (left axis 0-30, right axis 0-200)
• Move legend inside plot area and remove outline around it - more usable
  
• Remove bold effect on chart and axis labels - unnecessary
  
• Scale down font size of chart label (title)
• Tone down axes by changing from black to gray
• Quiet Y-coordinate labeling by increasing size of major divisions
• Tone down outline around each chart

I'm not saying that this version would escape criticism from ET (he presents an elegant re-design of bar charts on p. 128 in The Visual Display of Quantitative Information, 2nd edition). I'm sure it could be improved, and if I were going to submit it to him for a critique, or publish it in a professional journal, I would take more time to refine it. But this version meets the requirements I originally set for it, and looks respectable; it demonstrates that I care about the data and the viewer.

Summary

To summarize, these are the points I have tried to illustrate that apply to designing any informational graphic, such as a chart, table, diagram, or graph, for inclusion in a report, presentation, article, or publication:

1. Begin by defining the thinking task(s) the display is to help with - what questions does the viewer care about answering?
2. Sketch out some options; evaluate and choose the best one
3. Refine the look: Good design is unobtrusive and doesn't call attention to itself - it facilitates the thinking task.

Taking stock of 2006

Briefly, here are a few stats and highlights of 2006.

Stats:

• Software applications learned: 7
  
• Books read: 20 (13 non-fiction, 7 fiction)
• Windsurfing lessons taught: 130
• Days windsurfed: 161

Highlights:

• Getting married!
• Seeing my 79 years young windsurfing student progress
• Becoming a more proficient windsurfer: jibing consistently, sailing waves, clew-first waterstarts
• "Interactions of Color" class
• Life at "the compound"

I'll go into more depth on some of these in future posts, and look ahead to 2007.