Tufte Talk: Chapter 6

Fair warning: this one is kind of a bummer, toward the end.

Chapter six is a last bit of conversation about Data-Ink Maximization. This chapter doesn't have principles, there's no simple tenets to take away from the topics in discussion. Here we read through Tufte's examples of how to think about erasing and drawing your data.

Which is probably why the meatiest part of the prose comes at the very end, in the Conclusion. There are four facets Tufte considers to answer the question "are the transformed designs better?"

(1) The efficiency of the graphic. Does it display more information than noise?

(2) The thought put into it, or, put simply, did you revise and edit this thing?

(3)... Let's come back to this one.

(4) Familiarity. Tufte argues that newer, more efficient designs might at first demand a cognitive load of engaging with something neoteric. But he assures us that humans will adapt. We'll get used to newer stuff, and if the newer stuff is better, we'll adopt it happily. For this Tufte leans on the third facet...

So let's talk about facet three: Trusting your audience.

This is where I'm going to have to finally give in and tangent a little. Because increasingly, as I read this and other works about data analysis, I'm bothered by the fissure between what is ideal and what is popular.

Tufte has this great quote:

...it is a frequent mistake in thinking about statistical graphics to underestimate the audience. Instead, why not assume that if you understand it, most other readers will, too? Graphics should be as intelligent and sophisticated as the accompanying text.

As someone who would love to live by the simple axiom of "Believe in People," I'm worried there is a lot more to this assumption than good will.

I was drawn to adopting a new sphere in my vocation partially because the world is changing and being exclusively a print designer is an easy way to emulate the dinosaur. But my expansion into Data as a concept is driven by a desire to work with the objective, the empirical, the Truth. I loved the notion of going from a world of subjective criticism to a world of simple answers: does the Query return the right information? Does the graph represent accurate numbers? Does the dashboard update on time?

But just as I was falling in love with this notion, my community was changing rapidly, as well, it seems. I live in America, in the Pacific Northwest. And the many different cultures and conflicting perspectives that make up this massive country are increasingly in conflict about ... anything. Just... anything.

Part of this, I believe, comes from an increase in the abilities of media and spin to manipulate a message. It's my hot take that this ability has surpassed the average American's media literacy. That we are seeing not just an uptick of persons willing to ignore data for their own comfort, but a growing population of people that very sincerely think their political or religious perspective is irrefutable fact.

It begs the question: if my work is accurate, but my audience refuses to be skeptical in favor of outright denial... What does the data-ink matter? What does the familiarity or efficiency even mean?

More and more, I'm inclined to replace facet 3 with my own concerns: Is this chart hard to argue with?

And when you go into presentation with that as your opposition it can shape your efforts quickly and drastically. We are already seeing the results of the opposite. We see things like Ted Cruz presenting global warming data by only looking at the last few decades, and not the sum total of the graph. We see Alabama altering it's values for the Corona Virus outbreak, making the chart appear, at a glance, to show zero changes in the infected population.

We see manipulation of data that upholds dogmatic or political opinions being used egregiously. And in response, we have to build our data to show things in more and more 'argument proof' formats.

I'll admit. This wasn't the world I expected. But I am here for it.

My friend and I have a weekly chat to talk about history, politics and the future of our community. And something that really hit home for us both, I think, was talking yesterday about how the biggest struggle comes from the imbalance of workload.

I can spend hours perfecting the query that draws my data, the structure of my dashboard that visualizes my data, and organizing all the links necessary to show my work and back up my insights.

And a reply in the comments can throw out a straw man, a false equivalence, and just walk away. The effort it takes me to represent the Truth is measured in hours. The time it takes to poo-poo that effort is, apparently, moments. "The Audience" that Tufte is advocating for, can sometimes put more thought into their insults than they do into my data viz.

I'm not saying we shouldn't believe in our audience. I'm not saying we shouldn't believe in people. I'm saying: this is the world we get, the world we're in here and now, and to ignore that facet of it - to ignore that facet of people - would be an error.

Doubt and Distrust are powerful weapons. Equal, I fear, to the power of democratized data visualization and analytics.