When I was poking around on Valentine’s Day, I came across the coolest mathematical pictures to illustrate love. It got me thinking about how visually representing the information that we want to convey is so important. Now, I totally dig numbers. And I dig artwork. I took that test this week on whether you are more left or right brained, and I scored a 53 – ambi-brained. My undergraduate study was equal measures of English Literature and Accounting, but that shouldn’t be surprising. I worked with many excellent banking and finance professionals who had degrees in English, Religious Studies, Music, Art History, and Humanities. People who analyze often appreciate the aesthetic beauty of analysis for its own sake.
Hence, it’s no surprise that pictures of numbers and data can be inherently beautiful. For example, I found this one posted by Utsav Goyal, named “The Love Function.”
Change one of those squares to a 3 or a 1, and you’ve just got a squiggle. There are also tons of beautiful patterns in the natural world – the whole science of fractals blossomed a few decades ago to show just that. The Greeks understood the connection of beauty in math and aesthetics as they were passionate about both. Aristotle, for instance, coined the concept of the Golden Mean, a ratio in nature which would reveal everything from the structure of a nautilus shell to a rose to the human ear.
One of the gurus of the idea of “beautiful evidence” is Edward Tufte, a statistician who produced several books handsomely illustrating – literally – this very point: numbers in a pattern can be visually arresting. (For those whom I have already cornered to deliver my Edward Tufte lecture, all I can say is, you knew this blog was coming). When you figure out how to arrange the numbers you have into that pattern, you can make your point far more effectively than triple that amount of commentary. We know that adage that a “picture is worth a thousand words.” I don’t want to delve into the psychology of why that’s the case; but I’d like to give a few examples — non business flavored — to show how to bring pictures to the numbers or numbers to your pictures.
If the statistics are boring, then you’ve got the wrong numbers.
Some years ago, I challenged my spouse to use some graphs in an upcoming presentation on romance writing at a conference. It was intended as a joke – “throw some pie charts in there” – I snickered. She produced a fabulous analysis of the components of “What Defines a Romance.” Without replicating the whole thing (you can plead with her for the original powerpoint, R-O-M-A-N-C-E = R-E-S-P-E-C-T), the idea was to show examples of famous stories with some romantic elements juxtaposed with true Romances, to contrast a “real” Romance with plain old stories that include people in love. Take it as an assumption that Romance is a real genre that can be defined. (If you don’t agree, you will have to take that up with the Romance Writers who have their own Guild.)
One defining element Karin included was that in a Romance, the couple needs to live happily ever after. And she used this graph to make her point. I wasn’t even there, ten years ago, and I still remember the point, because I remember the graph.
Here’s another example – a living graph. My former corporate employer was a key sponsor of an Out & Equal conference last year on LGBT issues in the workplace. One of the presentations was on the fluid nature of sexual orientation, and how that contrasts with our perceptions of the rigid nature of sexual orientation. In the workplace, this complicates how orientation is viewed. It took years for corporate cultures to embrace the concept that there are lesbian and gay people at all, while the notions of bisexuality or fluidity in gender and orientation are still not as accepted. That was the point – here’s the illustration. The presenter showed a few slides reminding everyone of Kinsey’s work from the 1950s showing that people expressed their orientation along a continuum rather than only at the opposite ends. (you can google: Kinsey Scale). She put post-it notes of the numbers 1-10 along the wall of the room and asked us to physically stand under the number where we personally fell along the continuum – creating a Living Histogram, if you will.
She then asked a series of additional questions. How do you want people to think of you, as opposed to how you think of yourself? What have you publicly told people? What have you told your family? Your manager or co-workers? Your neighbor? How would you have identified when you were 15? Etc. As questions went by, we moved between numbers, because it was natural to do so. At the same time, we were watching other people move as well. The motion and shifts of the Living Histogram illustrated the idea of the continuum again better than mere words.
After thinking about all these pretty pictures, let me end with a nice tidy, Left-Brained To Do list for you to consider when you present.Because even if you aren’t formally presenting in Busi-ness, you all like to express your opinion.
- Work with your ideas – your data – to see what natural pattern emerges. You can’t change the data, but there is probably an angle that conveys your message. Even if there is no pattern – no correlation – no natural picture – then either your angle is wrong or that in itself is a point to be made.
- The picture needs to make the point for you. If you need a lot of words to explain the picture, it’s probably not doing it.
Note that the idea could be very complex. Tufte’s most famous picture – I won’t reproduce it because you have to see it and savor it like a chocolate ganache – is of a famous historical event (Google: Minard’s infographic of Napoleon’s campaign). It doesn’t have to be simplistic.
- The discussion of the graphic can also solidify the point for you. A ten minute conversation about one graph in a presentation is far better than ten one minute presentations on separate graphs.
I would be remiss if I didn’t at least try to practice what I am preaching. So here is my graphic version to now illustrate 1,100 words. The Visualization Function, which might be something like:
f(x) = nx
where if x is pictures, and the coeffienct n representing words is oh, about 1,000.