Teaching and doing digital history.

Data Visualization & The Historian

I was looking forward to revisiting Edward Tufte this week. In 2000, I scammed a free trip to one of Tufte’s day-long seminars in Crystal City. At the time, I was working for Sallie Mae as one of their web content writers and often, my work required drilling down complicated loan details into digestible webby bits. One of the ways we were hoping to make that easier was by (smartly) using visuals.

While sending a lowly web writer to a seminar discussing the 30,000 foot view of visualization was probably not the most utilitarian of ideas, I did appreciate the fact that I was getting a nice overview of the importance of proper visualizations and the theory behind what is good and what is bad. Not that I think I ever fully realized my goals of making kick-ass visualizations.

Years, age and more education later and I still know the difference between a good visual and a bad one and yet, I’m still not sure I could easily create one. I still architect visualizations for work and I think that a visual is only as good as the data you have underneath. Futhermore, visualizations are only as effective as person creating the visual has an understanding of the underlying data. And I think this is the rub. And the irony. Time and detailed study are required to create a visualization that people can quickly find and skim for the pertinent information. Gobs of info in a quick byte.

I’m still striving to deal with data in intelligent ways. When I come across something I really like, like the video in my very first post here, I spend some time and really look at what they’ve done and how. The Onion does them well, but I think it’s easier to build them when you aren’t as stressed about the integrity of the data, which I ended up having to do quite a bit during my five-year stint running an entertainment channel at AOL.

Complicated or nuanced data is a bit tougher to deal with. And as much as Tufte hates it, USAToday has created a level of expectation for these things. And if we are attempting to attract and audience larger than 20, it would do us well to consider that.

And again, we are back at the question that spans two Clio courses: who are building these suckers for?

I believe we can strive to cut a middle line where we remain academic and true to our standing as historians, and still make the data easy to digest and recall at a later time. After all, I think Dr. Hans Rosling toes this line pretty well in his video. I have no doubt about the veracity of the underlying data, nor do I think it’s shallow by any means.

So upon my re-read, many years later, of Visual Explanations, I think it still stands the test of time. We have more accessible data visualization tools now and a bit more experience in viewing these on the web than we did when it was first published. We can still strive to find the right balance between information and accessibility. As a budding historian, it’s certainly a goal that I have.

Though the fake ones are amusing!

funny graphs and charts
see more Funny Graphs


  1. The History Channel-o, sweet nemesis. Now known as just “HISTORY.” That didn’t make it at all awkward when the Library of Congress entered into a cooperative agreement with them. The first thing they came to me with was their desire for original WWII color footage, for their fantabulous production “WWII in Hi-Def.” Just think about that one for a moment. Uh huh. Hitler is, sadly, the fallback for all gaps in their programming. Crappy program after crappy program with third and fourth rate talking heads. Edward Tufte is someone who vision and love for the craft, and thus disdain for the hack. But the hack will always be cheap and easy, and perfectly suited for mass consumption and the internet. So we do face the problem of creating a beautiful, authoritative website that a handful of people will utilize, and that we can be proud of, or playing by another set of rules.

  2. Roger Connor says:

    I think that building web content is, by definition, building outreach, and as such, we are trying to improve accessibility. As Alexa notes, the noise of the “hacks” will always be present, but it has been since almost immediately after the invention of the printing press. Good content that is also nicely produced will stand out. Most non-historians that I talk to about the quality of what’s on “History” are already attuned to the nature of their sausage making (and as a sometimes third or fourth rate talking head – this is something I have experienced much too closely).

  3. I was reading you post just now and as I was reading, I realized how lucky you are to have experience with web design. Despite all we learn and all we read, I am not sure that I will ever make it to “aesthetically pleasing” for a large audience.
    Your comments about maintaining a balance between academically correct data and a presentation that is easily accessible mirror my concerns about this kind of work.

  4. “A visual is only as good as the data you have underneath” — this is very true, and I think the same can be said about websites. A website is only as good as the content it has on it, and its ability to convey that material to the audience.

    Just look at Tufte’s example of the Challenger — an intriguing visual cannot mask errors that are contained within it, and can lead to devastating consequences.

    That’s really interesting to hear your background story; as you mentioned, it looks like the book has stood the test of time after all these years.

    And very funny History Channel diagram you posted there — except you forgot UFOs and Nostradamus!!!

  5. first comment — LOVED the History Channel diagram. Back when A and E used to actually be about arts and entertainment, I called it arms and empire. Whenever they ran out of programming, it always boiled down to one country in Europe trying to blow the other one to bits.

    There’s an interesting discussion here between Tufte and us about content vs. form. Both are important, but as many of you all have commented, poor data shouldn’t be masked by good design. There seems to be a clamor among us all for content driving design and not the reverse. Which I think is absolutely on target.

    But I think there’s something deeper here. IF you have data AND its good and presumably important, BUT its not portrayed correctly, you’re doing yourself (and in the case of NASA, a group of astronauts and their families) as great, if not greater a disservice. Unintended consequences of poor design (or writing for that matter) is just as problematic as design without data. The latter is fatuous; the former could be dangerous.


  1. […] me grumpy. Thankfully the Clio II blogs have more to offer. This week, I commented on Erin’s blog, and Annalise’s. And rather than clog up the comments box, I thought I’d add a shorter […]

Speak Your Mind