Readings15: Uncertainty and Presentations

in All Readings

There are two topics this week: Uncertainty Visualization and Presentations. But since it’s the end of the semester, I won’t give you too much to read. (even though they are both really valuable topics!).

  1. (required)  Read this page (yes, it is content)
  2. (required)  Read one of the articles on Uncertainty
  3. (required)  Read my posting on presentations
  4. (required)  Watch a Hans Rosling video
  5. (optional)  Watch the Vis 2022 Capstone
  6. (optional)  Read more on Uncertainty Visualization

Presentations

Helping you think about presentations is something I like to do in this class (and all grad classes).

Being able to give presentations is an important skill: if you’re a researcher; even if you are a practitioner. But rarely do we discuss how to do it. We expect students to figure it out. I realized that the skill of giving a presentation is a critical thing - and that grad classes should try to teach it (rather than just require it).

Unfortunately, in the large grad class, we cannot let everyone practice giving presentations. Practice is an important part of learning to do it. Instead, you’ll get to hear my talk about, and read some of my thoughts. And watch a video or two.

My annual rant about presentations is an unnusual event. When I first did it, it happened by accident. But so many students have responded positively to it, that I keep doing it. Each year it ends up a little different. Over time, it has stopped being a random ramble (The first time I did it was spontaneous. I just started talking - and by the end of the 75 minutes there was a list on the blackboard, that became the outline for future things). But it’s usually highly interactive.

The “readings” are:

  1. This page (especially the section “My Notes on Presentations” which sets up #2)
  2. My “notes” on presentations (see the caveat below)
  3. Watching a recording of a Hans Rosling talk
  4. (optional)  Watch the capstone from Vis 2022

My Notes on Presentations

Before reading my notes, here are some caveats (note: this is taken from the 2012 class):

  • The goals and standard for presentation really vary across venue/discipline. What we value in computer science (in particular the areas I work in) are quite different than in other disciplines. It’s hard for me to discuss this without value judgement (since I am bred to believe in the “CS way”), but I also plead ignorance to the practices in other area. I’d like to use this as a chance to learn about others.
  • I don’t consider myself to be a great presenter. Do as I say, not as I do. The upside of this, is that it means I think about how to be better at it.
  • A lecture is not the same as a talk, so what you see in class is quite different than what you would see in one of my talks.
  • Even within a particular style/venue/type of talk, there is a wide range of opinions on what is good talk, what the goals should be, …
  • The “right answer” depends not only on the situation, but on the person. But that will be one of the biggest lessons I hope you get. I may not speak to your specific case, but hopefully, you can see how the general lessons apply.
  • As you might guess, I have strong opinions. But you don’t have to guess at what they are, since I’ve written them down.

Given that…

My real goal is to get you to think about what might make for a good presentation, and to form your own strong opinions – even if they are different than mine.

Given that, read my posting about presentations. Yes, it’s from a 2011 class – but I think if I were updating it, it wouldn’t be much different.

Video Presentations

Hans Rosling was a famous presenter – talking about social issues around the world in venues like TED, etc. He was famous for presenting data in a compelling way to make his points for a broad audience. Sadly, he died this year. But his influence is significant (both on presentating data and on the world in general).

If you haven’t seen a Rosling talk, you need to experience one. If you have seen one, you probably won’t mind watching another.

There are lots of videos of rosling presentations – here’s one I have handy, or here’s another one.

This page has an in-browser version of a famous Hans Rosling visualization (Gapminder), and a link to the video of him presenting it. It emphasizes (to me) that the visualization by itself isn’t what is so great: it’s his use of it in the presentation.

The actual point of Rosling is not his visualizations (he does use standard visualization effectively – often with animation), but rather as a way to talk about presentations.

You might also be interested in his son Ola Rosling’s keynote from a recent EuroVis. Factfulness on YouTube. It might be the “at home on Zoom” version of a Hans Rosling talk.

Optional - A very different style

And, for something completely different… This is the capstone talk from Vis 2022. I wanted to make it required watching since the content is so compelling, but I didn’t know how to squeze it in. And since this week you are doing your project you might either (1) need a break so watching a video is good, or (2) are too busy to watch a 30 minute long video. So I am making it optional.

The content is fascinating, but it’s also a great talk - in a different style than most computer scientists give talks. The content is good food for thought, but so is the presentation style.

Note: the actual talk is only about 30 minutes long. There is an introduction, and a lot of questions and other stuff at the end.

Uncertainty

Uncertainty Visualization is a critical topic in class. But, we’ve run out of time, so we won’t be able to do it justice. (even though in my notes from last time I put “prioritize uncertainty visualization”).

For a required reading, you need to read either #1 or #2. #2 is better, but #1 is shorter. There is substantial overlap.

  1. Jessica Hullman How to get better at Embracing Unknowns, Scientific American, Volume 29, Special Issue. (pdf) (official paywalled).

    A nice brief survey.

    If you want more, you could watch this talk at OpenVis. Although, these slides would have been even better.

  2. Padilla, L., Kay, M. and Hullman, J. (2022). Uncertainty Visualization (Chapter 22). In W. Piegorsch, R. Levine, H. Zhang, and T. Lee (Eds.), Computational Statistics in Data Science (pp. 405-421). Wiley. (pdf) (PsyArXov)

    A litte more technical than #1, covers some example techniques.