Progress in 2023, Jessica Edition

Since Aki and Andrew are doing it… 

Published:

Unpublished/Preprints:

Performed:

If I had to choose a favorite (beyond the play, of course) it would be the rational agent benchmark paper, discussed here. But I also really like the causal quartets paper. The first aims to increase what we learn from experiments in empirical visualization and HCI through comparison to decision-theoretic benchmarks. The second aims to get people to think twice about what they’ve learned from an average treatment effect. Both have influenced what I’ve worked on since.

7 thoughts on “Progress in 2023, Jessica Edition

  1. Nice! I also really liked the causal quartet paper.
    I don’t work in visualization research (at least formally, but I do make lots of plots for human consumption!), but reading the rational agent benchmark paper really makes me realize how difficult it is to research these topics! Generally, I make a plot, and then I either 1) ask several people to tell me what they learned, or 2) send it to them and wait until the discussion meeting to see what questions they have (pretty revealing sometimes, especially if I did a poor job translating info to consumable visual). It’s a lot easier to make plots for my own consumption vs someone else’s…

    • I think in many applied settings your best bet is to try to understand how the people you designed the vis for actually use it – you can get a lot of valuable information if your users are invested in helping you improve it for their purposes. But visualization interpretation has also become a research topic in its own right, and so there are more and more people running controlled experiments to try to establish what aspects of a visualization design or scenario lead to what kind of performance. The belief is that this work is more rigorous and allows for the development of generalizable theories about what makes an effective visualization. But it is often conducted without there being a well-defined standard for “good” performance (whether inference or decision-making) to compare people’s performance to. So the interpretations of the results can be misleading. It’s not clear to me looking at it that we’re getting closer to understanding what makes a visualization effective for reasoning under uncertainty. And all this is also playing out (with even more hype) in the growing base of research on how to design for AI-assisted decision making. My next benchmark paper will tackle that!

  2. The Reforms paper was nice to see as someone interested in reporting guidelines. Though by now, it seems like there are so many forming that a wiki would be a good idea to really scale up the practice. It would also make it easier to collect suggestions from diverse readers and evolve guidelines more readily.

    I also know a toxicologist who’s infusing reporting guidelines in the paper authoring stage to ensure that scholars actually use these guidelines.

Leave a Reply

Your email address will not be published. Required fields are marked *