Two main takeaways:
- The talk established an excellent mental representation of the author-user-principle of visualization.
- It emphasized that, every visualization should have a "point", i.e., to answer the question: what can the users get out of it? - Every visualization should tell a story.
Considering the place where the "point"(takeaway of the visualization) can be put between author and user, the visualizations can be categorized into a continuous spectrum.
- If the "point" is on the user's side, that means the visualization is author-driven. The visualization is embeded in the author's storyline. A good example would be Hans Rosling's TED talk: https://www.ted.com/talks/hans_rosling_shows_the_best_stats_you_ve_ever_seen
- If the "point" is on the author's side, that means the visualization is user-driven. Basically, the message to be conveyed is totally driven by the user interaction.
- Other visualizations stay in the middle, and they are an integration of the control from the author and the interaction from the user.
To extend this a bit from my point of view. (I mainly tried to put something that's known by all into the continuous spectrum.)
I think the "points" that the author-driven visualizations convey is usually smaller, more focused and well-controlled.
On the other hand, the "points" from user-driven visualizations are usually bigger, more flexible and out-of-control.
When we think about something like user engagement, the focused point from the former tend to make the user easy to be engaged, but the lack of interaction may prevent user from this goal.
In terms of the latter, we all know that the interaction can help people better memorize and get involved, while they really tend to get lost when trying to find the less-focused "points".
Although a popular solution to improve user engagement is to let the "point" stay somewhere in the middle of the continuous spectrum, is it possible to tweak the user-driven visualizations so that the user can better locate the "point" while messing around with the visualization?