I have always wondered about the role human perception plays in a user synthesizing data visualizations. Up to what point does subjectiveness overpower objectiveness in the human brain? I think that there can never be a definitive answer with data visualization, but there is a certain level of foundational visual structure that most people would agree is most effective. I like how the author visits this concept and dives into some of the visual building blocks of perception.
One of the important takeaways for the piece was to understand all components of a visualization and how they work with one another. The visual aspects are important, but don’t work without good data and a good story, and vice versa. To create this balance, you try to have the visuals reflect the data as accurately as possible, removing any superfluous, non-data driven aesthetics. However, I found the case of the monstrous data bar chart to be very interesting because, although it was overly graphic, it helped users/readers remember the bar chart much better than a plain one. So there seem to be certain circumstances that can allow for exceptions to the visual guidelines in data visualization.
I really enjoyed reading about human perception and the different visual cues are eyes look out for, like edges, contrasts, orientations, and the ways to highlight these through specific displays. Furthermore, the part of Gestalt inferences was incredibly interesting and, in a different way, offers an understanding of human perception having to do with assumed connectivity and similarity in visualizations.
It is not just bad visuals but bad data and bad perception that can harm the effectiveness of a visualization. I would ask my peers which would be the most significant variable to get right in data visualizations? Also which Gestalt inference is the most convincing to you?