Color & Information by Edward Tufte

Tufte argues that it’s quite easy to add color into visualizations, but certain mistakes (color violations) can lead to incorrect understandings or and even strong visual aversions. He claims that color decisions have rules, guidelines, and limitations, meaning that it’s crucial for certain aesthetics to correctly use color to properly translate ideas and values. You can use color to express differences, distances, measurements, and many other aspects of information.

Tufte also goes on to say that color can, in fact, communicate multidimensionality in displaying complex information. This reminds me of our discussion over the time mapping visualizations, where we attempted to dissect what certain color saturation overlays meant - we felt that the use of color saturation in that example was to indicate information that was multidimensional and in-depth, but it was hard to decipher exactly what the value was.

This chapter was interesting because it highlighted how color is a powerful tool, but it’s extremely easy to make mistakes and generate misunderstandings when one uses color incorrectly. I feel like students are always learning about certain rules they can or cannot break in the realm of their studies, but color interpretations and the definition of good aesthetics are always involving, so I wonder how strictly enforced these rules can really be.

**The Chartjunk Debate by Stephen Few

Few’s main argument is that Edward Tufte’s encompassing term of chartjunk, used to describe extra and potentially necessary embellishments to data visualizations, is too loosely defined and too critical of visual elements in charts that meaningfully support a chart’s message.

I generally agree with Few’s more moderate take on chartjunk. I’ve come across many unappealing graph visualizations and even disastrous graph visualizations that could’ve meaningfully depicted data if they weren’t designed so poorly, but I still believe that when done correctly, visual embellishments help relay a chart’s point.

When people see graphs and other kinds of data visualizations, they assume that the data is totally unbiased and is purely factual. We’ve talked about this inaccurate portrayal many times in class - data visualizations always have a goal, relay a biased opinion, and take a stance on data. Adding embellishments can better highlight that someone has a specific goal in preseting data information. As Few wrote, right now, “no one graph can display the full story that lives in a set of data, but it should provide the richest view possible for understanding what matters,” but it’s difficult to create a standard classification that teaches exactly what kind of embellishments are bad/unhelpful.