Envisioning Information: Color and Information

Tufte makes a strong case that color has the power to significantly affect the reading of visual information–both positively or negatively, depending on its usage. Lack of color may not adversely affect the information display, but bad usage of color can really inhibit one’s understanding of visual information.

He also proposes the use of color to display multidimensional information, due to its own inherit multidimensionality (color, hue, and saturation). The space of color isn’t at all linear, nor is the whole of color space spanned by the primary colors (red, green, blue in this case). This means that color space doesn’t behave like Euclidean 3-space, and mathematical operations cannot be performed as they would be there. The processing of 3-dimensional data also requires linearity, so in order to process color data, we must linearize the space of displayable colors (which doesn’t really make sense to do in the first place–there is no “right” way to do this).

The Chartjunk Debate

In this article, Few essentially takes a step back from the two sides of the debate about “chartjunk.” He brings up and questions recent study that found participants tended to remember charts and graphs that were better designed and a bit more embellished than relatively “plain” charts and graphs. I strongly agree with his questioning this study (which only had 20 college participants) as any evidence that one side of the debate is more correct, as their methodology for conducting it was clearly flawed. Additionally, he brings up a valid point that neither side of the debate should be correct–that is, charts should always strive to be well designed, and to “refrain from undermining the message by significantly distracting from it or misrepresenting it.”

I like his more neutral and common-sense position that takes a second to step back and not take a rigid side in this debate in the first place.