A1-NickDoty
From CS294-10 Visualization Fa08
Bad Visualization (or is it?)
Berkeley Campus Map, available everywhere on campus
I thought this was a bad visualization at first, but it's good for certain purposes. The only color on the map stands out and draws your attention to it, making me think that these pathways were what I should focus on. But in fact they cover only a particular case: wheelchair access. For that purpose though, the map is excellent and by making that layer the only source of color, they can be ignored by the rest of users.
Good Visualization (or is it?)
Diorama inside Doe Library
At first I thought this was a fantastic visualization since it gave me such an intuitive view of the area. It also affectively shows the outline of the underground extension. On the other hand, nothing is clearly labeled, so it didn't help me find my way or get situated and it takes up a lot more space than the simple paper map.
Deconstructing and Improving the Campus Map
I'll focus on the first visualization above, since it was the one I initially had the most problems with. Here's my brief reconstruction of a blown-up portion of that image.
To break down the variables here, the map is fairly simple. Position in both x and y dimensions shows geographic position of buildings and pathways. Two colors appear on the map, both used nominally. The large green dot names the position where the user is standing right now (the map is at a fixed kiosk), and the bright blue lines name the wheelchair accessible paths.
My problems with the map were:
- that the blue lines drew attention to themselves too much, so that I only focused on those lines even though they weren't relevant to me
- that the buildings on the map only showed outline of position, which didn't help me actually identify the buildings
Let's look at the first problem first. Just de-emphasizing the color makes a huge difference.
Making a thinner line or a dashed line can also help.
Just getting rid of the color though would be a huge mistake. The wheelchair accessible lines are then visually indistinguishable from the outlines of each path.
But we might be able to accomplish the same thing without the expensive (visually and financially) colored ink if the line is a lighter colored and dashed (changing value and texture).
The second problem I had was that the outlines of the buildings didn't help me identify each building as I was walking past it. There do exist campus maps on the Berkeley website that show more visual detail, but they don't include the wheelchair accessible paths as in the map above. Here's my reconstruction of the "3d" map with the original coloring of the flat campus map.
I think the additional detailed imagery helps a lot in identifying buildings. But it's not without its drawbacks. As you can see, it's not really possible to show a continuous line for wheelchair accessible paths without crossing over the Campanile tower.
If we use the lessons we learned above, we can lessen the intensity of this line while still keeping it on a separate layer of the visualization.
But it's worth noting that the gray versions from above don't work as well here since the pathways are drawn on this particular map as whitespace between gray (grassy) areas. Adding more gray half-tone just confuses the image (in Tufte's parlance, it's not on a separate layer and so interacts with the other information).
