Cheng Peng

1. Good Visualization:

This is a great graphics in the sense that it gives us a fairly intuitive grasp of how two most important factors in determining whether or not a planet is suitable for lives relate to each other. There are many praiseworthy qualities about it. First of all, the choice of color sequence is very reasonable. The color goes from red to green to blue, which is how the three colors are ordered in a color sequence, so the transition is very smooth. In addition, the color choice is great in representing temperatures. For example, red is typically associated with high temperature because of fire; green with moderate temperature because green represents lives; and blue represents low temperature because the color of sea (water) is kind of blue. Secondly, the author is very clever to use color as a means to group planets, because as human we are very aware of such changes in colors. If he/she instead used shapes, such as triangles and spheres, to represent groups it would have been a lot more difficult to see the pattern. Thirdly, using log chart is a smart choice. If the author chose to represent the data in linear scale, the square that represents potentially habitable zone would be much, much smaller as compared to the scale of the graph. It would be a lot harder to read details in that case. Finally, the degree of saturation of the circles are just right. Overly saturated colors tend to be distracting, and insufficiently saturated colors can make a graph pretty tedious. The author did a great job in controlling saturation.

2.

a. Bad Visualization #1

The biggest problem with this visualization is the amount of details displayed. Although the purpose of this graph is to show how painful tattooing could be for different parts of the body, it is absolutely necessary to show colorful tattoos all over an old man's body. When I first looked at this visualization, I thought the particular details of the tattoo contain meanings relevant to pain levels. The truth is the complicated colors and shapes on this old man's body have nothing to do with pain levels. The second problem is the choice of variable to represent level of painfulness. "Painfulness" is a continuous variable, which should be represented by something that's continuous, such as color. A major limitation of using size circles to represent pain level is that they circles are discrete. Besides, humans are not very good at comparing the areas of two circles. If I want to know whether it's more painful to tattoo on shoulder area or knee area, the graph tells me very little things. The third problem with this graph is that it's very difficult to find a pattern, since the level of painfulness is shown only at sampled area of the body.

b. Bad Visualization #2

The biggest problem with the visualization is that the differences in sizes of the "trees" create the illusion that there is depth, while actually they are just supposed to be some fancy version of the bar chart. Secondly, using "trees" or tree-like objects to represent different uses of smartphones camera is really confusing, because there is no obvious connection between the two. And thirdly, using areas of circles to represent numerical values doesn't seem to be a smart idea. This is because readers are not sure whether the radius or area is used as the scaling factor. Also, as was mentioned in the discussion above, as humans we are not very good at comparing area sizes.

Also, something that does not have to do with design: the percentages do not add up to 100%. What do those percentages mean?

3. First of all, if the author of these graphs expects the readers to get any non-trivial details from this visualization, he/she is seriously wrong. There are a few things here: 1) it's hard to compare the size of two spheres, because we're not sure if the smaller sphere is actually smaller or just deeper (farther away); 2) apparently lots of spheres are hidden behind stems or other spheres. If we have an interactive interface we can rotate the 3D graph to check out the hidden parts, but we can't really see much when it's a static graph in a paper; 3) there are both spheres and cones in the graph, but it's unclear what their differences are supposed to be. My guess is that a cone represents an "end object" that dose not have anything else to stem from it, but I'm not sure. Secondly, the use of colors is pretty confusing. For example, what do different colors represent? It's evident that the progression of colors in this graph is not continuous, so what are they supposed to mean? In addition, I'm not sure about others but certain combinations of colors and shape make me physically want to vomit. This is probably because evolution taught us to avoid extremely colorful things in nature, such as poisonous mushroom, bright-colored spiders/bugs etc..

Finally, I have questions about the directions at which those stems extend. Why does a stem go at one particular direction as opposed to any other? Are the directions of stemming supposed to mean something? It seems that if the author can require that all stemming be perpendicular to the original stems, the graph would be a lot more structured.