A1-RyanGreenberg

From CS 294-10 Visualization Sp10

Jump to: navigation, search

Good example

Image:kidnappings_cellphones.jpg

I spent some time trying to decide whether I thought this was a good or bad visualization. It appeared in the Economist alongside an article about research theorizing that mobile phone penetration in Colombia has decreased the number of kidnappings:

"For a given rise in police numbers, kidnappings fell in line with the expansion of mobile overage in each area. Mobiles enable kidnap victims and witnesses to inform the police swiftly. Since speed is of the essence in foiling kidnaps, this made police more effective and kidnapping riskier for its perpetrators."

First, the sparse style, typical of the Economist, minimizes the amount of information to digest; there is less chartjunk than you find on a typical graph. In terms of content, the graph encourages skepticism about the findings. You can see that the fall in kidnappings lead the decrease in phone antennae, and that the rate slowed while antenna installation continued apace.

Source: The Economist, December 19, p60.

Bad example

Image:statviz_logs.jpg

This chart, produced by the StatViz (http://statviz.sourceforge.net/) package, is featured in the book Building an Intelligent Web. It is a bad visualization because the first sensation the viewer gets is one of confusion and chaos. The point of this visualization is to make it simpler to see which pages on a website a user requested. Based on the visual complexity it's not clear that this chart is any better than a list of user actions. Each clicked link by the user is shown with a step number and a full hour:minute:section timestamp, adding clutter to the diagram, and forcing the user to scan the entire diagram to find the next step. Since there isn't any legend, the user has to guess what the dotted, red, and black lines mean.

Source: Building an Intelligent Web, Akerkar and Lingras, p236.



[add comment]
Personal tools