Hazards of Visualization: Misrepresentation of American Concern over Terrorism

Ross’s post about Vis Politics included commentary on the Political Climate site that made me realize the importance of having the message a visualization portrays be the message that was intended. I came to this realization because like Ross, when I first visited the Political Climate site I was amazed at the huge jump between 2001 and 2002 in how important terrorism was to Americans. It made sense that the events of September 11, 2001 would make Americans much more concerned about terrorism but I was still surprised at the jump from almost no concern to 83% of Americans who viewed terrorism as a top concern. Upon further inspection, I realized that the reason the visualization shows such a large difference is that there is no data available on terrorism from 2001.

We both interpreted the data to mean that Americans had an 83% jump in concern over terrorism when this was not actually the message the data was trying to get across. I would like to caution visualization sites against making their no data available visualizations look like their actual visualization. Political Climate could easily have grayed out terrorism in the 2001 plot to avoid any confusion or misinterpretation. The other side of this is that as consumers of information and visualizations in particular, we need to be careful about the conclusions we make. If an 83% jump seems unlikely, it is worth researching where the jump is coming from (or, in this case, not coming from).