A brief critique of 31 Days in Iraq NYT Infographic

Over the past few months, the New York Times has produced a series of graphs illuminating different aspects of the Iraq conflict. While most of these graphs are straightforward presentations of the information, the most recent infographic on the past 31 days of the Iraq war represents an highly editorial standpoint, and uses poor design, and several transparent emotional ploys to create a depiction of the past month’s body count. Although I deplore the actions taken by our government in Iraq, this type of slanted presentation represents the very worst of information design, and can only foster misunderstanding of statistical presentation and distrust of efforts to coherently visualize information.

Using Ben Schneidermann’s principles for creating visualizations of information as a metric for evaluation, we find several problems immediately. The graph does not successfully provide an overview of the data on any of the aspects that were to be included in the presentation. Viewing the presentation as a gestalt, the figures that are supposedly representative of categories of casualties are so self-similar that no coherent patterns can be discerned in the data. The three different types of soldiers are so similar in shape and orientation that the silhouettes read as nearly identical. Further, the graphic for police or civilian casualties are barely different enough to render them distinctive. The designer of this graph would have done far better to use more abstracted representations of each category, to permit the viewer to view the categories of deaths as an aggregate, and to be able to determine if there patterns in the data. Colored icons would be one possibility, but as only black is available, a stronger orientation of each graphic would have been a better solution. For example:

| | | \\ | | \\ \\ /
| | - - | | / - -
\\ \\ | | - - - | /
/ / | | - - \\ \\ |

As you can see above, if the designer had incorporated a strong axis of orientation in each of the icons, it would have made groupings of the information more discernible, and given the viewer a clearer visual “handle” on the information. Further, the icon created to represent civilian deaths is unnecessarily inflammatory. Are we to conclude that the only civilian casualties are mothers of infants?

Another aspect that this graph attempts to present is the location of the deaths. The designer chose to do this with thin lines linking sets of deaths to a geographic location. However, the lines that map sets of casualties to locations where the deaths happened are obscured by other graphics, intersected with other lines so much so that they are rendered completely useless as they depict any kind of cartographic relation.

The last aspect of the data – how the casualties died, is depicted by small, regularized icons in the upper-left corner of each casualty graph. Again, these icons are so self-similar that there is no way to detect any patterns or discern any trends in the data.

One would hope that a viewer could gain a better understanding of the impact of the war on Iraqis, US and foreign troops. This graph, however, does nothing to diffuse the fog of war. The nature of this graph prevents the communication of any understanding except that war has death as a component. If the designer’s editorial intent was to bring people to the conclusion that war has a negative impact upon a range of people in a range of ways, that point could have been far better communicated with a simple, clear visual.