According to Tufte's principles of data analysis and presentation, visualizations like the one Gapminder did for HIV infection trends fall slightly short of the mark of good design, in my opinion. For one, I'm just not exactly sure what Gapminder is trying to tell me, the viewer. Are they supposing I'll be astonished by the data, as if I've never heard that there's an HIV/AIDS crisis in the world? Also, on their website they seem slightly self-fulfilling, as though they believe their visualization is dispelling the myth that HIV isn't just an "African" disease. Who thought this anyway? I'm sure people are aware that North Africa is largely Muslim and particularly sensitive to the issue of sex outside of marriage, which I believe is largely the cause of its low infectious rate. And on a related, but much more serious note, the visualization largely ignores the fact that actually there are pockets in West Africa, concentrated in Guinea and Senegal, in which the spread of infection is increasing, based on data from UNAIDS. Gapminder writes, "The very high percentage of infected people is definitely not a general African phenomenon," and that the greatest rates of infection are in the south and east. While true, the effect of the visualization builds off a false sense of comparison, Tufte's first principle, and places focus not on those places that truly need aid and, more importantly, prevention.

The visualization also does a poor job creating a sense of what kind of dynamics exist in the region of Africa. Sure, the rate of infection in the south is dramatically increasing, in some places more than others (South Africa and Mozambique, particularly), but the visualization completely glazes over the fact that in neighboring Zimbabwe the number of infected has gone down by almost 33% in a six year period. That is astonishing, mind-blowing, and completely absent in the visualization. That kind of figure throws everything into relief and begs a serious question: how would a country so surrounded by ailing nations find a way to overcome the same difficulties? Of course, the answer is that the nation's efforts to quell the spread of infection has resulted in mass deaths that subsequently impacted the number of people living with AIDS, but this dramatic fall in numbers should be represented in some form, perhaps to point the finger at their poor support for infected individuals.  Refusing to address this fact and ignore Tufte's second principle, Gapminder's HIV visualization lets an opportunity to stagger the viewer go quietly and sadly by.

And finally, because of a lack of multivariate analysis, Gapminder's visualization uncovers no new questions about the spread of infection, other than to give the same trite explanation that, yes, the Third World needs our help. I feel disappointment, considering that they are getting so much press and are shown as the next wave in visualizations, and that given the opportunity to create an engaging collection of data they place it on something so banal and lifeless as what appears to be an Excel spreadsheet with some circles on it.  If you are going to instill some kind of importance on the message you are making with your visualization, think about the subject matter, think about what in particular this data is trying to tell us. My main problem with Gapminder is that it's a tool for anyone to make a visualization, but what about customization -- what about putting HIV infection and the CIA World Fact Book on the same page without the same drab color scheme as what the Gapminder application provides? What about creating something more enthralling than circles? Indonesia encountered an almost 200% increase in HIV infections and all we have to show is a small turquoise circle?

While Gapminder stays true to the dataset, it doesn't take it to another level in which one can be both moved to learn and compelled to move, largely because of its inability to provide context, causality and a wide array of important factors.