Graphic Representation

Earlier today a friend posted a like on Facebook of an infographic from a site that loves science almost as much as it loves adolescent vulgarity. It came originally from this article at Vox. None of the following is a reflection on my friend, because she is much smarter than me (her choice of friends notwithstanding) and loves science, the elimination of disease and approppriate vocabulary.

The infographic and article were originally published last summer about the time of the Ice Bucket Challenge meme, in which people challenged each other to have a bucket of ice dumped on them in connection with raising money for research into ALS, commonly known as Lou Gehrig's disease. The Vox writer had the laudable goal of pointing out that there are several things more likely to kill more people each year that could use some funding for research as well. She didn't address several of the other problems surrounding social media challenge activism, but that wasn't her purpose.

Critiques of the original infographic, which used circles of different sizes to show which diseases kill more people and which receive the most donations, caused its revision. The version you can currently find at both sites is supposed to have circles re-scaled to have their area represent the figures it's comparing, rather than their diameter. The latter exaggerates the real differences.

A post on this blog notes that problem, as well as several others, with this kind of chart. And as its update points out, the "money raised" category doesn't show the total amount raised to fight the particular disease listed. It shows the amount raised by a particular charity's event. In other words, the $54.1 million figure shown in the blue circle isn't the total raised in the U.S. to fight our number one killer, heart disease -- it's the amount raised by the "Jump Rope for Heart" event. If you donated to some other event that targets research into stopping heart disease then your giving wasn't represented in that $54.1 million. Are the events the same? Do they happen in the same time frame? If all we have is the infographic, we don't know.

You might think you would still be on pretty solid ground guessing that the total amount donated to fight heart disease isn't as much as the total amount given for either of the two causes that top it on the graphic. But science is about testing guesses to see which of them are more accurate. And as this comment on the post at the Statistical Modeling blog points out, figures for 2013 (the graphic dates from 2014, so those are the latest figures at the time) show the total funds raised by the American Heart Association exceed those raised by the Komen Foundation.

Which means that the graphic is accurate but needs a title that better describes what it shows: That the fundraising totals of selected events conducted by certain charities to raise money for their causes don't match up with the rankings number of people killed by those diseases. But not only is that a much more cumbersome title, it's not particularly useful information.

That, though, is par for the course for a site who's "crucial facts to understand the Israel-Gaza crisis" list in 2014 originally included a reference to a bridge between Gaza and the West Bank that would have been one of ten longest road bridges in the world. If it had ever existed -- and that fact, I believe, can be tested by experimentation.