Category Archives: Statistical Breviary

What’s killing our kids?

The American Academy of Pediatrics published a new report today about injuries kids sustain from firearms. Every day during the year 2009 (the most recent data available), about 20 kids were injured by firearms seriously enough to require hospitalization. For children and adolescents under the age of twenty, there were 7,391 hospitalizations, and 453 of those kids died while in the hospital. 85 percent of the hospitalizations were for kids between the ages of 15 and 19.

Just for reference and comparison, the CDC has published some statistics about teen drivers for the year 2010. There were 282,000 teens between age 16 and 19 treated in hospital emergency rooms for injuries suffered in motor-vehicle accidents, and 2,700 deaths.

As another data point, the CDC also has some facts about unintentional drownings. For kids 14 years of age and younger, there are 765 deaths per year and 4,069 hospital admissions per year.

I realize that the data for unintentional drownings does not exactly match up with those death from motor-vehicles and firearms. The drowning data includes a broader, and different age group of kids. However, I still believe the comparison is insightful.

Causes of death in young people, 2009

Please buckle up your kids, make sure your pool has a fence, and lock up your guns.

hat tip Dr. Claire McCarthy writing at

Exploring public datasets

Uncle Sam publishes an extraordinary quantity of data, much of which is difficult for an average person to assimilate. Today I stumbled upon MetricMash, an wonderful website for exploring a handful of important public data sets.

The tools are fully interactive, which means no settings the parameters and clicking refresh on the graph. You drag the sliders, and while you drag the graph changes. It looks like they use amCharts and Tableau to power their site.

Finding the Outliers

The Dartmouth Atlas of Health Care has done a great job of providing visualizations of some complex healthcare datasets released by the Centers for Medicare & Medicaid Services. While checking it out, I noticed the following graph:

Each blue dot represents a metropolitan area, and it has a value that makes it line up with the vertical axis. If you click through to the actual site, you’ll note the city and the value appear when you hover over a blue dot. This unique method of presenting information makes it trivial to see who the outliers are, and also clearly exposes the distribution of values amongst the field. Note also the high data-ink ratio; there is very little on the page that doesn’t convey useful information.

Smart Phone Marketshare

The Nielson Company just published some interesting graphs showing US smartphone manufacturer operating system share.

The colors are a bit loud for my tastes, but this is an excellent graphic. It cleverly combines market share by both hardware manufacturer and operating system into a single, easy to understand graphic. There is no chartjunk. You can look at the graphic without reading the article and instantly understand what they are trying to convey.

Weak Tea Indeed

Boston Review published an article entitled Weak Tea by two Harvard professors, Stephen Ansolabehere and James M. Snyder, Jr, in which they assess the impact of the Tea Party movement on the 2010 elections. Regardless of whether you agree with their assessment, the graph they include in their report is horrible.

This is a great example of a duck, Edward Tufte’s moniker for a graphic whose overall design purveys graphical style rather than quantitative information. The aged paper border, the flagged elephant in the background, and the horrible choice of elephants and the letter “T” for the data points lower the data-ink ratio considerably. Most of the ink in this graphic could be removed with no loss of data-information. Weak tea indeed.

William Playfair Would Be Disappointed

In the beginning we had geographic maps and tables of numbers. And that suited us just fine for hundreds of years. Fast forward to today, and we’ve got gaudy preso’s with ugly 3D stacked bar charts. Where did we go wrong?

In 1759 a kid was born in Scotland who would change the way we consume information. By the age of 18, William Playfair was a draftsman for James Watt, the guy who didn’t invent the lightbulb but somehow got his name on every bulb sold (sorry Edison). Playfair tried silversmithing, stormed the Bastille during the French Revolution, and started a bank. In between he wrote pamphlets and developed a taste for economics.

In 1786 he published The Commercial and Political Atlas which was the first major work to contain statistical graphs; 43 time series plots and the one bar chart, the first one ever published. Fifteen years later his Statistical Breviary introduced the first pie chart. Playfair wrote:

Information, that is imperfectly acquired, is generally as imperfectly retained; and a man who has carefully investigated a printed table, finds, when done, that he has only a very faint and partial idea of what he had read; and that like a figure imprinted on sand, is soon totally erased and defaced. The amount of mercantile transactions in money, and of profit or loss, are capable of being as easily represented in drawing, as any part of space, or as the face of a country; though, till now, it has not been attempted. Upon that principle these Charts were made; and, while they give a simple and distinct idea, they are as near perfect accuracy as is any way useful. On inspecting any one of these Charts attentively, a sufficiently distinct impression will be made, to remain unimpaired for a considerable time, and the idea which does remain will be simple and complete, at once including the duration and the amount.

Statistical Breviary, p. 3-4

And from that eloquent beginning we now have the Excel Chart Wizard. William Playfair would be disappointed.