« Can good charts be entertaining? | Main | The redundant dimension eye-trick »


Andrew Gelman

Take your graph of female/male life expectancy ratio and rotate it 90 degrees. That way you can spell out the country names and it's much more readable.

Actually, I'd also change the scale to difference in years rather than life expectancy ratio; the numbers are all close enough that there's no real need to divide, and years are clearer than ratios.


Population pyramids are used to show sex-specific age distributions, not to show country-specific life expectancies.

The bottom chart appear to be ordered by male life expectancy. If the point is that female life expectancy is higher, you might want to re-order by the gap. Alternatively, you could do a Tukey mean-difference plot. Here's one; here's another.


This nice posting reminds me of a slightly different type of comparisons in the statgraphics blog.


Great points Andrew. I like them both.

Robert, the Tukey mean-difference plot is conceptually similar to my plot except that it introduces a ranking on the x-axis. This does bring up another improvement to the plot which is that it'd be better to have sorted the countries in terms of increasing differences.

Finally, Aleks pointer to statgraphics currently has a dizzingly bad map from Forbes magazine. Fun to look at though

The comments to this entry are closed.


Link to Principal Analytics Prep

See our curriculum, instructors. Apply.
Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR.

See my Youtube and Flickr.

Book Blog

Link to junkcharts

Graphics design by Amanda Lee

The Read

Keep in Touch

follow me on Twitter