« Applying numbersense to a Bloomberg report | Main | A message worth repeating »


Mike Woodhouse

I like this. To address the jumpiness, you might take a weighted combination of the lines at, say, yearly intervals. So for someone aged 25 in '72, start with the 25-34 value. In '73, assume they've moved 10% up the ladder towards the 35-44 line, so they're at 0.9 * 46.5K + 0.1 * 54K, or 47.25K (assuming I'm more or less reading accurately off the chart). By 1977 they're halfway inbetween, in the 48-49K area. So a set of lines might be produced, showing the evolution for a number of starting points, maybe 5-year intervals?

Tom West

I wouldn't trace the line. The people who are 45-54 in 2011 were precisley those who were 35-44 in 2001 and 25-34 in 1991. So, for each point on the green line can be joined with the blue line ten years before and the yellow ten years after.

(If you do this, you find those aged 45-54 in 2011 are earnign *less* than those aged 35-44 in 2001).

Jeff Weir

It would be interesting to track cohort groups by plotting cohort groups taken at say 10 yearly intervals (e.g. 1970, 1980, 1990, 2000, ...) against years on the horizontal axis (e.g. year 1, year 2, year 3...) so you can see how the growth in income across time as each of the different cohorts matured compares.

Hard to describe. Will whip up a chart tomorrow, time permitting.

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