« A deeper look at the Bloomberg report | Main | Update on Dataviz Workshop 1 »


Tom West

The correct thing to do here is plot Dow 1928/29 on the x-axis, and Dow 2012/14 on the y-axis. That would reveal any correlation much better.

Shawn Shafer

And here I thought the correlation showed that the market grows over time.


Seems like your replot of the data is good, and you could simplify by giving it truly a single y-axis... say proportional change relative to Jan 1 or something like that. It would look identical to yours, but emphasize that the volatility observed in '28 & '29 isn't being seen today.

I get what you're saying about dual axes on the first viral chart. However I think the more dramatic sin is one of extreme cherry-picking. If you're looking for a time-period to compare '28 & '29 to, you could pick any period, say the 2 years after July 2nd 1983. Given enough patience or an algorithm that looked for meaningless correlation, you could probably do even better than this particular comparison.


Hig: The second version is not my work but that of the Atlantic blogger. Just thought I should clarify that.

The comments to this entry are closed.

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