Many sources date the classic "box plot" design to John Tukey and his "schematic plot" of 1970. The design seems to have stayed relatively static since then, with Edward Tufte's cut-down version of the box plot failing to catch on, while violin plots - though a more informative variant of the box plot - remain less popular. Cleveland's suggestion that whiskers extend out to the 10th and 90th percentiles has some supporters, see Cox (2009), but is not the norm.

Hadley Wickham and Lisa Stryjewski wrote an unpublished paper on the history of box plots but it does not seem to cover the historical precursors of the box plots.

So how did the current, ubiquitous "box and whiskers" plot come about? What kind of data visualisation did it evolve from, did those earlier designs have any significant advantages, and why do they seem to have been eclipsed so comprehensively in usage by Tukey's scheme? An illustrated answer would be a bonus, but to be directed to a reference that dives deeper historically than Wickham and Stryjewski would be useful.

References

• Cox, N. J. (2009). Speaking Stata: Creating and varying box plots. Stata Journal, 9(3), 478.
• Wickham, H. and Stryjewski, L. (2011). 40 years of boxplots. http://vita.had.co.nz/papers/boxplots.pdf