VisiCube is no longer supported:
for a more advanced version of the same product!
VisiCube's Box Array
A box plot is a univariate plot in which a statistical summary of the
distribution of a measure is presented. In such a plot, the measure is plotted along
the lone axis and the statistics are displayed in a graphical form along the axis.
(See the notes below to interpret a VisiCube box plot.)
Specifically, in VisiCube, a box plot is displayed as follows:
- The median is shown as a solid dot.
- The interquartile range is shown as a shaded box that encompasses the median.
- The upper and lower quartiles are shown by the ends of the interquartile range box.
- The upper and lower adjacent values are shown as brackets opening toward the median.
- Each measurement outside of the adjacent values is displayed individually as an open dot.
The following is an array of such plots. In a VisiCube array, you are
presented individual plots for a single measure...each for a different set of records.
An array gives you the ability to quickly compare the sets of records.
In this array, you can easily compare the height of singers across
voice parts. It can be seen that height generally increases as the voice deepens...as
expected. But some interesting exceptions can also be observed.
The median height of the high sopranos exceeds the median height of the low sopranos.
And the tallest singers in the sample are tenors, one of whom (in the group of low tenors)
stands out...literally, I'm quite sure...as significantly taller than the others.
The data presented here is from a study of the heights of singers in the
New York Choral Society by J. M. Chambers, W. S. Cleveland, B. Kleiner, and P. A. Tukey.
It appeared in Graphical Methods for Data Analyis which was
published in 1983 by Chapman and Hall (New York).
This data (and visual) is supplied with VisiCube as a sample project.