Ring out the praise for bell curves!

A bell curve distribution describes the normal range of many, many metrics. Try this. Array the amounts of all the invoices your department received during the past three years, and see the familiar hump with a few at the extremes (very small amounts and very large amounts) and many at the high point in the middle (perhaps $5,000?). Standard deviations describe how many invoices are within pre-defined distances on either side of the hump.

Approximately 80 percent of the numbers in a normal distribution lie within a standard deviation on either side of the hump.

For other analytic tools see my posts of Sept. 4, 2005 on Pareto’s 80/20 rule, of April 5, 2005 on correlations of innovation in law departments to their number of lawyers, May 10, 2005 on the correlation between firm size and billing rates, July 25, 2005 on power law distributions, and March 10, 2005 on the difference between linear and exponential growth.