Several previous posts have explained techniques to make sense out of data (See my posts of April 5, 2005 and May 10, 2005 on correlation; Nov. 30, 2005 on averages, medians, and modes; Aug. 14, 2005 on multiple regression; May 31, 2005 on bell curves; Nov. 13, 2005 on power law distributions; March 10, 2005 on the difference between linear and exponential growth; and of Jan. 26, 2006 on Bayesian statistics.).
Another methodology to explain metrics, with ample examples, applies the techniques of visual depiction.
Column charts, the picket fences, such as spending year over year, make immediate sense to everyone (See my post of Nov. 19, 2005 on Google showing its legal staff growth rate over a few years.).
Bar charts, fork tines facing to the right, are another familiar display of quantitative information, such as the number of cases by type of lawsuit.
Bubble charts let you depict not only where you are spending, such as by practice area or business unit, but also the relative amount of the spending. The size of the bubble combined with the bubble’s location against an axis or two efficiently says much about the data.
Stock graphs are sometimes the right choice. With their high-low-and-closing style they let you show the high billing rate for associates, the low billing rate, and the median in one familiar symbol.
Scatter plots can make more meaningful a large amount of data such as the total number of timekeepers compared to the total amount of a matter’s bills (See my post of June 6, 2006 on scatter plots.)
Double-axis graphs let you show data that has different scales on the two vertical axes, such as the number of months cases have been pending versus dollars spent, or years out of law school against base compensation.
Trend lines. A trend line on data that otherwise might appear to be random can squeeze some sense out of it. A trend line is actually a calculation of the least-squares line, but all users need to know is that it can discern patterns, such as that more senior lawyers are paid more.