Span

Hi Anonymous,

Here is a simple and practical example of Span. Let's say that you're installing cable (for television, high speed internet access, etc.). We all know the standard clause: be home between this time and that time -- usually a 3 hour window. At the end of the 3 hour block, you are one upset customer; you waited at home and they didn't come. So, if they come anytime during the 3 hour time period, they're considered on-time (not ideal, but go with me). If they're after the 3 hour time period, you can quantify how late they are. Let's just use half-hour increments as the measurement time period (I know, I know, the granularity probably isn't right).

Let's look at the following 11 installations in half-hours late:
0
0
0
3
2
0.5
0
0
5
3
0.5

The P-5% and P-95% are easily calculated in Microsoft Excel, Minitab, JMP, etc. (In Excel, use the function "=PERCENTILE(A1:A11,0.05" where the A1:A11 is the 11 data points above and the 0.05 is the 5th Percentile, although you can also substitute 0.95 for the 95th percentile.)

Here are the results:
P5% = 0
P95% = 4.24.0

Span = P95 - P5 = 4.24.0 hours. Your job, should you choose to accept it, is to find root causes and poka yoke them to reduce the span (sorry, it's late).

I hope that example helps. Good luck with Span!

Kerri