The goal of Six Sigma is of course for your organization to reach a Six Sigma level of defects, so how do you know when you’ve got there? There are plenty of lookup tables around in textbooks for you to use (make sure you get a Six Sigma one, as most don’t go high enough for what we need). But what if you’re a bit of a perfectionist / geek like me and want to be in more control of your data? For that you need to calculate it yourself.
Sigma levels are based on the normal distribution curve. The normal curve (or bell curve) is the standard distribution for random distributions of data around a central average, with no bias above or below. Many things follow normal distributions, such as heights of people, grades in an exam, and errors in a process. You therefore use the ‘NORMSINV’ and ‘NORMSDIST’ functions in Excel to find the Sigma level or find DPMO (defects per million opportunities) from the Sigma level.
Converting DPMO to Sigma level
Sigma Level σ = ABS(NORMSINV(DPMO/1,000,000))+1.5
e.g. σ = ABS(NORMSINV(3.4/1,000,000))+1.5 = 6
Converting Sigma Level to DPMO
DPMO = NORMSDIST(1.5 – σ)*1,000,000
e.g. DPMO = NORMSDIST(1.5 – 6)*1,000,000 = 3.4