The statistical thinking underlying Six Sigma is based on the following three principles:

All work is subject to varying performance outcomes; the amount of variability can be limited with proper application of statistical methods; and the tighter the use of statistical controls the lower the degree of variability.

All work is a statistically controllable process; a tightly$-$controlled statistical process can reduce variability in performance outcomes; and every properly$-$controlled statistical process can achieve fewer than $5$ errors or defects per million iterations.

All work is subject to varying performance outcomes; the accuracy with which work is performed is statistically controllable; and every well$-$controlled work activity can achieve fewer than $15$ errors or defects per million iterations.

All work is a process, all processes have variability, and all processes create data that explains variability.

All work is subject to varying performance outcomes; these performance variations are chiefly due to a lack of proper focus and attention to detail on the part of workers; and insufficient worker focus and attention to detail can be controlled by proper application of statistical controls.