3-sigma is a statistical boundary representing plus or minus three standard deviations from a measure of central tendency (i.e. mean or median) for a group of values. Walter Shewhart, W. E. Deming’s teacher and mentor in the field of statistics, determined that when a group of measurements comes from a system of “common” causes, also referred to as a “stable system”, they will vary in predictable ways. Virtually of that variation will fall within +/- 3sigma, indicating that the events represented by the data vary in a random fashion. In other words, this variation is a product of the system as a whole. Systems of causes that do not show this predictability are unstable and therefore, unpredictable.
In stable systems, data depicting events outside of 3-sigma provide a reliable signal that the system is being influenced by ”special” causes that can be identified. These special causes can then be dealt with in order to re-establish predictability in the system of causes.
This simple idea about the nature of variation provides a foundation for the systematic design, management, and continuous improvement of work processes.
(Read a more extended discussion of 3-sigma in my sticky post, “What’s so Special about 3-Sigma?“)
plz. guide 3 sigma & 6 sigma