July 2002

Vol. 4, No. 7


Six Sigma: Whose idea was it, anyway?
Speculation about origins

How does a term like “Six Sigma” migrate from a purely statistical application to mean an entire program? Whose idea was “six sigma quality,” anyway?

A recent forum on www.iSixSigma.com discussed the origins of Six Sigma, both as a program such as that utilized by Motorola and other industry giants, and as a statistical term that relates to process capability.
 
English is a dynamic, living language, and terminology often migrates from a specific application to a wider use. For example, brand names such as “Kleenex®" or “Scotch® tape” are so widely recognized that they are used generically to refer to tissues or cellophane tape even when they are not actual Kleenex or Scotch brand products. In a sense, the Six Sigma reference was applied to a specific program, where it has stuck. But of course that program is derived from an understanding of six sigma as a statistical term, and of its role in producing products and services of consistently high quality.
 
One of the contributors to the iSixSigma.com forum discussion, Michael Ervick, was knowledgeable about both past and present applications of the term. He attributed the original—or at least an early—use to W. Edwards Deming in a book published in 1925, Biometrika. Ervick also noted that it appeared in Japanese in “Theory of Process Capability and its Applications” (JUSE Press, 1975). Engineers at Boeing in the 1950s and Hewlett Packard in the 1960s, he says, “get credit for applying it to the reliability measures of individual components to measure the impact on entire systems. As a measure of process capability, it was part of TQM’s statistical process tool box.” He has promised to keep looking for the term’s “discovered” birth place in the statistical world.
 
Certainly, Six Sigma’s application as a program has its history in the archives of those industries that utilized the statistical measure to assess and improve their systems. But why six sigma, rather than, say, 3 or 12? The answer lies in an understanding of process capability itself. Process capability is the “range over which the natural variation of a process occurs as determined by the system of common causes.” It represents the ability of a system to produce a product that consistently meets specifications (James R. Evans and William M. Lindsay, The Management and Control of Quality, West Publishing, 1993, p. 436. If the natural variation of a process is larger than the specification, it will be impossible for that process to meet specifications even when it is in control. “Six sigma” quality means that while the natural variation is defined by 3 sigma, the design tolerance should be 6 sigma—or twice as large as the natural tolerance.
 
Specifications and process capability link design, manufacturing, and quality, and it is this link that is the underpinning for Six Sigma, the program. Motorola, among others, decided that 3 sigma was not good enough to assure consistently high quality. Six sigma tolerance means that 99.99966% of the yield will meet design specifications.
 
To engage in further discussion about this concept and its origins, you may want to visit http://www.iSixSigma.com/forum/showmessage.asp?messageID=14547


Copyright 2002 PQ Systems.

Please direct questions or problems regarding this web site to the Webmaster.