• ## Follow Blog via Email

Join 778 other subscribers

## Six Sigma and Statistics

In a previous post covering the book Six Sigma:  The Breakthrough Management Strategy Revolutionizing the World’s Top Corporations by Mikel Harry, Ph.D., and Richard Schroeder, I mentioned that Kaizen and Six Sigma are both predicated on the improvement of processes, but that Six Sigma has the statistical rigor that Kaizen does not necessarily have.   That is why Kaizen is suitable for improving quality up to about the 3-3.5 Sigma level, but once a company’s processes reach the 3.5 Sigma level, Six Sigma is the only way to reliably improve beyond that point.

The statistical rigor of Six Sigma is something I want to discuss in this post.   Sometimes statistics seems very mysterious to many Americans, but that is only because it is taught in college, and only in certain curricula.   In Japan, however, calculus and probability/statistics are taught in high school, which is why a high school graduate in Japan can handle the mathematics behind quality control, whereas a high school graduate in the U.S. needs additional training to be able to handle it.   The difference between the educational systems in Japan and the U.S. was one of the realities Japanese companies needed to face when hiring American workers for their manufacturing plants in the U.S.

As an example of how mysterious it is to American students, I can relate a time when I was in graduate school at the University of Illinois at Urbana-Champaign, and I met a fellow graduate school student at our Friday night “graduate student support group”, which we jokingly referred to our group that met at a local bar to unwind.    One Friday, he looked kind of glum, and I asked him what happened.   He said, “our statistics professor said today that up to 50% of us in the class were doing below average work–I really thought we were doing better than that.”   I told him honestly that I didn’t think the professor was maligning the students’ academic performance as much as he was giving a definition of the word “average” in a tongue-and-cheek way (one which my friend obviously failed to detect).   That considerably brightened his mood, although it firmed up my resolve to never go with my friend to Las Vegas.

But for those who work with statistics know that it is a way of teasing out of the apparent chaos of individual data measurements the coherence of explanation.   It allows one to make the invisible visible.   How do you do that?  Well, as Galileo Gallilei once said:   “Measure what is measurable, and make measurable what is not so.”    In fact, the power of statistics is so compelling that a famous science fiction writer H.G. Wells wrote in 1935, “Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write.”

In fact, the authors of the book go so far as to say that “statistical knowledge is to the information and technological age what fossil fuel was to the industrial age.”   Fossil fuel is what makes engines go, and statistical knowledge is what makes industrial processes which create those engines improve.

However, Six Sigma is not without its detractors, and one the reason why people are skeptical of Six Sigma is because they perceive it to be yet another quality initiative, another fad that gets hyped by management every time there is a new CEO at the helm.   Although this skepticism is healthy to a point, it is caused by a misunderstanding–Six Sigma is not just a quality initiative, but rather it is a business initiative which can affect the entire enterprise, and not just operations.   That topic is the subject of the next post.