informs ny Meeting Notes
Have news to share? Email your news announcement or press release to the webmaster.

Notes from December 10, 2003
Surfing 101 - Riding the Six Sigma Wave
Dr. David Burn, Chief Statistician (and Master Black Belt) at Boston Scientific. 

The talk presented insights into Six Sigma (6s) process/movement that focuses on quality improvement (total quality management, e.g. 'zero-defects') while attempting a universal framework across industries and technological processes. 

Several names are recognized in the area of (pre Six Sigma) quality improvement; beginning with Deming of NYU and Fordham, whose statistical quality control went appreciated in post WWII Japan,
Ishikawa - seven tools ('fish-bone/cause-effect diagram', flowchart, Pareto chart, histogram, control chart, scatter diagram, notepad), or
Juran - with the breakthrough improvement approach…

The term Six Sigma (6s) was coined in 1985 by Bill Smith of Motorola.  Six Sigma - is defined as 'a vision of quality which equates with only 3.4 defects per million (pm) opportunities (dpmo) for each product or service transaction.'  The idea couth on, and similar programs started in many other companies, (GE Work Out, Kodak, Allied Signals and others) which jointly claim billions of dollars in savings ($500m/yr) from 'following the Six Sigma Way'.

Perhaps because of great amount of material, perhaps (also) out of the attempt to serve such a wide diversity of production processes, the 6s presentations often stop with (useful) generalia before/instead of getting into 'juicy technicalities', (e.g.  'three essential elements: Delighting Customers, Outside-In Thinking, Leadership Commitment', or 'Six Sigma is embedding quality thinking - process thinking - across every level and in every operation of our Company around the globe.' 'Our Customers Feel the Variance, Not the Mean').  The 6s process also introduces a number of colorful acronyms (e.g. DFSS - (Design for Six Sigma), DMAIC - (Define, Measure, Analyze, Improve and Control), DMADV (Define, Measure, Analyze, Design, Verify.) …) and snappy role names (Green Belt, Black Belt, Master Black Belt).

As opposed to Project Management (PM), there is no single organization defining the 6s body of knowledge.  Instead, dozens of institutions train in 6s and issue certifications.  (PMI (Project Management Institute) maintains the body of knowledge and manages exams and certifications).  With so many vendors there has to be some variability in the material (and I doubt the differences are at the 6s levels), e.g. some include a Yellow Belt role, others do not.

Even the flagship 6 sigmas have decentralized interpretations.  It may be attributed different values (I recall values in range 3.2-4.5 per million).  In the presentation (a view shared also by Mulbury of, 6s translates to 0.0002 pm (give or take a decimal place) and is modified by accounting for a 1.5s long-term shift/drift/bias (which makes 6s more of a 4.5s?).  According to another source, in a perfectly centered distribution, 6s would yield 2dpmo.

On the positive side of decentralization, since applicable quantitative approaches vary depending on industry, even production process and they cover virtually all of OR/MS, the 6s process is a dream chance to apply quantitative person's/statistician's skills.  In author's words, 6s is a '… tremendous opportunity for statistics professionals to make an impact on their organizations and their careers if they learn to ride the Six Sigma wave.'