Hi Jim!

Your follow-up comments on Six Sigma are well stated and I agree with you.  I have seen too many people torture data until it confesses.  People often distort data in order to tell a story that they want to tell.  PROPER analysis of the data allows us to understand what our process are really doing.  Too often, people misinterpret their results because they don't really understand the tools they are using.

As a result, people develop and hone their "story-telling skills", which they like to show off in front of the boss.  I've seen it many times.

The most abused tool is the trend chart.  I really hate trend charts because they don't tell the whole story without proper analysis via a properly constructed process behavior chart (aka, control chart).

The tools of Six Sigma have been around a long time.  "Six Sigma" is merely a package or tool box, but it has gained an aura of mysticism among upper management.  The DMAIC process of a Six Sigma project is a discipline as much as anything and, properly used, can result in huge process understanding and improvements.  The key understanding, however, is understanding the nature of variation in processes.  You MUST understand whether your variation is "common cause" or "special cause" variation (Deming's terms, not Six Sigma) before trying to improve any process.  Otherwise, we simply "tamper" with the system and usually make things worse.  

It is a perception, but not true that "You can prove anything with statistics".  THE FIGURES DON'T LIE, BUT LIARS FIGURE - and that is the key issue!

Steven J. Moore

Rhinelander, Wisconsin

****

Have a comment? Send your email to jthompson@taii.com. Unless you tell us otherwise, we will assume we can use your name if we publish your letter.