By Kim Niles
I'm sure all of us experience daily memos, emails, voice mail, procedures, forms, and reports that are nearly meaningless due to lack of context. Some ramble on with obvious passion but fail to explain what the heck they're talking about. Some are very beautiful with lots of statistical charts and graphs yet without simple contextual statements, these potentially higher order documents (informative beyond normal human perception and thought) are no more than any other type of first order information (i.e. 'I think it's the operator's fault').
Every Quality document should be fully self-contained or reference applicable context elsewhere. Simple statements of 'the assumptions made', 'what the objective of the project was', 'how the data was taken', 'how to read the graph', etc., MUST BE INCLUDED, or the document can be nearly worthless.
"A fact is like a sack- it won't stand up if it's empty. To make it stand up, first you have to put in it all the reasons and feelings that caused it in the first place."
- Luigi Pirandello (1867-1936), Italian author, playwright.
Contextual areas important to Six Sigma are as follows, in order of importance:1. Primary: Context that is absolutely necessary for long-term understanding such as the 5 W's (who, what, why, where, when) or related to subject, object, verb.2. Secondary: Scope or breadth, which might contain important aspects of our projects (i.e. analytic or enumerative studies), applicability, simplified abstracts, etc.3. Tertiary: Extended background or situation based, which might include important historical, political, cultural, or multilingual information, etc.
The Future of All Science Depends On Context
Recently while sifting through my pile of 63 printed postings from what might be iSixSigma's longest bulletin board thread ever (see What is a Stable Process), I began to realize that much of our frustration was contextual. Our goal of saving the future of all science from endless bickering over what process capability, control, and stability are comes down to simple context.
One driving force behind our discussion was a paper by William H. Woodall which just won him the Brumbaugh award for the largest single contribution to the development of the industrial application of quality control. Woodall stated that 'the distinction between common causes and assignable causes is context dependent.' As we discussed in the bulletin board thread, depending upon context used, one can say that a process can be stable but not capable or visa versa, both capable and stable producing no good parts, neither and produce good parts, both and be non-normally distributed, and both with either common or special cause variation. One can see the danger here of misleading others without proper context up front.
Walter A. Shewhart defined statistical control as follows: 'A phenomenon will be said to be controlled when, through the use of past experience, we can predict, at least within limits, how the phenomenon may be expected to vary in the future.' Note that in order to use this definition one is required to provide context of time and reference.
Next Page > Context As Applied To Six Sigma
Page 3 > Exponential Requirements / Exponential Gain
