The aim is to reduce variation

["John McConnell" <wysowl@msn.com.au>]

The discussion on control charts has been interesting, but one can't help
but wonder about their usefulness in some organizations.

I use control charts extensively, and often use both an average and range
chart as well as a single point and moving range chart.  As the discussion
has mentioned, there are many other useful statistical methods, but if my
aim is to discover whether the process under examination is stable, I use a
control chart first.

However, examples abound where control charts are prepared and are obvious
everywhere one looks, but where they are next to useless.  This often is the
case even where the charts are properly prepared.  The most common problem
is that the management team do not have reducing variation as a primary aim.
In addition, many still think control charts are primarily for operations.
Often, a comment to the effect that the best place to start using control
charts is with the board/executive reporting system has been made here.
Sounds like good sense to me.

We are not surprised to note that companies where the management are
committed to reducing variation in everything they do generally use control
charts and other statistical methods to far greater effect.

What is the AIM?  In one Industrial Gas facility, nearly all production
processes were in control and displaying very low levels of variation.  Not
one control chart had been produced.  Nonetheless, the production manager
had led a crusade against variation in all aspects of production, and
without the support of his superiors.  He was very successful.  The addition
of control charting techniques helped to further reduce variation and to
create a solid approach to minute-by-minute process control.

Nonetheless, we would be wise to recall that this manager was very
successful at improving quality and productivity as well as reducing costs
without any statistical techniques more sophisticated that run charts and
histograms.  He was successful because he dedicated himself and his staff to
reducing variation.

Those of you who attended Dr. Deming's seminars will recall him pausing
occasionally as he discussed control charts, single source supply, the
Nelson Funnel Experiment and a host of other issues to remind his audience
that the AIM was to reduce variation.  Nothing in our world is good or bad;
neither right nor wrong, until couched in terms of the AIM.

At one optical lens factory the manufacturing manager took his business from
one that was struggling to one that was the benchmark in two years by giving
everybody the same single, superordinate AIM: to reduce variation.  This aim
replaced or became superordinate to all other objectives such as
departmental objectives, productivity targets, machine
availability/utilisation, and so on.

It does not matter how many control charts are prepared, nor how technically
well they are prepared, if the AIM is not to reduce variation, and if
management are not leading the organization on a crusade to conquer
variation, lots of pretty charts will not help much.

Remember the 15th Point!

John McConnell
wysowl@msn.com.au
====================================================================