Another good new blog on Statistical and Scientific Thinking [the broken link was removed] that is discusses the application of Deming’s ideas, by John Dowd. Short quotes from two posts:
The predictions made on the job are usually a lot less obvious and in most cases managers may not even be aware of the fact that they are making predictions. For example, much of management time is spent reacting to events. Some problem takes place and it’s management’s job to fix it. Some process of assessment is done and then a ‘corrective action plan’ is put into place.
So where’s the prediction.The plan is the prediction. The plan is developed in the hope if carried out in the proper way, some desirable result will occur that will eliminate the problem, solve it, or whatever. That plan is picked (one would hope) as being the one most likely to bring about the desired outcome.
Shewhart went at the problem a different way. He was trying to determine what the characteristics of process behavior were and began studying the output of different types of systems to see what happened. Knowing that the variability of processes took place over time, he began plotting time-series charts (sometimes called Run Charts) to see what patterns of the variation looked like.