Cause source - system v. employee

[wayne.mack@pec.com]

Regarding the statement that "95% (or whatever) of the problems are caused by the system," I would comment that the measurement is highly subjective and that identifying the "source" of the problem is not necessarily conducive to solving it.

If an employee makes a mistake, who is at fault?  Is it the employee, for not being careful?  Is it the system, for not providing enough training or fail safe systems?  Was it an assignable cause or a common cause?  What action, if any, should be taken to prevent reoccurrence of the mistake?

If you use a control chart as your definition of assignable causes for variation, you will come up with a very small percentage of assignable causes.  With the various permutations of rules that can be used, I don't think a precise, mathmatical value can be assigned, but it would probably lie at less than 10%.  This means that at least 90% of the causes, by definition, are not assignable and unknown.

This implies that many improvements can only be made through experimentation and not analysis.  This is where PDSA comes into play.  Propose an alternative, try it, measure the results, see if there is an improvement, and incorporate it if there is.

Determining who or what is the cause of problems is immaterial and shows you are on the wrong track.  Most variation is due to indeterminate (non-assignable) causes and it is not economically feasible to determine the cause.

Wayne Mack


Message posting through the Clemson CQI Web Server.