© 2004 Cayman Business
Systems
Rev:
Pre-G3 - Rendered Thursday,
February 12, 2004
•If the problem is a
manufacturing process, begin to establish a stable
process. Once the process is stable,
definition of the potential cause will be
clarified.
•If design
causes are identified, screening experiments
may help identify the key variables
which are affected by subsequent processes.
Design changes may be appropriate.
•Four or
five potential causes should be identified
to investigate. Identifying several potential causes forces the team to address multiple possibilities rather than searching endlessly for a single cause. An implicit part of problem analysis is investigating potential causes in parallel rather that in series.