Design of Experiments

A general overview of concepts is in the presentation file "ConcExperDesign_200801.ppt" (on this page). This presentation covers more than just DOE, and includes other concepts such as Regression, comparison of means, types of data, and many other topics.

(Thanks to Thom Nichols for permission to use this presentation.)

There are many different types of statistically designed experiments.

These include:

See also Relationship Between SPC and DOE

For a good overview of Experimental Design, see the presentation "Concepts in Experimental Design" (courtesy of T. Nichols) on this page.

Goals of a DOE study may include one or more of the following:

  • Finding the process inputs, and interactions, that most affect the process output(s) (understanding the system)

  • Finding a setting in the design space that optimizes some criteria of the process output (maximize or minimize process mean and/or process variation)

  • Finding an area of the design space where the process output is robust to changes in process inputs