A Two and a Half Day Course in the Design of Experiments

This course is designed for engineers, and is designed at a fairly sophisticated level. It works best if at least some in the class have some previous experience, whether good or bad, with designed experiments.

Each participant should have access to a computer with either Statistica or JMP available on that computer.

It is best for the course to be taught no more than 4 hours per day, with at least one multi week break between, for instance, the first 2 four hour sessions and the last three four hour sessions.

Topics covered:

Introduction
    Terms, Definitions for experimenting
    Learning Statistica or JMP basics

Comparative experiments - comparing two processes (or machines) or one process (or machine) with a standard

Split lot experiments - special case in common use for comparisons

Factorial experiments - looking at many factors

Fractional Factorial experiments - looking at many factors with fewer runs

Response surface methods (brief intro)

Review of strategy of experimentation

Note: The introduction to Response Surface Methods which can be done in this course is brief, and in no way a substitute for a full course on the subject.