SOT 570 is the first course in a three-week series designed to teach students the principles
of test design and analysis. Satisfying the need for statistics to deal with variation in test data and the concepts of the scientific method (hypothesis testing), this course introduces multifactor designs for varying multiple test conditions simultaneously while retaining the ability to link cause-to-effect. It explores design options with factors at two or more levels, including
strategies for running a fraction of these combinations, designs for nonlinear response spaces (response surface methods), and a number of design augmentation schemes to resolve ambiguities in test outcomes.

SOT 580 is the second course in a three-week series, building up to the fractionated factorial and second order designs with any number of variables at any number of levels. Students learn basic techniques and processes needed to create a statistically rigorous test
in the physical sciences. A well-designed experiment can lead to reduced development lead time for new processes and systems, more rigorous test and evaluation, and systems that have superior function and reliability. Using examples from military T&E and the student’s own problem field, the course has a strong applied flavor and is intended to graduate practicing
experimentalists. Both design and statistical analysis issues are discussed.

SOT 590 is the final course of three, one-week applied statistics courses covering the fundamentals of experimental design. This third week extends DOE concepts into the full complement of test matrix alternatives to handle many variables over multiple testing periods (missions, days, and sessions). Efficient fractional factorial designs for 2- or more level variables are addressed. Another primary topic is response surface methods, including designs for nonlinear systems and strategies to find optimal settings for the test variables. Multiple effectiveness measure studies are also covered. In addition, we discuss proper experimentation and analysis practices when the principle of randomized run order is not practical. Design Expert and JMP software programs are used during this course. The graduate will be fully trained using Design Expert to practice the basic concepts of real-world DOE and able to exercise the discipline independently.