Spring Seminar Series - March 4, 2003
University of Minnesota
School of Statistics
College of Liberal
Arts
Generalized Resolution and Minimum Aberration:
Ideas and Examples
Boxin Tang
Department of Mathematical Sciences
The University of Memphis
Tuesday, March 4, 2003
4:00 PM, 115
Ford Hall
Minneapolis, East Bank Campus
Social at 3:30 PM, 300
Ford Hall
Abstract
The talk discusses the ideas of generalized resolution and minimum aberration
in experimental design. The approach is informal, mainly through examples.
I first review the criteria of resolution and minimum aberration associated
with regular fractional factorial designs, and then examine how these criteria
can be generalized to non regular fractional factorial designs. Two key
ideas are involved in this process of generalization. One is that we would
like a design to have good projection properties, and the other is that
a good design should be robust in some ways. As it turns out, the criteria
of generalized resolution and minimum aberration do just that.