Degree Type

Creative Component

Semester of Graduation

Summer 2019

Department

Statistics

First Major Professor

Max D. Morris

Degree(s)

Master of Science (MS)

Major(s)

Statistics

Abstract

The foldover technique allows for an expansion of a fractional factorial experiment with two level factors to achieve a resolution IV design from a resolution III design. Resolution IV designs allow for unbiased estimation of main effects if third order interactions and higher are 0. This paper lays out a scheme that generalizes the foldover technique to regular fractional factorial experiments with three level factors by augmenting the design matrix by a rotation vector. The criterion for choosing an optimal rotation vector is discussed in detail. It is shown that no perfect analog exists to always achieving resolution IV designs for regular fractional factorial experiments with three level factors although minimum aberration can be gained in all situations where resolution IV designs are not available.

Copyright Owner

Paris, Reid Vincent

File Format

PDF

Share

COinS