Sankhya: The Indian Journal of Statistics

2007, Volume 69, Pt. 2, 344--357

Maximal Rank - Minimum Aberration Regular Two-Level Split-Plot Fractional Factorial Designs

Mike Jacroux, Washington State University, USA

SUMMARY. Regular two-level fractional factorial designs are often used in industrial experiments as screening experiments. When some factors have levels which are hard or expensive to change, restrictions are often placed on the order in which runs can be performed, resulting in a split-plot factorial design. In these cases, the hard or expensive to change factors are applied to whole plots, whereas the easier or less expensive to change factors are applied to the subplots within the split plot designs. For such experimental situations the minimum aberration criterion has been used by a number of authors to find optimal regular fractional factorial split plot designs. In this paper, we suggest an alternative criterion called the maximal rank-minimum aberration criterion for selecting optimal fractional factorial split plot designs and study how this alternative criterion performs in terms of the optimal designs it selects, and how it compares to the minimum aberration criterion.

AMS (2000) subject classification. Primary 62K15, secondary 62K10.

Key words and phrases. Minimum aberration, sub plot, whole plot, word length pattern, estimable effect.

Full paper (PDF)