Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1149238 | Journal of Statistical Planning and Inference | 2013 | 8 Pages |
Abstract
For two-level factorials, we consider designs in N=2 (mod 4) runs as obtained by adding two runs, with a certain coincidence pattern, to an orthogonal array of strength two. These designs are known to be optimal main effect plans in a very broad sense in the absence of interactions. Among them, we explore the ones having minimum aberration, with a view to ensuring maximum model robustness even when interactions are possibly present. This is done by sequentially minimizing a measure of the bias caused by interactions of successively higher orders.
Related Topics
Physical Sciences and Engineering
Mathematics
Applied Mathematics
Authors
S. Huda, Rahul Mukerjee,