Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
10526613 | Statistics & Probability Letters | 2005 | 8 Pages |
Abstract
When our purpose is to rank the factors by the size of their effects, estimation of the factor coefficients should be the criterion for selecting a design. Thus, the average or maximum VIF values would appear to be the best measure for selecting a screening design, in addition to the behavior of the design under the well known alphabetic optimality criteria. In this paper, we perform an evaluation of several non-orthogonal saturated designs with two levels that are well known in the literature by studying their behavior under certain optimality criteria. Some new designs that have been found through a computer search are also considered.
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Statistics and Probability
Authors
H. Evangelaras, C. Koukouvinos, S. Stylianou,