Efficient Nearly Orthogonal Deletion Designs
Abstract
There is a vast literature on the construction of orthogonal single replicate factorial designs in incomplete blocks. The reader is referred to Voss for the list of references. The concept of deletion designs was introduced in Kishen and Srivastava. The deletion technique in deletion designs was then used by many authors. This article considers deletion designs in three incomplete blocks and then presents a systematic method for finding the u.e. (unbiasedly estimable) and n.u.e. (not unbiasedly estimable) factorial effects. For n.u.e. factorial effects, the biased estimators (biased w.r.t block effects) are called the unadjusted estimators. Under the assumption that certain higher order interactions are negligible, the unbiased estimation of block effects contrasts and n.u.e. factorial effects, excluding the general mean, are possible. This makes the deletion design an orthogonal design. The unbiased estimators of n.u. e. factorial effects under the assumption are called the adjusted estimators. The relative efficiency in the estimation of a factorial effect is the ratio of the variance of the unadjusted estimator divided by the variance of the adjusted estimator.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1988
- Accession Number
- ADA197923
Entities
People
- Joan Mahoney
- Subir Ghosh
Organizations
- University of California, Riverside