A class of designs for mixture experiments based on augmented pairs.
A class of designs for mixture experiments based on augmented pairs.
A class of three-level response surface designs involving augmented pairs, which has the property that the initial factorial set of runs and the set of augmenting runs are orthogonally blocked, as regards estimation of the linear and bilinear terms in a second-order model, is extended to the case where orthogonality applies additionally to the second-order square terms. The advantage of this is that orthogonally blocked mixture designs may be derived from these extended augmented response surface designs by projecting them into the restricted mixture simplex space. A general result for complete orthogonality using augmented full or fractionally replicated factorial designs is provided, and several illustrative examples of designs suitable for projection are used to demonstrate the method.
central composite designs, mixture ingredients, process variables, projections, response surface designs
911-919
Prescott, P.
cf0adfdd-989b-4f15-9e60-ef85eed817b2
2004
Prescott, P.
cf0adfdd-989b-4f15-9e60-ef85eed817b2
Prescott, P.
(2004)
A class of designs for mixture experiments based on augmented pairs.
Communications in Statistics: Theory and Methods, 33 (11), .
(doi:10.1081/STA-200037934).
Abstract
A class of three-level response surface designs involving augmented pairs, which has the property that the initial factorial set of runs and the set of augmenting runs are orthogonally blocked, as regards estimation of the linear and bilinear terms in a second-order model, is extended to the case where orthogonality applies additionally to the second-order square terms. The advantage of this is that orthogonally blocked mixture designs may be derived from these extended augmented response surface designs by projecting them into the restricted mixture simplex space. A general result for complete orthogonality using augmented full or fractionally replicated factorial designs is provided, and several illustrative examples of designs suitable for projection are used to demonstrate the method.
This record has no associated files available for download.
More information
Published date: 2004
Keywords:
central composite designs, mixture ingredients, process variables, projections, response surface designs
Organisations:
Statistics
Identifiers
Local EPrints ID: 30007
URI: http://eprints.soton.ac.uk/id/eprint/30007
ISSN: 0361-0926
PURE UUID: 4a24176b-1ffa-4098-9f75-514cb2decc12
Catalogue record
Date deposited: 11 May 2006
Last modified: 15 Mar 2024 07:36
Export record
Altmetrics
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics