TY - JOUR
T1 - A comparison of three approaches for constructing robust experimental designs
AU - Agboto, Vincent
AU - Li, William
AU - Nachtsheim, Christopher
PY - 2011/3/1
Y1 - 2011/3/1
N2 - While optimal designs are widely used in the design of experiments, a common concern is that they depend on the form of an a priori assumed regression model. If the assumed regression model is not the same as the true, unknown regression function, the optimal design might not be a good choice. Several useful criteria have been proposed to reduce the dependence of optimal designs on a single, assumed model, and efficient designs have been then constructed based on the criteria, often algorithmically. In the model robust design paradigm, a space of possible models is specified and designs are sought that are efficient for all models in the space. The Bayesian criterion given by DuMouchel and Jones (1994) posits a single model that contains both primary and potential terms. In this article we propose a new Bayesian model robustness criterion that combines aspects of both of these approaches. We then evaluate the efficacy of these three alternatives empirically.
AB - While optimal designs are widely used in the design of experiments, a common concern is that they depend on the form of an a priori assumed regression model. If the assumed regression model is not the same as the true, unknown regression function, the optimal design might not be a good choice. Several useful criteria have been proposed to reduce the dependence of optimal designs on a single, assumed model, and efficient designs have been then constructed based on the criteria, often algorithmically. In the model robust design paradigm, a space of possible models is specified and designs are sought that are efficient for all models in the space. The Bayesian criterion given by DuMouchel and Jones (1994) posits a single model that contains both primary and potential terms. In this article we propose a new Bayesian model robustness criterion that combines aspects of both of these approaches. We then evaluate the efficacy of these three alternatives empirically.
KW - Bayesian designs
KW - D-optimality
KW - Model-robust design
KW - Supersaturated design
UR - http://www.scopus.com/inward/record.url?scp=84860815267&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84860815267&partnerID=8YFLogxK
U2 - 10.1080/15598608.2011.10412046
DO - 10.1080/15598608.2011.10412046
M3 - Article
AN - SCOPUS:84860815267
SN - 1559-8608
VL - 5
SP - 1
EP - 11
JO - Journal of Statistical Theory and Practice
JF - Journal of Statistical Theory and Practice
IS - 1
ER -