Abstract
D-optimality is a commonly used criterion to evaluate a design with respect to parameter estimation. The variance of the optimal dose estimate is another criterion for evaluating a design. The quantitative dose-response experiment involves fitting a model to data and estimating an optimal dose. Two techniques for estimating an optimal dose and three models are used to compare the variances of optimal dose estimates over nine equally spaced balanced designs and five fixed unequally spaced six-point designs. The results show that a design which is more D-optimal than another design does not necessarily produce optimal dose estimates with less variance.
Keywords
Dose-response, optimal design, optimal dose.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Recommended Citation
Rolka, Henry R.; Milliken, George A.; Schwenke, James R.; and Remmenga, Marta
(1991).
"THE EFFECT OF DESIGN AND DOSE LEVEL CHOICE ON ESTIMATlNG THE OPTIMAL DOSE IN A QUANTITATIVE DOSE-RESPONSE EXPERIMENT,"
Conference on Applied Statistics in Agriculture.
https://doi.org/10.4148/2475-7772.1425
THE EFFECT OF DESIGN AND DOSE LEVEL CHOICE ON ESTIMATlNG THE OPTIMAL DOSE IN A QUANTITATIVE DOSE-RESPONSE EXPERIMENT
D-optimality is a commonly used criterion to evaluate a design with respect to parameter estimation. The variance of the optimal dose estimate is another criterion for evaluating a design. The quantitative dose-response experiment involves fitting a model to data and estimating an optimal dose. Two techniques for estimating an optimal dose and three models are used to compare the variances of optimal dose estimates over nine equally spaced balanced designs and five fixed unequally spaced six-point designs. The results show that a design which is more D-optimal than another design does not necessarily produce optimal dose estimates with less variance.
