Planning Herbicide Dose-Response Bioassays Using the Bootstrap
- 239 Downloads
Dose response herbicide bioassays generally demand large amounts of time and resources. The choice of doses to be used is thus critical. For a given model, optimum design theory can be used to generate optimum designs for parameter estimation. However, such designs depend on the parameter values and in general do not have enough support points to detect lack of fit. This work describes the use of bootstrap methods to generate an empirical distribution of the optimum design points, based on the results of a previous experiment, and suggests designs based on this distribution. These designs are then compared to the Bayesian D-optimum designs
KeywordsD-optimum designs Bayesian D-optimum designs Bootstrap Non-linear models Dose-response models
Unable to display preview. Download preview PDF.
- Fedorov, V.V. (1972). Theory of Optimal Experiments. London: Academic Press.Google Scholar
- Powles, S.B. and Holtum, J.A.M. (Eds) (1994). Herbicide Resistance in Plants: Biology and Biochemistry. Boca Raton: Lewis.Google Scholar
- Souza, G.S. (1998). Intwdução aos Modelos de Regressão Linear e Não-linear. Brasília: Embrapa.Google Scholar