Determination of the No-Observed-Adverse-Effect-Level by controlling over-estimation probability

Research Report SRR98-007

Determination of the No-Observed-Adverse-Effect-Level by controlling over-estimation probability

Toshio Sakata and Takashi Yanagawa

Abstract: NOAEL determination problem in a toxic dose-response model is discussed for binomial responses. A new procedure and its modified version are proposed. They focus on controlling the probability of over-estimation of NOAEL. The methods use the observed 2nd kind of errors and its estimates. The performance of the procedures is investigated both by theory and by numerical calculations. For lage sample cases we also propose a procedure using estimates of 2nd kind of errors by Metropolis walk. It is shown that this procedure has a similar performance with the one using exactly calculated estimates of the 2nd kind of errors.

This service is maintained by the Mathematical Sciences Institute (MSI)
Comments to webmaster@maths.anu.edu.au URL: http://wwwmaths.anu.edu.au/