No CrossRef data available.
Published online by Cambridge University Press: 15 May 2009
Optimum forms of moving averages are derived for the estimation of LD 50 in the situation where no assumptions are made about the form of the dose-response distribution. The theory is also applied to uses of moving averages based on the angular transformation of the percentage response. A numerical example illustrates this application, and its results are compared with the corresponding probit and logit estimates.