Multi-variate probit analysis.

SUMMARY This paper is concerned with the situation in which samples from a population of biological organisms are exposed to a stimulus at various levels, the effect on the organisms being manifested in terms of the observable quantal responses of two or more separate physiological systems. A method of analysis is put forward, based upon the notion that each organism may be characterised by a tolerance vector, the components of which correspond to the reaction of the various physiological systems. In this way, it is possible to assess the relation between the responses of the different systems in terms of the correlation structure of the components of the tolerance vector, as well as the dosage-response relation for each separate system. The problems of estimation and hypothesis testing are discussed and the proposed methods are illustrated by a study of the prevalence of respiratory symptoms in a human population.