Key points are not available for this paper at this time.
The use of discrete dependent variable regression models is well established in insecticide research. The policy impact of such research has increased due to concern over environmental effects of insecticides. One result of this concern was the creation of the U.S. Environmental Protection Agency, which registers insecticides and recommends dosages for their use (Anon. 1974). As part of the registration process, data to support minimum effective dosages and effective dosage ranges are required (Anon. 1975). The registration procedure is predicated on the hypothesis of a fixed relationship between dose and response for different populations.
Savin et al. (Thu,) studied this question.