Methods and Concepts of Biometrics Applied to Teratology

Some procedures are described which are used frequently in the statistical analysis of data from teratological studies. For the most part, the discussion here focuses on those species that produce litters with more than one offspring. There is a question of what teratological indices provide the most useful information. A discussion is presented on the proper choice of the experimental unit, the litter or fetus, for determining sample sizes. Owing to the large experimental variation often encountered in teratological studies, the desirability of replicated experiments is stressed. Measures of lethality and teratogenicity are discussed. Statistical tests for comparing treated and control animals are given. The use of fitting mathematical models to describe doseresponse relationships is discussed.