2x2 Factorial Experiments in Incomplete Groups for Use in Biological Assays

Abstract

Factorial assays with 2 doses of the standard drug and 2 doses of the unknown have proved experimentally practicable and easy to analyse when based on a graded response. In many cases, potential sources of variation can be segregated by means of randomized block or Latin square designs. When homogeneous groups of 4 are less practicable than those of 2 or 3, the balanced incomplete group may be substituted. The design and analysis of factorial expts. in balanced pairs and triplets are discussed and illustrated by 2 parathyroid assays upon the serum calcium in dogs.