Statistical Analysis of Results for Successive Tests on the same Organism

Abstract

IN a recent letter¹, Leech has pointed out an error of method which sometimes occurs in the analysis of experimental data; this error “arises from regarding successive tests on the same animal as independent”. With this criticism I am in complete agreement. However, as an example of this error, he refers, inter alia, to a paper² for the statistical analysis in which I am responsible. Bailey also, in a review³ of this paper, has made a similar criticism. The purpose of this note is to explain the method of analysis used, to indicate how it may be employed where the independence of successive observations cannot be assumed, and to point out that the criticisms of Bailey and Leech do not apply in this particular case.