Planejamento de ensaios changeover na experimentação animal

Abstract

In experimentation with large animals, especially with dairy cows, it is very common to use alternative tests. In this type of trial, animals receive two or more treatments in sequence. The main justifications for the use of this particular type of assay are due to the high cost of the animals and the heterogeneity of these animals. These tests are basically classified into two types: Rotating Tests (Changeover) and Reversal Tests (Switch-back). To ensure that the effects of treatments are properly evaluated, pre-established rules and restrictions on randomization of treatments are necessary for planning such trials. Thus, the objective of this article is to present possible ways to plan changeover designs, and a routine was developed in the R software to determine the number of balanced changeovers of an order n, since in the literature the number of possible balanced changeovers was found only for experiments with a maximum of four treatments and the algebraic demonstration was not trivial until the time of obtaining the simulated results. It is concluded that the number of options to plan in changeover is much smaller and more restricted compared to planning in Latin square.