A computer program for selecting animals for control and experimental groups in biochemical studies
- 1 June 1986
- journal article
- research article
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 2 (2) , 107-109
- https://doi.org/10.1093/bioinformatics/2.2.107
Abstract
The standard method for selecting experimental animals for control and experimental groups is by a randomization process in which each animal is placed into one of the groups without looking at any of its individual differences. An alternative, active matching process is proposed by which three to five animals are chosen to form ‘experimental units’ that are closely matched with respect to average weight and with respect to distribution of weights (standard distribution values). Several such units are then randomly assigned to each control or treatment group and, at the end of the experiment, the animals within each unit are pooled prior to analysis. The merits of this approach and a computer algorithm for making the selections are described.This publication has 0 references indexed in Scilit: