Standardizing tests of mouse behavior: Reasons, recommendations, and reality

As more investigators with widely varying backgrounds enter the field of mouse behavioral genetics, there is a growing need to standardize some of the more popular tests because differences between laboratories in the details of behavioral testing and the pretesting environment can contribute to failures to replicate results of genetic experiments. It is argued here that we have sufficient knowledge to warrant a wise choice of a short list of standard strains and even details of apparatus and protocols for several kinds of behavioral tests. Equating the laboratory environment does not appear to be feasible. Instead, we need to learn what kinds of behavioral tests yield the most stable results in different labs and what kinds are most sensitive to the ubiquitous variations among test sites. Methods for making an informed choice of sample size for evaluating interactions between the laboratory environment and genotype are available and should be utilized in standardization trials. New resources for convenient sharing of data will greatly aid in collaborative and comparative studies involving several sites. Like the sequencing of an entire genome, test standardization is something that needs to be done only once if it is done properly, and the work will then benefit the field of behavioral and neural genetics for many years. (C) 2001 Elsevier Science Inc. All rights reserved.