Determining the Validity of Preclinical Behavioral Assessments for Extrapolation to a Clinical Setting

Abstract The use of animals is essential to understanding the relationship between the brain and behavior. While animal models of human cognition provide a unique opportunity to study the output of the brain in its entirety, they also present unique challenges. To bridge the gap between preclinical results derived from research using laboratory animals and clinical predictions, five types of validity must be taken into consideration: ecological, face, convergent, construct, and predictive validity. Just as there is typically no “best test,” there is typically no hard rule about which of these different validities to emphasize: this will depend on the particular research question at hand. However, utilizing testing endpoints that are well defined and that can be obtained from human and nonhuman subjects will enhance the likelihood that the testing of a novel hypothesis using a preclinical species will result in scientifically relevant predictions for the clinical setting.