Replicating Empirical Research In Behavioral Ecology: How And Why It Should Be Done But Rarely Ever Is

Abstract

That empirical evidence is replicable is the foundation of science. Ronald Fisher a founding father of biostatistics, recommended that a null hypothesis be rejected more than once because "no isolated experiment, however significant in itself can suffice for the experimental demonstration of any natural phenomenon" (Fisher 1974:14). Despite this demand, animal behaviorists and behavioral ecologists seldom replicate studies. This practice is not part of our scientific culture, as it is in chemistry or physics, due to a number of factors, including a general disdain by journal editors and thesis committees for unoriginal work. I outline why and how we should replicate empirical studies, which studies should be given priority, and then elaborate on why we do not engage in this necessary endeavor. I also explain how to employ various statistics to test the replicability of a series of studies and illustrate these using published studies from the literature.