摘要: Wildlife ecology and management studies have been widely criticized for deficiencies in design or analysis. Manipulative experiments-with controls, randomization, replication space time-provide powerful ways of learning about natural systems establishing causal relationships, but such are rare our field. Observational sample surveys more common; they also require appropriate More important than the analysis individual is metareplication: entire studies. Similar conclusions obtained from same phenomenon conducted under differing conditions will give us greater confidence generality those findings would any single study, however well designed executed.