For a bat-response study, researchers randomly assign the order in which trails are visited. At each site, they record two minutes of silence, play a short call, and record the next two minutes. Later, assistants who do not know which clips came before or after the call count the bat sounds.