Pavlov's strategy was to study salivary processes in individual dogs over many test sessions. During each session, he placed dry food powder inside the dog's mouth and then collected the saliva. All went well until the dogs became experienced subjects. After several testing sessions, the dogs began salivating before being fed, usually as soon as they saw the laboratory assistant enter the room with the food powder. What Pavlov discovered was a form of learning in which one stimulus predicts the occurrence of another. In this case, the appearance of the laboratory assistant predicted the appearance of food (Carlson, 1997).
Pavlov designed experiments to discover exactly why the dogs were salivating before being given the opportunity to eat. He suspected that salivation might be triggered by stimuli that were initially unrelated to eating. Somehow, these previously neutral stimuli came to control what is normally a natural reflexive behavior. After all, dogs do not naturally salivate when they see laboratory assistants (Carlson, 1997).