Most training involves the use of operant conditioning - the dog performs a behavior and the trainer delivers specific consequences for that behavior. There is sometimes a misconception that operant conditioning consists of the exclusive use of positive reinforcement. Both reinforcement and punishment procedures are aspects of operant conditioning, as Figure 2.1 demonstrates.
| Add (+) | Remove (-) | |
|---|---|---|
| Pleasant | Positive reinforcement |
Negative punishment |
| Unpleasant | Positive punishment (follow behavior with an aversive =>behavior descreases) |
Negative reinforcement |
Some examples:
- Owner gives a treat when dog sits to greet at the door. Behavior of sitting increases in frequency. (Positive reinforcement.)
- Owner steps back outside and shuts the door when friendly dog jumps up to greet. Behavior decreases in frequency. (Negative punishment.)
- Owner makes startling noise when dog jumps up to greet. Behavior decreases in frequency. (Positive punishment.)
- Person lifts up on leash and head halter and maintains tension until the dog sits, then releases tension. Behavior of sitting increases in frequency. (Negative reinforcement.)
Some humane trainers emphasize the exclusive use of positive reinforcement to establish desired behaviors and negative punishment to discourage unwanted behaviors. Others find that in some circumstances, with some dogs and owners, the inclusion of positive punishment and negative reinforcement procedures may be warranted.
The goal is always to minimize the dog's exposure to unpleasant events. Figure 2.2 shows in broad terms the relative emphasis that should be placed on each type of procedure if all four are to be used.
Figure 2.2: Operant Conditioning with Emphasis on Reward-Based Training
Certain unwanted behaviors, such as avoidance and aggression, are symptoms of underlying emotional states, such as fear. These emotional states are influenced by classical conditioning. The dog has learned to associate certain stimuli with certain other pleasant or unpleasant events. For instance, a man in a uniform might signify a threatening event to the dog, thereby triggering fearful or aggressive behavior.
Trainers should recognize that behaviors linked to emotional states can be influenced by operant conditioning without changing the emotional state itself. For example, even if a trainer teaches the dog not to lunge at men in uniform, the dog might still perceive them as threatening.
One option in such a case would be classical counter-conditioning to improve the dog's attitude towards uniformed men. If the dog is taught to associate these men with food or fun activities, the fearful responses may cease.
Trainers should not expect a dog to learn new behaviors by imitation - that is, by purposeful, goal-directed mimicking of another dog. However, social facilitation does occur, in which an existing behavior is triggered by a social situation. For example, a dog may be motivated to drink water out of a bowl when other dogs are doing it, because they are causing the water to splash in a way that attracts the dog.
Trainers should also recognize that a dog that has learned a behavior in one context will not necessarily perform it in another. They should design a training plan that will help the dog learn to perform the behavior to a range of different contexts. This tendency is known as generalization.
Trainers should:
- Have a working knowledge of how to consistently and effectively deliver both cues and consequences.
- Clearly identify target behavior, the behavior that the trainer seeks to influence.
- Continuously assess training success, discontinuing or avoiding the use of techniques that do not work.
- Give dogs frequent feedback about appropriate behavior.
- Choose an approach that gives the dog the best chance of success, rather than set the dog up to display undesirable behavior for the primary purpose of punishing it.
- Base training programs on the principles of learning rather than on anecdotal remedies.
