DOG TRAINING IN PRACTISE: Teaching conditioned reinforcer to the dog

Dog training should be fun but also effective. Using positive reinforcement requires that we can "tell" the dog precisely when it is doing something right. If we use only unconditioned reinforcer (for example food and toys), particularly teaching new things to the dog, reinforcing could be inaccurate and too slow. Trainer rewards the dog always few seconds too late. Conditioned reinforcer (for example clicker, some sound or light) makes dog training much more faster and effective. Utilizing classical conditioning we can teach to the dog that some kind of sound signal predicts a reward.

You can use any kind of reward, however food is often simplest. Play the sound signal of your choice (clicker, good-word, whistle etc.) and then give a reward to the dog. Pay attension that for example your hand doesn´t move until the sound signal ends. Repeat it many times and remember to take breaks. Add some kind of distractions to the learning situations step by step, for example train your dog in different kind of places. Other people and dogs are also good distractions. Remember that when a new distraction comes along, your dog could stay and stare it/those. Deliver the reward still and don´t repeat the sound signal! Finally you can test the signal by playing it when dog is looking somewhere else. If your dog immediately looks at you and comes to you, it has learned the signal.

I speak in next dog training in practise-articles how to teach your dog some new tricks by using conditioned reinforcer!

DOG TRAINING THEORY: Reinforcers and punishers

Dogs gain practical information about the social and physical environment through the consequences of their behavior so they can control and manipulate significant events vital to their interests. Animal behaviorism defines reinforcements ans punishments as follows:
Reinforcement (R) increases the relative probability of the behavior it follows.
Punishment (P) decreases the relative probability of the behavior it follows.
So dog decides what is reinforcement or punishment, and what is not. We can only observe the dog and its gesture and behavior. It is also good to decide and know what kind of behavior you want before you start training, because reinforcement will affect to dog´s behavior. Example food is calming reinforcement to the most of the dogs, but playing with toys is exciting.

Reinforcers and punishers can be divided to positive (+) and negative (-):
Positive reinforcement (R+) occurs when dogs behavior is strengthened by producing or prolonging some desirable consequence.
Negative reinforcement (R-) occurs when dogs behavior is strengthened by reducing or avoiding some undesirable consequence.
Negative punishment (P-) occurs when dogs behavior is weakened by omitting the presentation of the reinforcing consequence.
Positive punishment (P+) occurs when animals behavior is weakened by presenting the previosly escaped or avoided consequence.

Positive punishment involves lots of risks and disadvantages. It is associated with the elicitation of various concomitant emotional states, especially frustration, anxiety and fear. Punishment usually makes dog to fear something that is present in the training session. Its very likely that dog´s fear is directed to the owner, because she/he is always present. Classical conditioning occurs.

In dog training, using both positive reinforcement and positive punishment has lots of disadvantages. First of all, the power of rewards weakens. Secondly, withdrawal of reward (negative punishment) isn´t so good tool in dog training anymore. It is very powerfull to "tell" to the dog when it is doing something wrong by withdrawing reward. Punishment doesnt also tell dog what to do. Its important to remember that severe punishment causes severe risks and disadvantages. Altough we wouldnt think the ethical side of punishment, there´s still lots of good reasons not use positive punishments. When we reinforce dog´s behavior, it becomes entrepreneurial. But using positive punishments in dog training could lead to dog´s passivity, it starts to avoid the things that could lead to punishment. And if punishment has been unpredictable and uncontrollable in dog´s perspective - dog is never quite sure when punishment is going to come, and never quite sure why it is administered - it can start to avoid all activity. In that case dog´s training is nearly impossible or at least very slow, because we have to use lot of time to get some action in dog. Positive reinforcement satisfy some psychological and physiological need. For example, when dog is hungry and it learns that sitting leads to food, after several positive experiences (sitting-->food) the probability that the dog will sit on cue is increased. Trough this simple lesson, the dog not only learns how to sit, but, more importantly, the dog learns that its action can control the environment. And that is a outcome that makes learning itself intrinsically rewarding. Reinforcement also exhibits many irregular and, perhaps, unanticipated characteristics. For example, while the opportunity to eat represents a very strong reinforcer for a hungry dog, the dog may also find just smelling the food reinforcing.

When we look reinforcers and punishers in dog training, it is good to remember that the provision of anything that the dog would rather be doing at any given moment may function as a reward. And on the other hand, anything that the dog would rather not be doing at any given moment might be used as a punisher. During the training process, dogs form certain expectations and predictions about outcomes associated with their behavior. So if a dog receives a reward that is significantly smaller than expected, the outcome is perceived as punitive (and disappointing), resulting in the trial rendering the response weaker. So I repeat myself: dog decides what is reinforcement or punishment, and what is not.

Timing and repetition play very important role in dog training. For a reinforcer to be effective, it must closely follow the target behavior and optimally, the reinforcer should be presented immediately after the target behavior is emmitted. Dog only learns if the consequences of its behavior follows relatively fast. Using only unconditioned reinforcer (example food and toys), particularly teaching new things to the dog, reinforcing can be difficult. Clicker trainers use clicker as a conditioned reinforcer, but it could be any kind of stimulus. Conditioned reinforcer is needed only when we want to reinforce very precisely, in second ot two. But teaching dog long-lasting activities, example to "stay", it is not usefull anymore.

When using reinforcement in dog training, it is important to remember chedule them right. During the early stages of training, a new behavior is reinforced every time it occurs and once a stable operant level is obtained, dog´s behavior is usually brought under the control of an intermittent schedule of reinforcement. For example when you are teaching your dog to "follow", intermittent reinforcement makes the behavior more resistant to extinction. Dog never knows when the reward is presented to it and it maintains good performance through the hole exercise.

DOG TRAINING THEORY AND PRACTICE: Classical conditioning

Russian psychologist, Ivan Pavlov, discovered Classical, or in the terminology of behavior analysis respondent conditioning, in 1920s. It seems that classical conditioning was discovered by chance. Pavlov was investigating dog´s digestion when he noticed that the most experienced dogs that he had been testing began to salivate before the samples of food were presented to them. He started to investigate this systematically. Throughout Pavlov´s experiment, dogs were exposed to a wide variety of arbitrary stimuli (example bell and light) presented contiguously with food. Pavlov has a special laboratory constructed in wich the experimenter and a dog were separated from each other in soundproof rooms. Discovery was that conditioned stimulus (bell) began to trigger the same reactiong (salivate) in dogs that unconditioned stimulus (food) did. The associative bond between the conditioned stimulus (CS) and the unconditioned stimulus (US) is strengthened when the CS consistenly occurs just before the representation of the US and is weakened when the CS and US occur indepently of each other. So classical conditioning is unintentional and subconscious reaction to a certain stimuli.

It is important to understand how classical conditioning works and hopefully a few everyday examples will do that. Most dogs respond readily to sound of doorbell ringing and after several repetitions, in which the bell signals the arrival of someone at the door, the dog may begin to example bark. The dog has learned to correlate a previosly insignifant event (doorbell) with a significant on (visitor) and now when the doorbell rings the dog dances with excitement anticipating the visitor´s entry and greetings. But what starts to happen if dogs owner doesn´t open the door when doorbell rings? After several sessions of nonreinforced exposures to the bell, the dog will gradually inhibit its anticipatory reactions and finally ignore the sound altogether. We can say that the conditioned effect has been extinguished.

Classical conditioning has also an important role in the development of anxiety and fears. For instance, dogs easily develope fears associated with the veterinary clinic, expecially if they have undergone painfull procedures there. It may help to prevent negative associations to the veterinary clinic when we provide dogs or puppies treats and other pleasurable experiences while being examined. But what if dog doesnt it eat in that situtation? Its probably too stressed (or not hungry) and you have to start giving food to the dog earlier (example in the hospital parking lot). Lifelong phobic reactions can also occur as the result of a single traumatic event and dogs suffer a broad spectrum of phobic fears, most of which are established and reversed through classical conditioning.

But does Pavlov have anything really good information for us to use in dog training? You may have heard about clicker training. When we examine clicker training through classical contioning, clicker is conditioned stimulus (CS) and food (or some else reinforcement) is unconditioned stimulus (US). Clicker is a mechanical device that makes a short, distinct "click" sound which tells the animal exactly when they're doing the right thing and its combined with positive reinforcement. You can also use something else than clicker, example light, whistle or some other sound. Other important information of Pavlovian conditioning is counterconditioning. It plays avital role in the learning and unlearning of emotial reactions. So in dog training, resolving fears and other problems (example phobias, separation anxiety and aggressiviness), classical conditioning may be required. Classically conditioned responses are largely autonomous and independent of central control, unlike instrumental behavior. Dogs dont choose to feel fearfull or anxious. To be controlled, an aversive emotion like angre or fear, must be countered by the elicitation of an ever stronger and incompatible emotional response.