Basic principles of rewards-based (clicker) training

Basic Principles of Training

Why Rewards?

Training is a dog’s work.  When you work, do you work to please your bosses?  Would you work as hard if you stopped receiving a paycheck?  If you would work just to please your boss without a paycheck, you a one-in-a-thousand employee, just as a dog that would work without a paycheck (i.e., strictly for praise or to please) is a one-in-a-thousand dog.  We should not expect our dog to work without pay any more than our boss should expect us to.  We should also realize that we don’t really work for money but for the things that money can buy:  security, food, and other comforts and pleasures.  Since dogs lack our abstract reasoning abilities, they need more direct motivation to work:  food, attention, praise, sniffies, freedom, and toys/fun. 

Traditional training involves a corrective collar (“choke chain”) or prong collar and leash-corrections as punishment.  This is a shame, because the use of rewards is scientifically proven to work.  The use of rewards is based on behavioral science, mostly on a principle called operant conditioning, the same kind of sophisticated training that marine-mammal-trainers use on dolphins.  Dog-trainers often refer to this type of training as positive reinforcement or, more simply, as clicker-training.  Dogs trained using this method often become more like the way most of us think of dolphins:  friendly, intelligent, and very willing to work.

A clicker used along with rewards results in a trainer that has outstanding observation skills and timing, which are vital skills for any trainer to learn.  A clicker-trained dog is much less likely to develop bad side-effects, such as anxiety or unwanted aggression, sometimes associated with “traditional” training methods.  He often becomes much friendlier toward people.  He is also more willing to offer new behaviors because he lacks the inhibition that punishment can cause.  Finally, a reward teaches the dog what to do, which is more efficient than the dog figuring out through punishment and trial-and-error what not to do.  The clicker is a form of communication that the dog understands.

Many traditional trainers who believe their dogs work to please them really have dogs that are working to avoid punishment.  And well-meaning trainers who avoid supplying treats or other rewards to their dogs often inadvertently resort more and more to punishment/corrections when petting, praise, and a “happy voice” fail to sufficiently motivate.

Why use a clicker? 

The click itself is a powerful communication-tool.  According to Don’t Shoot the Dog, by Karen Pryor, clicking works for your dog better than rewards alone, because it bridges time and distance.  For example, let’s say you want to train a puppy to go potty outside.  The puppy goes potty, but by the time you deliver its treat, it is chasing a dragonfly.  You just rewarded the puppy for chasing a dragonfly!  If you click the dog just as it finishes potty, it will understand more easily when and why it earned a treat.  The clicker can bridge distance:  in a more sophisticated training session you may want to train the dog to sit reliably 30’ away.  You can’t deliver a treat from that distance, but you can deliver a click!  (By the way, if your dog is deaf, you can use a hand-signal, such as a quick “thumbs-up” sign, instead of the click, in all the exercises in this flyer). If your hands are full, a special word that the dog doesn’t hear very often, like “yes!” spoken quickly and enthusiastically, can be used as a substitute.

Types of Rewards to Use

What the dog finds more rewarding or less so depends on the particular dog and his situation.  A dog that has been recently deprived of human contact, for example, such as a dog that has spent a few weeks at a shelter, will be more willing to work for praise, petting, and attention.  A dog that is hungry will be more food-motivated, and a not-so-hungry dog might be more willing to work for some stupendous treat than for kibble.  You need better rewards in a high-distraction environment than in a low-distraction one.  A dog that loves to chase squirrels more than life itself is probably going to disregard your orders even if you are holding steak in your hand.  In situations like this, you need to use management to get your behavior, such as using a leash and head-collar for better control. 

            Another important point about rewards is variety.  The dog will work harder if you vary the types of rewards, even the types of food rewards, and if you surprise him once in a while (such as he’s expecting a sliced carrot but you give him a small piece of a peanut-butter treat, instead).  For the first time the dog gives you a behavior you’ve been working on, for a particularly well-executed behavior, or for a behavior done well in a difficult circumstance, such as a distracting environment, give the dog a jackpot:  a particularly good reward after the click (like a hunk of steak) or ten pieces of his usual treat given one-after-another rapid-fire, with lots of praise and enthusiasm.  If you have been working on fetch, for example, and the dog brings a ball to your hand for the first time, it might be best to jackpot that event and switch to some other activity during your training session.  Make bringing the ball to hand pleasant and memorable.

Be careful what you give to young puppies, since their stomachs are sensitive.  For them it is best to reserve some of their regular kibble and use it as training treats.  Older dogs will work for kibble, too, provided that they are in a low-distraction (boring) environment such as inside the house.  Some dogs will even take “diet” treats in such an environment, such as Cheerios, coarsely chopped raw carrots, or pieces of apple.  Experiment to find out.  A training-treat should be pea-sized or smaller and moist.  Small amounts of “people food” can be a terrific motivator!  Hard, crunchy dog-biscuits and similar treats are not good for training because they take too long to chew and swallow.  The extra time slows the training process considerably and can become frustrating for dog and trainer alike.

You will want a variety of higher-value treats for use in more distracting or difficult situations.  The training treats need not be nutritionally balanced as long as you use a limited supply of them (no more than 10% of the dog’s caloric intake).  If the dog has a health condition such as pancreatitis or you or otherwise in doubt, ask you veterinarian for advice.  To save money, you may want to bake your dog’s training treats, yourself.  Check under the “Further Resources” section for some ideas on home-made treats.

Treats are very convenient to use, but they are not the best reward for all dogs in all situations.  If your dog is tug-crazy or ball-crazy enough, he will prefer his toy to even the tastiest of treats, even if offered a hunk of steak instead.  In a situation like this, use the toy(s) instead of the treats.

If your dog is overweight, you have several options to be decided in consultation with your veterinarian:

1. Cut back on the dog’s regular food and replace the calories with the training treats (not to exceed 10% of total caloric intake unless nutritionally balanced). Try using a type of kibble different from what he usually gets.  I would not use semi-moist dog foods as the dog’s regular fare, but I might as a training treat!
2. Bake skinless, boneless chicken breast (purchased on sale or in bulk) and chop it up into treats. Store the excess in your freezer and thaw in the refrigerator as needed.
3. Use treats to train your dog for some fun exercise:  to jump through a hoop, to play fetch or to tug, etc.  There are explanations on how to do this in Section III.  Once he has learned, use the game itself as a reward whenever possible instead of treats.
4. Use low-calorie treats.  Take Cheerios and store them for a while in the refrigerator in a bag with something greasy or stinky, like hot dogs, to make them smell more interesting to the dog.
5. Use “life-rewards” such as the dog’s regular feeding times, or the chance to go through a door, through a gate, or for a walk, as training opportunities.
6. Measure the dog’s kibble but do not give it to him in a bowl.  Make him work for every handful at various times throughout the day.  Kibble works better in low-distraction environments and situations.

What Not to Feed Your Dog:

Spoiled or moldy foods, yeast dough, salty foods, and food with a high fat content can upset your dog’s stomach.  Chicken bones have already been mentioned as a choking hazard. Onions, grapes or raisins, alcoholic beverages, rhubarb leaves, onions, chocolate, coffee, and macadamia nuts are all poisonous to dogs.  Also beware of anti-freeze, particularly those that contain sweet-tasting ethylene glycol.  A few teaspoons of it can kill a dog.  (Some types of antifreeze made with propylene glycol are less toxic.  Further information on potential toxins and hazards to your dog are available through the HSUS and ASPCA websites).   

A good trainer is constantly observing a dog to see what interests him the most in various circumstances.  One of my dogs will work in a moderate-distraction environment for tug and for really good food.  My other dog is easily distracted, very interested in sniffing and the environment, and hard to motivate consistently.  A chronic puller, she will heel nicely during a section of our walk near some wet drainage-ditches.  If she is a very good girl, I will reward her by letting her race up to the culverts to sniff and hunt for frogs.  In this particular circumstance, for this particular dog, the sniffing/hunting reward is much more motivating than food.  Alas, I cannot carry those drainage-ditches with me for use in other parts of the neighborhood!  For most dogs and trainers, food is one of the most useful motivators because it fulfills one of the dog’s primary (survival) needs and because it is convenient for the handler to carry and dispense.