Please take a moment to read the following wall of text. I will define a number of concepts that will make following along easier.
A behavior that elicits a consequence is called an operant behavior. Operant conditioning concentrates on the relationship between various outcomes of operant behaviors. Outcomes can be positive consequences, negative consequences, or punishment.
Positive reinforcement is when you encourage an operant behavior by rewarding it with something good. Examples of rewards include treats, playing with a toy, silly talk, petting, running and pretty much anything that your dog likes. In terms of physiology, most positive reinforcement methods stimulate a dopamine response at the basal ganglia (brain stem). It makes them happy, feel good, and makes him want to do the action again.
Negative reinforcement is a very subtle concept, and it can have profound effects on what your dog learns. Negative reinforcement is when you take something bad/unpleasant away. Contrary to popular belief, negative reinforcement is NOT a punishment.
Punishment is when you do something unpleasant in response to a behavior your dog exhibited. This includes spraying with a water bottle, smacking, pinching, blowing, shouting, rubbing noses in urine and feces, news paper hitting, and any other aversive behavior. I personally would never condone punishment as an effective way to train your dog. Punishment tends to stimulate an adrenal response, which is also known as the "fight or flight" response hard wired into pretty much everything that lives. In addition to the adrenaline response, all sorts of stress hormones can be released as well. These stress hormones interfere with the learning functions of the brain, which is completely counterproductive when it comes to training. Using punishment can lead to a fearful dog, which in turn can lead to a dog that is unstable or one that will bite (which is a fear response) in stressful situations.
Working from the outside in.
Traditional training methods involve physically manipulating your dog's body and then using a cue with those manipulations. For instance; Steadily pulling upwards on a collar, the dog has two choices. Choke or sit. The dog Sits. You say "sit" as soon as his bum moves towards the ground. Another example would be pulling your dogs collar towards the ground to teach them "down".
These methods do work, and the dog does learn, but it is not the most effective way to teach your dog.
Working from the inside out.
This is where clicker training shines. You are teaching your dog to WANT to listen to you. Every interaction you have with the dog ends in a positive outcome. When you use clicker training, you are teaching your dog how to learn. You open up an amazing avenue for communication which helps improve the canine-human bond, which results in your dog wanting to listen and learn. In short, you get your dream dog and your dog gets his "dream" owner.
Contrast the above example of "sit" and "down" with the following.
toy poodle pit-bull terrier, will not stop barking at everything that moves. You've tried yelling (which only makes her bark more, because she thinks you are barking too), smacking her with a news paper, spraying her with water. Nothing seems to stop her from barking. You ask a friend how they managed to train their dog to be a good canine citizen. He replies "Well gee-whiz, clicker training." You decide to learn all about clicker training.
Understand this; If a dog is bored, under-exercised, and generally under stimulated, barking turns into a very rewarding and fun game to pass the time. In most cases, the reward is the intrinsic value of barking itself. It's fun, it's something to do, and it normally gets a response. We need to figure out a way to make it so that barking is no longer rewarding
The most common way to teach a dog to stop barking with clicker training is to use the idea of behavior extinction. Extinction is when a behavior or action is no longer rewarding, so the dog stops doing it because it simply isn't worth doing. The way we accomplish this is to change how your dog is receiving his reinforcement to bark. We are going to teach your dog a new trick, and get rid of an annoying habit all in one swell foop (or is that fell swoop). More on this later, if I get enough requests I will put up a step-by step on how to teach a dog to stop barking.
Read the next section for clicker training theory.
Step 1: Dog Training Equipment
There are many types of dog training collars, I will outline the most common ones here.
This is my personal preference, it is cheap, functional, and humane. I don't use a collar to control my dog, I use it to hold her dog tags and my phone number. If a dog is trained appropriately, you shouldn't need to control them via a leash or collar. Applying pressure around the neck of a dog actually causes an adrenal response, which results in keying your dog up even more. In most cases, pulling on your dogs leash will be counter productive.
For years, this was the classic choice for obedience training. When the dog pulls ahead of you, the collar tightens around their neck. The discomfort caused by the collar teaches your dog to not pull (in theory) although I have seen many a dog hacking up a lung with a choke chain.
I personally think that there is a special place in hell reserved for the person that invented this collar (well maybe not, but I think they are totally useless and cruel). Many people claim they need to use a prong collar to penetrate their dogs long, thick hair. People claim that the prongs don't hurt their dog. I have seen gouges in dogs necks where these "harmless" prongs have been used to teach a dog to heel. Again, pain and tension on a dogs neck results in an adrenal response which is always counter productive to learning. Show me a dog in a prong collar, and I'll show you a dog that isn't' well trained (i.e a lazy owner).
The harness was developed to allow a dog to pull loads safely, and without discomfort to their neck. You often see small dogs wearing harnesses, which I think is kind of silly. The great majority of non-working dogs that wear a harness, wear them because the owners haven't taught them to heel. Lets think about this for a minute. Your dog chokes on their collar when you are walking them, so you put them in a harness instead. Congratulations, you have just taught your dog how to pull.
Leashes can be divided into a couple of different categories. Extendable and non extendable. The materials that leashes are constructed out of vary widely. For training a new dog, I prefer a 6 ft long 1 inch wide nylon leash. They are cheap, functional and get the job done. If I am training a puppy, or a small dog then I generally use a 1/4in wide leash. The full inch wide leashes are much to thick for a small dog :) I prefer the 1in wide leash, because if the dog suddenly bolts for something it is much easier on your hands. Once my dog is trained up, I tend to use a simple leather leash. This is purely for aesthetics and reflects more on my sense of style than anything else.
Yet another silly type of dog accessory. This type of leash was invented for working Hounds. The trainer could give the dog a lot of head room to track scents, and then quickly take up the slack to prevent tangling on ground debris. In an urban environment, these leashes are downright unsafe. A strong dog can pull hard, overwhelm the locking mechanism, and then run right into the road. Additionally, these leashes are nearly useless for everyday training. Save your money, get a generic 6 ft nylon leash.
Any clicker or short noise maker will do. Heck, you could even just snap your fingers.
What makes your dog tick? Food, play, silly talk? Make sure you have a ready supply of dog treats. Nylon treat bags that clip onto your belt or waist band are particularly handy. If not, a zip-lock bag and an alligator clip works just as well.
If you want to train a dog, you need a dog. Duh. Any mutt will do. Some are smarter than others, but I haven't met a dog that couldn't learn the basics. I personally prefer working dogs, or poodle crosses.
Like any field, you will find a wide range of contrasting ideas. The above are my views on what works and what doesn't. At the end of the day, it is your dog. If you want to use a $60 Dolce and Gabanna leash with a diamond studded collar...go right ahead. I prefer a 6 ft length of nylon rope, but then again I have a rather utilitarian outlook on life. If you want to fill your closet with a variety pointless dog accessories and accoutrements, go right ahead! Whatever makes you happy.
Step 2: Reinforcement Schedules
One click, one treat. Every time. When teaching a new behavior, always use a regular reinforcement schedule. This means that as soon as your dog is doing what you want him to do, you click and treat. This will encourage them to repeat the behavior. A human example of this would be a salesperson earning a commission on every sale they make.
Random Reinforcement (The slot machine principle)
Clicking randomly earns a treat. This is the principle that casino's use with one-arm bandit slot machines. The reward is sufficiently frequent that it makes it worth your investment in time and money, but not so frequent that the Casino loses money. An example of inadvertent random reinforcement , in terms of dogs, is begging at the table. Your dog begs, and you give him some food. The next day, you are annoyed with his begging and you send him away. A couple days later, your dog is begging again, and you give him a treat. Now your dog begs every day, all the time. The reason? Sometimes the behavior pays off, and other times it does not. The positive reward for the operant behavior occurs frequently enough that the behavior continues. He might as well beg, because he might get a reward. Random reinforcement is THE most effective way to guarantee (proof) a behavior. Once your dog is responding to a cue reliably, you are going to switch to a random reinforcement schedule. This will help to solidify the behavior.
As an aside, a sneaky way to break begging is the following. Use your knowledge of operant conditioning to pull a fast one on fido. Give your dog a treat every single time your dog begs for a week or two straight. You are setting up a consistent reinforcement schedule, the stronger the association the better. After a week of reinforcement, STOP giving him treats cold turkey. Send him away onto his bed. All of a sudden the behavior no longer pays off, and the annoying habit goes away. This is called behavior extinction. I have done this with a couple of dogs, and it works a treat.
Step 3: Clicker Training, How It Works.
If you still adhere to various traditional dog training methods that use punishment as a motivator, DO NOT combine clicker training with punishment. Can you imagine how much you would screw your dog up if every time you clicked they expected a punishment to happen? You would have a nervous wreck on your hands. I have seen this happen from as little as one repetition. so be very very careful not to combine punishment and clicker training.
What is a clicker?
First off, what the heck is a Clicker?
A clicker is a small widget that makes a short, sharp clicking noise. Don't buy the expensive ones, with all the extra gadgets. Buy the cheapest one you can find, it works just the same as the deluxe versions. Your dog doesn't care what colour it is, or if it has a GPS locator built in. If you are really opposed to buying one, use the lid from a glass ice-tea bottle. Snapple brand works particularly well (although a bottle of snapple costs about the same as a cheap clicker...) At the cost of $ 0.25 each (see above link), you could buy 8 clickers for the price of a single bottle of iced tea. Include a bunch of clickers in your next amazon.com order. :)
Is the clicker magical?
There is nothing inherently magical about the sound of the clicker itself. It is convenient to carry around with you, it is lightweight, and it makes the same sound every single time. Some people use the word "yes" instead of a clicker, which works just as well. I personally prefer the clicker over saying "yes", because my tone of voice isn't as consistent as a mechanical clicker.
If the clicker isn't magical, how does it work?
In order for the clicker to be an effective training tool, you need to teach your dog what the click means. In technical training circles, this is known as "priming the clicker". Essentially, you are associating the sound of the clicker with a positive outcome, be it food, praise, play, walkies etc.
Marking and Cue words.
The idea behind clicker training is that you use the clicker to MARK behaviors that you want your dog to repeat. This is more effective than traditional positive reinforcement, because your dog associates the click with a positive outcome, you don't need to reward as soon as your dog exhibits the behavior, you only need to click.
With traditional training methods, you would generally introduce a cue word and then manipulate your dog's body into doing that behavior. With clicker training, you wait to introduce the cue word until your dog is performing the behavior reliably. Once your dog is performing the behavior reliably, you only click and reward when you provide the cue. The cue can be a hand signal, a word, or body language.
Won't my dog only listen to me if I have a clicker?
No. The clicker is used to mark or identify the behaviors that you want your dog to do. For each behavior, you will eventually phase the clicker out. You can carry a clicker around with you every where you go. Every time your dog does something cute, funny, weird etc. Click. Eventually, you'll be able to train them to do that behavior on command. You just have to catch your dog doing something right!
Step 4: Priming the Clicker
Priming the clicker is very easy, just follow these steps.
- Get a bag of treats, a favorite toy, or any other high-value reward (i.e real honest to goodness meat, none of that dry kibble crap). Keep the size of the treats small (one chew and they are done), dogs generally don't differentiate between a huge piece of steak and a tiny piece of steak. To a dog, steak is steak, and it is all good. If you want to emphasize a reaction, you prolong the reward. i.e several small treats in a row last longer than one big treat. I have a 60 pound dog, and my treat size is always the size of a pea or smaller.
- Click the clicker in the general vicinity of your dog, and then give them a reward.
- Repeat about 20 times in a session, do this 3-10 times a day. Keep the sessions short and sweet, they are more effective. The short session principle holds true for all training types. 5 sessions of 5 minutes is more effective than a 25 minute long marathon-session. The breaks between sessions allow your dog's brain to form the neural pathways associated with the learning, which results in quicker and more effective learning.
- You know your dog is "primed" when you click, and they come running, or they act like they expect something to happen. Don't move on until your dog REALLY understands that clicking results in something good.
When I am training my dog, I like to mix the types of rewards up, sometimes I click and stroke, sometimes I click and play a quick game, sometimes I click and give her a food treat. The important thing is, make sure something GOOD happens every single time you click that little box. I like to mix the rewards up, because it keeps it fresh and prevents your dog from getting bored of the same thing. During priming though, make sure you use HIGH VALUE treats. I found all-beef, no preservative hot-dogs to be particularly effective.
Step 5: Tricks or Obedience
Something to think about.
If you want to be good at teaching your dog tricks, you first have to be DAMN good at teaching them obedience. Obedience work is where you learn to communicate with each other. The basic sits, stays, downs, are all important foundation skills to any trick. If you can't communicate the basics to your dog, how can you ever expect him to learn a complicated multi-stage trick like jumping onto your back or into your arms?
Step 6: SIT!
- Wave a treat in front of your dog to get them interested.
- Slowly move the treat above and behind their head.
- As soon as their butt moves towards the ground, CLICK!