free offline website design software

METHODS OF TRAINING


Shaping With Markers

Marker training has been one of the most effective of any method I have seen in my 40 plus years of training dogs. Marker training is mostly a black and white method of indicating what is the right movements that make up a behavior.


Marker training provides a method that allows a human to communicate with his dog at the exact moment his dog does something he likes. It provides a non-punishment based method of telling a dog that you like what he is currently doing and you want him to continue to do exactly what he is doing at this moment in time.


The process of marker training looks something like this:

1) Get a dog

2) Decide what you are going to use to "mark" behaviors. This can be a word, a sound you make, or a sound you create with a tool like a 'clicker'.

3) Decide what behavior you wish to teach

4) Picture in your mind, or write down, the different movements that produce the final behavior that you want your dog to do

5) Watch your dog. As soon as he makes one of those movements, "mark" that movement and reward the dog.

6) Keep watching your dog. He should make the same movement, if he does, mark and reward.

7) Repeat #6 several times and then wait a little longer to see if the dog can add to the movement the next logical move toward the behavior you have decided upon.

8) Mark and reward all such movements in sequence that will finally produce the wanted behavior


For example The first steps of Take It

1.Final Behavior: Take the toy from the human and hold it

2.Closest behavior the dog already does: Looks at toy.

3.Reinforcer for each approximation that meets the criterion: freeze dried liver.

4.Progressive approximations – Mark each time the dog attempts any step, raising criteria as proficiency is gained. 1.Look at toy

2.Move toward toy

3.Touch nose to toy

4.Grab toy with teeth

5.Grab toy with teeth and pull away from human

6.Hold toy for some duration without chewing

7.Repeat previous approximation for longer durations

Lure Reward Training

(lo-or)
n.
 1.
 a. Something that tempts or attracts with the promise of pleasure or reward: the lure of the open road.

 b. An attraction or appeal: Living on the ocean has a lure for many retirees.

 2. A decoy used in catching animals, especially an artificial bait used in catching fish.

 3. A bunch of feathers attached to a long cord, used in falconry to recall the hawk.

tr.v. lured, lur·ing, lures

1.To attract or entice, especially by wiles or temptation:

According to Ian Dunbar:

"All you need to train your puppy is the inclination, a few sparks in your brain, a couple of pieces of kibble in your hand, and...the puppy. So, enough said — let's get going. Ask your pup whether it is ready to proceed by moving a food lure up and down in front of his nose. If your puppy nods in agreement, you're off and running."


Mr Dunbar uses kibble and that's all fine and you can usually get a puppy or a really motivated dog to work for kibble, but my dogs are spoiled by being fed raw and think kibble is on par with dirt. So I use "treats" generally Natural Balance food rolls cut up into little bits. Natural Balance is dog food but I'm assuming tastes a lot better then kibble as I haven't found very many dogs that dislike it.

 Lures

When using this method of training, there are a few rules


A lure is presented before a desired behavior. It is an enticement to either follow the lure to get to the desired position or for the dog to do a previously learned behavior that isn't on stimulus control yet. 


A lure helps teach the behavior and the meaning of the cues that signal that behavior. Because of this, the cues come first, before the behavior is learned unlike the marker methods. This is based on the fact that the dog has no clue what we want him to do or what the sounds or body movements that come from us mean. The lure shows the dog the path he is to take to do the behavior. The path helps the dog choose the right movements. This works whether the dog is willing to offer behavior or not. Over the top enthusiasm or offering 20 behaviors in 3 seconds doesn't affect the response to the lure. So the dog learns what is being asked for and that all the other movements aren't needed right now.


Just like with marker methods of training, luring can be used to teach progressive approximations to a complicated behavior like "around" as well as a simple one movement behavior like "sit". With luring you can break a behavior down into simple movements, lure each movement, rewarding at the completion of each part and continuing in a smooth progression to the finish. The "chain" of movements is learned fast and smoothly.

Do As I Do - Mimicry

When one animal copies the appearance, actions, or sounds of another animal, the first animal is called a MIMIC. You can think of it as a copycat! Usually, an animal will MIMIC another to avoid predators. If it can trick its enemy into thinking it is something less tasty or more dangerous, it will survive. Mimicry of action and appearance is quite common in the animal world. Generaly the "mimic" copies only the outword appearance of another animal which usually has toxic means of defence. This mimicry is used to prevent certain predators due to the association with the toxins.


Many animals also learn through mimicry, or as it is sometimes called "social learning". 


One of the most influential learning theories, the Social Learning Theory, was formulated by Albert Bandura. Social Learning has pieces of both operant and classical learning theories but is more in the realm of natural learning then contrived learning. With humans, Bandura posited that there were three ways that social learning occurred.

A Live Model, which includes an actual person performing a behavior.

A Verbal Instruction Model, which involves telling of details and descriptions of a behavior.

A Symbolic Model, which includes either a real or fictional character demonstrating the behavior via movies, books, television, radio, online media and other media sources.


Dogs of course, cannot read and although I've seen a few watch TV I seriously doubt they could learn from it. Dogs do learn their communication signals, their body language, through watching other dog do certain movements and actions under particular circumstance. These action mostly communicate an emotion, a warning, or an intention. Puppies try these signals out on each other and hone these skills just like a human baby learns to walk and talk by watching mom and dad and possibly other siblings. Studies are now showing that animals can learn from other animals, even animals not of their own species, through imitaion. In some cases, with complex behaviors, learning by imitation may be quicker and more efficient then marker based or other reward based training methods.

Social learning has recently made a splash in the dog training world with a book by Claudia Fugazza Do as I do, a new training method based on social learning. Claudia, as far as I know, has been researching this method for a few years (Fugazza & Miklósi, 2014) and has discovered that teaching by imitation can be faster and more reliable then other methods. Her book outlines a method of teaching using imitation or what is called social learning.


The method described in Fugazza's book and DVD set is as follows

1.Demonstrate the new behavior.

2.Give the new cue

3.Give the cue 'do-it'

4.The dog should attempt to perform the new behavior that you have demonstrated

5.Reward!!

According to Fugazza, you can eventually dispense with the "Do It" and the demonstration and just use the cue. I've found that in most cases, cases where the dog is not incredibly anxious or fearful, dogs learn fast. I've had to add the cue very early on in the repititions and the "Do It" becomes confusing. Once the dog understands the "do it" cue however, mimicing what you did in demonstration is pretty solid.

SHAPING WITH GAMES

A game is a challenge, created by the rules that govern it, bound by the cooperation between the players of the game who all have the same purposes, intentions and focus; all of which results in a quantifiable goal.


The Goal of Shaping


Shaping is about going through the steps of an action and reinforcing each step until the final action is achieved. Shaping with markersworks best with those fine motor skills and intricate movements that conprise a simple behavior. Shaping with games can increase the amount of behavior that can be learned in short period of time. 


Timing, criteria and reinforcement are critical in marker shaping


Timing: Are you marking at the exact moment of the action or non-action? One of the nice things about marker training is that you don't "ruin" a dog if you make a mistake. But to actually teach a behavior, the timing of the mark must be nearly perfect. This is one of the reasons clickers are thought to be best. It's a sharp, short, unique sound that can be (according to clicker trainers) done faster then any other form of mark. Personally, I'm all thumbs and never did get the timing right. My thumb would press too fast and hurt me, too slow and not make the sound, off center and the clicker went flying. But there are many great clicker trainers out there who either have better manual dexterity then I do or they took the time to get it right. Timing a "yes" may (and that's debatable) be slower, but for me it works. I've trained things as intricate as an ear flick using only "yes" as the marker.


Criteria: This is where many trainers seem to not understand. You MUST know what the final behavior should look like and the actual actions that build that final behavior. Planning is one of the most important steps in shaping. It doesn't matter whether you are shaping with markers, lures, targets or games. Break the behavior down, understand each step. Then when it comes time to increase criteria as you are working with the dog, you know what comes next. Without the plan, your dog with be stuck in a guessing game as you try and determine whether the action he made with lead to the final behavior or not, which often causes you to miss marking a specific movement until you figure it out. 


Reinforcement: This is another area, especially with beginning marker trainers, that needs work. Reinforcement should be fast and abundant; especially in the beginning stages of teaching a new behavior. Reinforcement should be appropriate. If you are teaching an ear flick, you don't want hard to chew large pieces as you will be reinforcing tiny movement which would get lost in the effort of chewing. But if you are teaching "go around", treats could be larger and more rewarding for the larger movements of the dog.






Timing, criteria and reinforcement while shaping with games.


Timing: When shaping with Games, timing is not that critical. Reinforcement generally happens at the end of the game, occassionally inside the game if there are a lot of steps to the game, and it's always what would be considered a jackpot in marker training. The entire sequence of actions is reinforced, not a series of progressive approximations. As you play the game, the movements should flow one from the other for both the human and the dog. Flow is what is critical in shaping with games, not timing.


Criteria: Criteria are planned out. Just like marker training you must know what the steps are to creating the final behavior. The difference with shaping with games is that the criteria become separate games. For instance in training a "go around". The first game would be runing to the object and running back. The second game could be putting the reinforcement on the ground next to you and by using game one, the dog would run up to the object, run back and be rewarded with what's on the ground. By manipulating the environment, walking the dog through the steps of the game and rewarding each step, the dog gains understanding from the beginning. There is no guessing game.


Reinforcement: Reinforcement is large when shaping with games. You want the dog to learn that the action just performed is very valuable and could serve on it's own and not just a bridge to a final behavior. The placement of reinforcement when shaping with games can be vital. You are asking your dog to play a game and showing him the "prize" that can be won if he plays the game to a win. Rate of reinforcement isn't as crucial here. If you have broken down a complete behavior into simple parts and created a game for each part, reinforcement automatically happens very fast. One of the biggest differences between shaping with markers and shaping with games is that with games there is no need for variable or differential reinforcement schedules to create motivation. The game itself takes on the value of the reinforcement and because it's own secondary reinforcer. This can be seen when you teach a dog to get "up" on objects. After less then a week of "up" on various objects, your dog will offere this behavior everywhere. It becomes something that is so reinforcing, it can even be the reinforcement for other games.


How Shaping with Games Differs


End Goal: First identify the goal of your end behavior. What are you trying to train, what does the end behavior look like in your mind's eye? What is the criteria; how fast/how high/how many repetitions/how much or how little movement?


Stationary or Motion: Distinguish if you are shaping a stationary behavior or behavior of motion. The answer to the question will determine your reinforcement strategy. For example if you are shaping a stationary behavior your goal will be to build duration for that response. Rewarding the same response but working towards extending the interval of the dog maintaining that response (a sit or down stay or holding a sit pretty). By way of contrast, if you were to consistently reward the same moment of time for a behavior of motion (always rewarding the exact same number of steps away or number of rotations around), you run the risk of turning that behavior of motion into a stationary response (the dog will predict always getting rewarded at the same place and will want to stop when he reaches that place). 


Know What You Don't Want: Almost as important as having a vision of what you would like your dog to do is clearly knowing what you DO NOT want your dog to do. There are some responses that can be toxic when trying to shape your dog. That is, if these responses were inadvertently rewarded they would make it difficult for your dog to do what you are really after. For example, imagine you are trying to shape your dog to back up away from you. You see his paws move back so you mark the response, but at the same time your dog sits. If you feed the sit, you make it difficult to get the back up; how can a dog back up when he is sitting? By knowing ahead of time what responses you do not want your dog to offer, you will be prepared to not mark or reward the dog should your dog offer one of those responses.

Map it Out: All good training sessions start with a written plan. It is plan of action of how to get your end or "goal" behavior. Bob Bailey suggests to us all that we "Be a splitter, not a lumper." By that, he is encouraging us to chunk the vision of our larger behavior down into smaller responses that we can build separately and eventually put together to give us our targeted response. This sort of planning gives me the flexibility to move ideas around and be flexible with the elements of my plan as they move from my mind's eye to the computer planning page to the dog training session. 


Manipulate Your Environment: Your first step towards you actually training the dog should be to manipulate the environment in order to set the dog up for success. Environmental manipulation reduces the number of failures your dog has, increases the success of your session as it decreases the options your dog has while working with you. In my opinion there is nothing else that is unrelated to the dog himself, that has more impact on your dog training than environmental manipulation. You can manipulate your dog training environment by:


Space: Limiting the space your dog has to work within. The larger the environment you train in, the more potential distractions to your training and the more options your dog has available to him. Keeping the environment small and sterile (without exciting distractions) helps to keep you as the focus and your training session as the focus of the dog's attention!


HOW TO PLAY THESE GAMES

There are very few trainers that teach using games. There are rules to this type of training that you should be aware of.
1.Do not look for perfection, you are playing a game, eventually your dog and you will gain proficiency, but for now, just play the game as instructed.
2.Do not play for more than 5 minutes at a time. You can do 3 or 4 sessions of 5 minutes each in any one day, but no more than 5 minutes. This way of training is exhausting, demanding and powerful. 5 minutes is all you need.
3.Play the games in order. Do not skip around. Each game builds on the ones before it. If you skip around not only will your dog not learn the lesson, but she will get confused and you will get frustrated. These games each teach only a piece of a final behavior in obedience. If you skip around, the dog doesn't figure out what that final behavior should be.
4.Do not over think the instructions. If you wonder if you should put your dog on a leash, read the instructions. I will tell you when the leash is needed and when it's not.
5.This is not mindless obedience, we are teaching behavior not how to sit when told. You should be building self-control in your dog because you will not always be there when she might encounter a snake or other dangerous animal or situation. A dog that has self-control will be willing to live by your rules and you will have built in an understanding of the rules by playing these games.
6.Even in your 5 minute sessions, take breaks to play with you dog, just play. A little tug after your dog does the steps of a game goes a long way to building your dog's self-confidence and motivation to play again and again.
7.Play only one game per day.
8.When a game calls for you to push your dog back so you can run away, you need to watch your dog's reaction to being pushed. Start with barely pushing your dog a quarter of an inch and watch the response. If your dog responds well, try a little harder. Shy dogs are not good with the push, so you may want to just run with no push. For dogs that get too excited, keep the push very light or delete the pushing.
9.Do not correct your dog for what you consider bad behavior or the wrong choice. Dogs do what gets them rewards and what avoids pain and discomfort. If in the teaching process, before your dog fully understands the game, you correct him for a wrong choice, what are you teaching him? You are teaching him to avoid this game.