Ever hear people say how much their dog "hates" to be groomed, have its nails clipped, or wear a muzzle? I hear this all the time. It's quite rare to hear anyone say "boy does my dog love the hair dryer and nail clippers, that just makes his day". However, this statement can be the norm instead of the exception with some work.
Enter classical conditioning. Classical conditioning is similar to the idea of clicker training but with some very marked differences. In clicker training (or operant conditioning) the clicker indicates to the dog that reward (food) is coming upon hearing the sound. You then use the clicker when the dog is performing a behavior in order to teach the dog a behavior is wanted and will be rewarded. This behavior then increases and can be put on cue.
However using Classical conditioning in this sense we pair food/reward with some "thing" in order to build a positive association with the "thing" in the dogs mind. He isn't required to perform a behavior and there is not a defined relationship with the "thing" and with rewards as there is with the clicker and rewards (one click = one treat). The "thing" could be a muzzle, gentle leader, nail clipper, brush, or hair dryer.
The general idea is to replace a neutral or negative feeling in the dogs mind in relation to the 'thing' by pairing it with something pleasurable like food. (If the dog already has a negative opinion of the "thing" in his mind we call this counter-conditioning). This is not necessarily a conscious thought in which the dog learns that nail clippers = treats, but rather to give the dog a general 'good' feeling of the items presence. This also must be done very slowly with slight increases in duration and ferocity each time. I'll lay out a standard hierarchy of how to do this at the end of the post.
The majority of owners however (and certain TV personalities) use whats known as flooding. Rather than pair the item, say a hair dryer, with food and slow gradual increases in duration/ferocity, they simply lay into the dog by drying him all over without any build up or pairing with positive associations like food. Normally this is done by restraining the dog, possibly also with a muzzle, to ensure he accepts whats being done to him. While this accomplishes the task of actually drying the dog, there is huge psychological damage. While repeated use of this will likely show less signs of the initial panic, your relationship with your dog will be damaged and trust will be greatly reduced. He will also never really 'like' whats being done.
However by using classical conditioning and slow desensitization we can get dogs to actually accept things willingly and without restraint. Its also important to note that the earlier you do these the easier it will be. My puppy classes spend some time focusing on this important and overlooked topic. Here's a hierarchy of classical conditioning and desensitization you can do with your dog to get him to enjoy nail clippers :
1) Put the nail clippers behind your back, show them to your dog, then treat and put the clippers behind your back. Repeat this 10 or so times.
2) Continue the above but this time increase the duration for the visibility of the clippers. Continue to treat whenever they are shown.
3) Change locations in your house or go outside and continue the above. Also change where your treats are coming from (ie if you are wearing a sack, change sides, or change pockets). Its important to teach the dog that the clippers mean treats, and not location or where your hand is going.
4) Touch the clippers to the dogs front paw and remove then treat. Continue this several times. Do not actually clip or put the clipper over the nail, just touch it to the paw. Gradually increase the duration of the clipper touching the paw. You may want to get your dog into a 'down' position for this exercise. Do this exercise for EACH for your dogs 4 paws. Do several paw touches with just your hands and without treating but with praise. This teaches the dog that paw touching = good, but touching with clippers = awesome!
5) Put the clippers in the cutting position over the nail, remove and treat. Continue this several times. Do not actually clip the nail. Increase the duration and do reps with each nail and each paw.
6) Put the clippers in the cutting position and slowly cut one nail. Treat the dog then remove the clippers. Continue this on each nail. The relationship for this should be each cut = one treat.
The key to this program is that if at any point your dog does not appear comfortable or shows a negative reaction - go back to the previous step and do more reps. Your dog should show you a clear message of satisfaction during the procedure and upon the appearance of the clippers. The dog should look like he's expecting a reward when seeing the clippers. We refer to this as a conditioned emotional response. If your dog hasn't learned that clippers = good times, you need to go back a step.
As you do this more and more, just like clicker training, you reduce the amount of rewards. Do NOT progress with this too quickly, it must be a gradual reduction. For example go to 2 nail clippings per one treat, then 3 clippings per one treat, etc.
Think of an employee getting a pay cut. You might accept a 1% cut each year and still stay in your job, but if suddenly the pay cheque went down 75% one pay, you'd be gone!
Slowly move this towards one big reward after all nails have been clipped, and you will have a dog that is excited at the prospect of nail clipping, rather than running in fear.
My dog use to be fearful of the hair dryer, now when she's wet she waits in the living room FOR the hair dryer and will actually demand (paw, stare) that I dry her with it. It's now HER idea! This is real training and I'm more than happy to follow her direction!
For an excellent video example of how this works, please visit the following link:
http://www.youtube.com/watch?v=1wakterNyUg
This shows Jean Donaldson, an excellent trainer and author - doing this same process for a gentle leader on her dog buffy.
Friday, August 27, 2010
Monday, August 23, 2010
Living for the Moment
As Humans we sure have a lot going on in our lives. Work, Friends, Family all consume our time and thoughts. We are always looking in the past thinking about what could have been in our lives, analyzing past conversations or maybe just thinking about a movie we saw last night. We also worry about the future and how to pay our bills, if our job is always going to be there, or for the health of ourselves and family.
So how do Dogs view the passage of time and do they ponder the past and future? I never admit to knowing what a dog is thinking, sadly we will never know for sure. But we do know certain laws of learning when it concerns animals and dogs.
Dogs will continue to do what has worked for them in the past. If your dog is scared of strangers and growling at them makes them retreat - your dog will continue to growl. He's knows that in the past, growling worked. He doesn't ponder the morality of growling or how humans view this, but simply that it worked in the past to get rid of a scary person. Likewise if sitting down always gets treats and praise, he's likely to do a lot of sitting around you, specially if the same conditions for being rewarded exist again. You dog will remember that before he got a treat, your hand went into that cookie jar! One of the keys to obedience is teaching dogs that they will be rewarded in many different locations, with a variety of preemptive cues (IE, cookies don't ALWAYS come out of a jar or a certain pocket). This is also how we get a dog to perform without treats! We need him to want to gamble that he never knows for sure when he might get a reward. How many of you play the lottery and haven't won yet? Still playing?
However when it comes to teaching and rewarding behavior - dogs live very much in the moment. The most common example of how humans get this wrong is the classic housebreaking example. Man comes home after work to find that Fido has pooped all over the brand new rug and the owner proceeds to tell Fido what a 'bad dog' he is. The owner then brags that his dog must have known what the punishment was for because he 'looked guilty' soon as he pointed out the poop on the rug.
What really happened here was the Dog got worried once the Owners disposition turned sour upon seeing the poop. Perhaps that dog has seen that look and behavior just prior to something bad happening for the dog (another cue like the cookie jar). He has no idea that the physical action of pooping is what is upsetting the owner.
This example of course deals with a behavior that has happened hours in the past, but what about behaviors only seconds apart? Take this example:
Two dogs pass each other on leash. One dog lunges at the other in an aggressive motion before being pulled away by the owner. The quiet dog's owner hurries him away from the lunger while the aggressing dog's owner apologizes and hurries her dog past.
Once separated, the owner proceeds to verbally punish the dog saying what a bad dog he is for lunging at that poor dog. I've seen owners hold their dogs face and give them a lecture or a stern "NO" after such an incident.
This much like our housebreaking example - also doesn't work. Once the encounter is over, the dog has moved on and is thinking about the current moment, while the human owner is still lamenting on the incident. Any reference to that incident for the dog, is now gone. Punishing (or rewarding) for the behavior of lunging will have no effect. Sadly I see this same incident time and time again in parks and on walks.
One of the reason a clicker is so effective is because it signals to the dog at the EXACT moment a behavior is happening. With something as simple as a sit, if you click the dog too early in the sit, you'll teach your dog that 'sit' means to hover his butt over the ground. The reason clicker training was invented was because it was sometimes impossible to reward the animal the moment the behavior was happening if the animal wasn't in reach. Clicker training started with dolphins. By the time mr dolphin swam over to get its fishy reward from the trainer, it didn't know that it was getting the fish because of that 10 foot jump it just performed. A sound like a whistle or a clicker bridged that time gap.
So while dogs may not ponder the past and future as we do, they certainly remember how things worked out for them before, and apply that in the future. The good news is we can use this knowledge in teaching and training. So keep rewarding those behaviors that you DO like, and you'll see a lot more of them!
So how do Dogs view the passage of time and do they ponder the past and future? I never admit to knowing what a dog is thinking, sadly we will never know for sure. But we do know certain laws of learning when it concerns animals and dogs.
Dogs will continue to do what has worked for them in the past. If your dog is scared of strangers and growling at them makes them retreat - your dog will continue to growl. He's knows that in the past, growling worked. He doesn't ponder the morality of growling or how humans view this, but simply that it worked in the past to get rid of a scary person. Likewise if sitting down always gets treats and praise, he's likely to do a lot of sitting around you, specially if the same conditions for being rewarded exist again. You dog will remember that before he got a treat, your hand went into that cookie jar! One of the keys to obedience is teaching dogs that they will be rewarded in many different locations, with a variety of preemptive cues (IE, cookies don't ALWAYS come out of a jar or a certain pocket). This is also how we get a dog to perform without treats! We need him to want to gamble that he never knows for sure when he might get a reward. How many of you play the lottery and haven't won yet? Still playing?
However when it comes to teaching and rewarding behavior - dogs live very much in the moment. The most common example of how humans get this wrong is the classic housebreaking example. Man comes home after work to find that Fido has pooped all over the brand new rug and the owner proceeds to tell Fido what a 'bad dog' he is. The owner then brags that his dog must have known what the punishment was for because he 'looked guilty' soon as he pointed out the poop on the rug.
What really happened here was the Dog got worried once the Owners disposition turned sour upon seeing the poop. Perhaps that dog has seen that look and behavior just prior to something bad happening for the dog (another cue like the cookie jar). He has no idea that the physical action of pooping is what is upsetting the owner.
This example of course deals with a behavior that has happened hours in the past, but what about behaviors only seconds apart? Take this example:
Two dogs pass each other on leash. One dog lunges at the other in an aggressive motion before being pulled away by the owner. The quiet dog's owner hurries him away from the lunger while the aggressing dog's owner apologizes and hurries her dog past.
Once separated, the owner proceeds to verbally punish the dog saying what a bad dog he is for lunging at that poor dog. I've seen owners hold their dogs face and give them a lecture or a stern "NO" after such an incident.
This much like our housebreaking example - also doesn't work. Once the encounter is over, the dog has moved on and is thinking about the current moment, while the human owner is still lamenting on the incident. Any reference to that incident for the dog, is now gone. Punishing (or rewarding) for the behavior of lunging will have no effect. Sadly I see this same incident time and time again in parks and on walks.
One of the reason a clicker is so effective is because it signals to the dog at the EXACT moment a behavior is happening. With something as simple as a sit, if you click the dog too early in the sit, you'll teach your dog that 'sit' means to hover his butt over the ground. The reason clicker training was invented was because it was sometimes impossible to reward the animal the moment the behavior was happening if the animal wasn't in reach. Clicker training started with dolphins. By the time mr dolphin swam over to get its fishy reward from the trainer, it didn't know that it was getting the fish because of that 10 foot jump it just performed. A sound like a whistle or a clicker bridged that time gap.
So while dogs may not ponder the past and future as we do, they certainly remember how things worked out for them before, and apply that in the future. The good news is we can use this knowledge in teaching and training. So keep rewarding those behaviors that you DO like, and you'll see a lot more of them!
Wednesday, August 18, 2010
Welcome to Golden Rule Training
Thanks for coming to visit my blog and hopefully my website (www.goldenruletraining.ca)
My name is Tristan and I train dogs in the Halifax area, providing group classes and individual consultations. I'd like to use this blog to discuss recent happenings in the doggie world and also training issues with my personal views on things.
Dog training sadly is one of those fields where you can get 5 different answers on the same subject from 5 different trainers. Imagine taking your car to a mechanic and having one tell you to change the oil, one telling you that you don't need oil, and one telling you to fill it with beer instead. People sure wouldn't stand for that, however that's a regular occurrence in the dog training world. One of the biggest reason for this I believe is there is no set standardization for dog trainers. While many people are trying to do this through sites like www.apdt.com and www.ccdpt.org, the most popular dog trainers on TV have no such certifications or training. Dog training seems to fall on different ends of a straight line, with positive trainers (those who use clickers, food lures) on one end, and dominance force based training on the other end where all methods focus on being the 'alpha' of the group or 'knocking the dog down a peg'. There are many trainers that fall in between, such as those that will use food for rewards, but also force if faced with non-compliance to commands.
I personally fall very high towards to the positive end of that line. I use clickers in my training but not as religiously as some use clickers, I use food lures and reward training and I don't jerk leashes or inflict pain. I'm against shock collars and invisible fences, since training can accomplish those goals without shocks and pain. If you'd like to learn more about my resume please visit my site. I believe there are two important things a person must do to be a good trainer. 1) Read as much as possible by respected authors. I have a large library from trainers like Ian Dunbar and Jean Donaldson. 2) Practical experience - volunteering at shelters and just working with dogs and specially your own dog is the best way to hone your 'chops' of training.
Thats it for the first post - if you'd like me to discuss a certain topic send me an email. In the next post I'll discuss a bit about learning theory and exactly how animals learn.
My name is Tristan and I train dogs in the Halifax area, providing group classes and individual consultations. I'd like to use this blog to discuss recent happenings in the doggie world and also training issues with my personal views on things.
Dog training sadly is one of those fields where you can get 5 different answers on the same subject from 5 different trainers. Imagine taking your car to a mechanic and having one tell you to change the oil, one telling you that you don't need oil, and one telling you to fill it with beer instead. People sure wouldn't stand for that, however that's a regular occurrence in the dog training world. One of the biggest reason for this I believe is there is no set standardization for dog trainers. While many people are trying to do this through sites like www.apdt.com and www.ccdpt.org, the most popular dog trainers on TV have no such certifications or training. Dog training seems to fall on different ends of a straight line, with positive trainers (those who use clickers, food lures) on one end, and dominance force based training on the other end where all methods focus on being the 'alpha' of the group or 'knocking the dog down a peg'. There are many trainers that fall in between, such as those that will use food for rewards, but also force if faced with non-compliance to commands.
I personally fall very high towards to the positive end of that line. I use clickers in my training but not as religiously as some use clickers, I use food lures and reward training and I don't jerk leashes or inflict pain. I'm against shock collars and invisible fences, since training can accomplish those goals without shocks and pain. If you'd like to learn more about my resume please visit my site. I believe there are two important things a person must do to be a good trainer. 1) Read as much as possible by respected authors. I have a large library from trainers like Ian Dunbar and Jean Donaldson. 2) Practical experience - volunteering at shelters and just working with dogs and specially your own dog is the best way to hone your 'chops' of training.
Thats it for the first post - if you'd like me to discuss a certain topic send me an email. In the next post I'll discuss a bit about learning theory and exactly how animals learn.
Subscribe to:
Posts (Atom)