Click to Guide
Click to Guide. 1
What follows is a compendium of posts that were sent to an Internet mailing list that was established in May of 2007. There is no identifying information about who sent the post etc. Some of the content of the postings has been edited, not to change its meaning at all, but to keep things as clear and concise as possible.
Here’s a quick note from your Click to Guide crew that explains our list ordinances! Please save this message so you can refer back to it if needed.
This is a list of and for Seeing Eye graduates but it is not sponsored by The Seeing Eye.
Your CTG crew is Sue Martin, Bill Engleking and Ginger Kutsch. . The three of us are Seeing Eye graduates.
Several people have asked what clicker training is so here’s a brief explanation: Hmm, this is difficult to boil down to a couple of paragraphs. Okay, clicker training is operant conditioning. Basically, you pair the sound of a click, which is made by any one of several gadgets out there, with a yummy treat. The yummy treat is the primary reinforcement and the sound of the click becomes a secondary reinforcement. In other words, the click takes on the same wonderful connotation as the treat itself. The click is used as a very precise marker of behavior. It’s way more precise than verbal praise. For example, If I’m trying to teach my dog to show me an outside trash can and the dog actually does this, by the time I finish saying, “Good girl,” she might be already doing some other behavior that’s not so desirable…like sniffing or snatching something off of the ground near the trash can. So, with this very precise marker I can communicate to my dog exactly what it is that she does that I think is good. I then follow the click with a treat. The good thing is that the treat doesn’t have to be given at the exact moment that the dog does the good thing. The click has already marked the good thing and the treat can be given a few seconds later when it’s more convenient. The clicker can be used to gradually shape just about any behavior you could imagine. However, there’s absolutely no coercion involved. The dog must offer the behavior by itself. That’s why it’s called operant conditioning.
I am the machine that my dog tries to operate. I’m a click and treat machine and my dog is trying to figure out what behavior she can do in order to get the click and treat machine to operate. It’s simply another means of teaching my dog to do stuff. What makes it great is that it’s a really positive experience for both of us. Kismet loves it when I get out the clicker and we go in our bedroom. That’s where I’ve taught her all of the new commands that she’s learned. She knows the game is on and she’ll throw every behavior I’ve ever taught her at me trying to get me to click and treat.
Clicker training is not an expensive game to play. You can get a clicker for a couple of bucks at many pet stores. You can also get fancy more expensive clickers if that floats your boat. Bottom line is that anybody can do this and you don’t need to sink much of anything except time into it.
Question: This post brings up the question that no one can answer to my satisfaction. If you are going to click then praise then treat, why do you bother with the clicker? Praise and treat would seem to me to be sufficient.
Answer 1: From my personal experience, and as a moderator of another positive training list—from my observations of handlers on that list asking the same question, one point I’d like to make is that experience is really the key to the answer to your question. You can hear many of us talk about our own experiences, or read articles that explain why the clicker works better than a word, but until you try both a marker word and a clicker, no matter how much information you’re given, you really won’t be able to see the difference. That said, as to why a clicker and not a word? First I’ll use an excerpt from the post of Gary Wilkes article and then explain my own experience. “Why use a clicker as the “hot” signal instead of verbal praise? The clicker accurately identifies correct behavior. Because the clicker is faster than verbal praise, it is more precise. In the time it takes to say “good boy” an animal may perform the desired behavior and then move to an unwanted response, before the praise has time to register. In this scenario, the animal can’t tell if the trainers liked the “sit” or the “jumping up on the trainer” that occurred a split second later. The clicker can also work well from a distance. It is impractical to try to toss a treat at an animal’s mouth at the exact moment that a desirable behavior occurs. The clicker bridges the gap from the instant the animal performs the correct response and the time it takes to actually deliver a treat — you can click in a timely fashion and then deliver the treat as much as three or four seconds later. With clicker training, the timing of the click is far more important (and practical) than the timing of the treat.” I first started with positive training using a “marker word,” which in my case was “treat.” I learned the concepts above by TSE instructors. I thought, “Hmm this seems to work OK, and why do I have to use something else if a word works just fine to get the behavior I want?” Then I attended a Clicker Expo put on by Karen Pryor (the guru of clicker training.) One of the sessions I attended was a “learning Lab,” where members of the Clicker Expo faculty gave attendees exercises to practice, similar to the ones Sue and Ginger have created for this list. We were given clickers and asked to get into teams of two, and in case it was three. We were given the instruction to have one person click when a signal was given and the other person was asked to observe the timing of the person who was doing the clicking and give them feedback. Since I couldn’t see the person who was giving the signal to click, I had that person touch my arm and then clicked. First, I was amazed at how confusing it was to get the right timing! By the end of the activity, I could see improvement in my timing, meaning that muscle memory was beginning to take hold. I took the clicker home and dropped the marker word and just started using the clicker to shape a behavior with my dog. She quickly began to learn what the clicker meant, (she’s a lab and in her mind treats easily became synonymous with the clicker). I started to see a difference in my own timing and in her reaction. I was truly amazed at how quickly my dog began to learn after I switched to the clicker. It was also exhilarating for me! All of the behaviors we are talking about so far on this list can be trained using traditional training. IMO there is no one right method of training, only techniques that work better in some situations with some dogs. I don’t always use the clicker with my dog, but I do always use it to teach her a new behavior.
Answer 2: Darn good question. The reason for the click instead of just the praise is that it’s way more precise. That’s why all of the practice. Your dog is going to learn that the sound of the click marks the exact moment that he or she did something that you like. If you simply praise when your dog does something you like he may be doing something else entirely by the time you get the words, “Good boy,” out of your mouth. You almost have to see this to believe it. Once Kismet was clicker wise, i.e. she knew that the sound of the click marked the exact moment that she did something good, it took me, literally, two minutes to teach her to shake hands. The very first time she lifted her right leg off of the floor I clicked. Then I praised and gave her her treat. I told her to sit again and she promptly threw that right paw up. It was incredible.
Question: I’ve been told before that you click then treat immediately, but no one can explain how you do that without moving the clicking hand, usually the right, toward a pocket or bait bag?
Answer: The ideal place for the clicker is either nestled in the palm of your hand or somehow attached to the harness so that your finger can rest comfortably on it just waiting for a moment to click. That way there is no movement that the dog can notice and this click comes out of nowhere. After the click you can move your hand to wherever the treats are and get one out. The click has already marked the instant that you are trying to capture and the treat can come along afterwards. The word, “Immediate,” is relative. While the treat should follow fairly soon after the click it doesn’t have to be so immediate that you fall all over yourself to deliver it. Think of the training of marine mammals. A whistle is usually substituted for a clicker. So, here’s this dolphin swimming around and he jumps in the air. The trainer blows the wistle to mark the event and then tosses the dolphin a fish. You can bet there are going to be several long seconds that pass before the trainer can reach down, pick up a fish, and toss it to the dolphin.
Whether we are visually impaired or totally blind, we all learn to read our dog through the leash and/or harness handle. We probably don’t think about this much, but if we are in tuned to our dog, we pick up on our dog’s movements and have a good idea of what the dog is doing even though we may not be able to “see” the dog.
The leash and harness tell us a lot. We can feel if the dog is turning its head. We can feel if the dog is changing pace. We can feel if the dog is lying down or up. We can feel the dog tense. We can feel if the dog is distracted and so on.
These are just a few of the things we “feel” and it is not necessary to see the dog to feel this.
What we *need to spend more time on is feeling our dog by touching our dog’s body. Touching our dog’s body directly is different because we are able to “see” firsthand what the dog is going to do.
Holding the leash/harness gives us an idea of what the dog has done after the dog has done it. Whereas touching the dog tells us what the dog is getting ready to do as well as what the dog is doing.
We usually notice the latter first, but with time and practice, we will be able to feel the dog gather the muscles that bring about the movement necessary to do what the dog is “about to do next”.
By touching our dog, we will learn to feel what “calm” muscles feel like and what “tense” muscles feel like. A dog that is about to do something will tense those muscles. That is how muscles work.
It is important that our dog get accustomed to being touched when we are training with a clicker. Yes, yes, yes…I know all the stuff you read about
clicker training says never to touch your dog…hogwash! There will be times when we will need to feel our dogs and that is okay. How else are we going to know if our dog is turning its head, etc.
While it is important for us to understand the “rules” of clicker training, it is equally as important for us to understand that we will have to adapt some training methods to suit our capabilities. We’ll talk about using sound later but for now, let’s just focus on learning more about our dogs by touch.
To start, position your dog within your reach so that you can rest your hand lightly on the area between the dog’s neck and shoulders. This is a great area for sensing the dog’s movements and one can learn to pick up even the subtlest of signals from the dog.
It is not a bad idea for those who may have some sight to practice such methods as well, for you will learn a lot about your dog when doing this. It will teach you to “watch” your dog in ways that your eyes will not see. Remember, dogs always give signals to what they are about to do!
What does all this have to do with using a clicker? If we are going to learn how to use a clicker effectively, we must first learn to be “proactive”. Proactive means to be prepared for whatever behavior may come from a dog.
By touching our dog with our hand, we will begin to understand what the dog is *going to do* rather than what the dog has done. Is everyone here with me?
If I’m not making sense, please ask for further explanation. So, here’s our first assignment. We have along weekend coming up so let’s focus on touching our dogs! When you’re hanging out at the family picnic, waiting for a ride, lounging in front of the television…or any chance you get…remember to rest a hand on your dog and study its movements. I think you will be surprised at what you will learn about your dog.
Then, it would be cool if we could all share our experiences with each other. We are pouring the foundation right now so please be patient. Remember, slow is fast in dog training!
Hi All, After reading the posts on getting started and preparing the Human”, I thought I would share some thoughts and emotions on these subjects with the list. When I first began researching the practical use of clicker training, I thought to myself, “Great Scott! What a terrific concept, but there is no way a blind person could respond quickly/appropriately to the dog’s subtle body language.” Then I began to imagine all sorts of disasters taking place not only with the training of a behavior, but also with the effective management of the tools required. Nervousness and fear tickled me because while I may have gone to college, juggling was not my major. It is important at the beginning to go as slowly as you need to so as to set yourself up for success before beginning the planning necessary to set the dog up for success. In my case, this meant learning how to communicate with my dogs effectively, before I attempted to juggle new equipment. It was also critical not to take myself so seriously and remember that this should be fun for all parties concerned — especially the dog! In the formal training process we were all encouraged to praise our dogs. Can you hear the instructors? “Mrs. Gillis, PRAISE that dog!!!” We were also asked time and time again if we had done our obedience exercises — sit, down, rest, come etc. Okay, now that we are all on the same page, let’s look at Ginger’s recommendation to touch our dogs in a quest to here what they have to tell us and Sue’s explanation of the clicker as a distinctive sound that marks a behavior. For me, getting started was as simple as touching more while engaging in what the dogs already knew how to do and praising in either a more exuberant fashion or by using a new distinctive verbal cue — “YES!” or “Great Job!”. This eliminated the frustration and awkwardness of having to hold something new in my hand and trying to concentrate on how and when to use the clicker as I built on my ability to know more about what my dog was going to do and when. I started by touching Ivy on the neck near the shoulder, asking for a behavior she already knew (sit,down) and as soon as I felt she was headed in the right direction with my hand, I praised that Dog! While working in this fashion, I found it easier to hold the leash in my right hand, while resting my left hand on my dog. I also used a flat collar rather than a training collar during the process to ensure that nothing even remotely feeling like a correction could occur. It was also helpful to attach a bell to the flat collar so I could hear certain movements. It seems to me that this approach is very safe because it buys you time in learning how to “see” your dog and since we have altered the usual exercise of obedience the dog views it as something new rather than a change to the routine. Remember this is supposed to be fun and as with anything new there is a learning curve, in this case for the human, and touching in this fashion was not part of the “formal” equation. I also mixed it up a great deal from the “formal obedience” as we performed in class — sit left, sit right, sit facing me, sit in front of me (butt to my toes) etc. Then, how many ways can we do a down? Getting your hands on your dog while at the same time asking them to do what they already know in a little easier and different from “normal”, it’s what makes the learning fun for both of you. However, don’t be surprised if your dog tilts it’s head and seems to say, “Have you lost your mind? WE never did this before.” When this happens, praise yourself because you are on the right track in the important steps of communicating as well as getting your dog’s undivided attention. The ho hum days are gone and now the dog knows to stay on its paws! I look forward to hearing the experiences you have over the weekend as we all strive to get in “touch” with our dogs. I also hope this helps to alleviate any human anxiety about getting off on the wrong foot. Remember go slow and make it fun!
Brushing kismet is a bit of a chore. She’s quite body sensitive and doesn’t much like being brushed. To keep her put while I brush I get her to lie flat on one side. I do that side and then flip her over to do the second side. When I was brushing her just now I kept my hand resting gently on her shoulder with my fingers on her neck. When she seemed to be resigning herself to the brushing her head would point in what would be an upward position if she were standing up. When she got tired and wanted to get up she’d tuck her chin towards her chest just before making a bid for freedom.
What’s interesting is that I think I knew what this slight change in her position meant but I’ve never exactly analyzed it.
I’ve also spent some time just sitting with Lukas, the puppy. He loves banging me in the face with his nose. I’ve noticed that he sort of tucks his head just before he makes this upward motion towards my face. When he’s flipped over on his back he sometimes stretches out his front legs so they’re pointing up and away from him. I have no idea what this means or what other action it might clue me into. It’s just something I’ve noticed.
I love shepherd ears. I was just sitting on the floor with Kismet. She was sitting in front of me looking towards the kitchen where Jim was on the phone. I placed my hand on her head, just behind her ears. At first she pinned her ears back and looked at me. When I proved to be boring she pricked up her ears again and looked back towards the kitchen.
I repeated this out on the front porch. There are lots more interesting things to look at out there and, I’m sure, lots more interesting smells.
This time Kismet was sitting beside me and we were facing in the same direction. When her attention was focused outward, her ears were erect.
When she attended to me she’d lower her ears backwards and then turn her head towards me. By keeping my hand on her head just behind her ears I could feel her ears begin to move backwards and anticipate her turning her head towards me.
I spent some time with Iggy outside, not just our usual outdoor walk and travels. I usually do some obedience outside when the weather is nice. I noticed that while he is sitting quietly, if he hears a sound, that big Labrador head turns slightly that way if only for a second. I also noticed that his nose is twitching and his ears move a lot. He doesn’t seem to pay attention to birds but turns his head to follow the movement of cars, etc.
Once I got Griffin settled on the bus this morning en route to work, I used my feet, which I keep in close contact with Griffin, to tell me what he was doing. I could not only tell when he shifted positions, I could detect when he was preparing for the position change. it was kind of neat to know what he was about to do before he actually did it. So, feet work, too.
Okay,we have touched upon “touch”, but there is also sound that can be used in training. Obviously, touch only works when the dog is close enough that we can reach out. There are times, however, when the dog may be outside our immediate touch and we need to add sound into the training equation. How can we use our hearing to listen for our dogs. You probably already do this but never really stopped to think about it much, right? Now we’re going to pay attention to every sound our dog makes and learn how these sounds can tell us more about our dog’s movements. Later, we will talk about how we can *add sounds to help us detect these movements. But for now, let’s share some of our experiences with sounds and what these sounds tell us. For instance, Payton’s nails make a clicking noise when she walks on the hardwood floor or tile at our house. This helps me figure out what she is doing and where she is going. I’ve also noticed that her nails stop clicking when she walks onto the area rug in the living room. Okay, so big deal…how does this help us with training? Well just to jump ahead for a quick moment, imagine that you are trying to train your dog to go to its place. You set up the dog’s mat on the kitchen floor. The mat is soft so when your dog steps onto the mat, the sound of clicking nails stop. This tells you that your dog is now on the matt, exactly where you want it to be! Yea. Many of us were taught to use the bell in class when we were teaching our dogs to go to their place and this may work as well. How can we use sound to tell what objects our dog is touching. What sounds help us distinguish if our dog is using its nose or paws to touch an object?
Payton will push the water dish around with her nose when it is empty. I can tell she is using her nose because she is not flipping the pan off the floor, merely sliding it around on the floor. If she was using her paw, I would hear the pan tipping up and crashing down, or maybe even flipping over. What other types of sounds can give us information about our dogs? Let’s all chime in and share some of our observations. This is a great chance for us to learn from each other.
thinking about your example of teaching the dog to go to her place, i can always tell when Kari has actually done so because when she lays down she lets out a huge groan. If she is just standing near her place, she doesn’t groan until she lays down. Both of my dogs have found ways to walk quietly so as to not let me know what they are doing so i have had to find other ways to figure things out. my retired dog has tags that jingle on her collar so i can tell which dog might be getting into mischief. Kari isn’t often quiet. she’s always snorting or groaning. She is a dog who enjoys touch very much and always wants to be right next to me.
It’s tough to come up with something new as obviously Ginger is secretly writing in the middle of the night — making a list and checking it twice just like Santa –to cover all the important bases in one clear and concise post. You go girl! An observation about Ivy is that she has learned to say “yes” and “no”. I did not teach this either she came up with it all on her own and I have to ask if anyone else’s dog does this. If I pick up the leash or harness by the front door, she gets out of her throne on the master bed and comes to the top of the stairs to look at me. Then if she wants to go she says “yes” by sort of sneezing and “no” by shaking her head. If Ivy gives me a “no” I have to convince her of the importance of the trip by promising a cookie if she comes with Mom. If the answer is “yes”, I can actually hear her dancing at the top of the stairs waiting for me to say, “Well, okay then let’s go!” Then she comes bounding down the stairs so fast I think she is going to run into the front door sometimes. Strange huh? She does this at bath time, car ride time and at church if I simply ask, “What do you say Ives?”. Another peculiar observation is the body language I get when putting her in harness. She comes, but just out of reach a tiny little bit. She will touch if I ask, but I really think she believes this a game I like to play. She also lowers her head for both the harness and leash. Once it goes on then she gets all serious and is almost snobbish about praise — no wags no eye contact, just a simple can’t you see I am working here air about her. If either of the dogs somehow hear me silently lift the clicker from its resting place, they both come running and begin battling for who can get closer to Mom as if to say “Pick me first please.”. Ivy will go through my legs when she is excited and wants to play the game and she has also been known to freeze under my legs and bark at Lily telling her that she is going first no matter what.
The other morning I was asleep and Payton was watching me like a hawk. When I awoke, I felt her chin resting on my leg…she probably actually woke me up…she sees nothing wrong with eating breakfast an hour or so early! Anyway, I kept my eyes closed and laid very still but I couldn’t help smiling. The instant I smiled her tail started thumping on the bed. This made me smile even wider and she responded by thumping her tail faster and harder. I stopped smiling and her tail stopped…smile and thump…no smile and no thump…it was so much fun! I marveled at how we were communicating. I felt like I was truly “seeing” my dog! Remember, eyesight is only part of seeing. To see is not necessarily to understand and many stop at seeing. They take it no further than that. They do not attempt to interpret the behavior or see if from the dog’s point of view. Even if we can never know for sure what a dog is thinking, we can definitely get an idea of what the dog thinks by the behaviors he/she exhibits. Isn’t it fun and interesting to discover movements from your dog that you haven’t noticed before? All part of getting to know your dog and truly seeing it. <grin>
There are several clickers out there. The cheapest and easiest to find is a simple plastic box. The one hooked to my belt loop is about an inch and a half long, an inch wide, and maybe half an inch tall. On one of the long sides there’s an opening with a metal strip filling it. One end of the metal is firmly attached to the box and the other end can be moved in a downward movement towards the inside of the box. When it is depressed it makes a distinctive clicking sound. Most of these little guys have a loop on one end through which you can run a string, rubber band, or anything else you like which can be used to attach the clicker either to yourself or to your harness. The box clickers cost a couple of dollars and I’ve found them at many pet stores.
This simple, box clicker is the only one I’ve ever used. It’s not elegant but it gets the job done. The first time I started with the clicker I duct taped it to my harness handle. This worked fine except that it wasn’t in reach if I needed to click a behavior that happened when the harness wasn’t actually in my hand. The duct tape also made a mess of my harness.
Next I tried running a rubber band through the little loop on the clicker and then putting the rubber band around my wrist. This was okay except that the rubber band was a tich uncomfortable. Nowadays I use a hair rubber band. It doesn’t pinch and it’s unobtrusive. I can have the clicker nestled in the palm of my hand when I need it and it stays around my wrist if I need to use my hand for something else. There are also spiral loops that you can get that look kind of like telephone cords. You can loop these around your wrist.
I know there are lots of more sophisticated clickers out there but since I’ve never used them maybe someone who has can chime in. These more sophisticated models have different volumes and even different click sounds to mark different behaviors. Bottom line is that you don’t have to spend much money to get started with clicker training but you can get more sophisticated if you want to.
Heads up all!
Before we get started, everyone is going to need a box clicker or i-click so if you do not have a clicker yet, you have a week or so to find one!
You will also need to decide what sort of attachment e.g. wrist coil, lanyard, duct tape, etc. you will use on your clicker. I’ve listed a few web sites below where clickers and other equipment can be purchased. You can often find box clickers in your local pet stores but I have yet to see i-Clicks for sale anywhere but on the internet.
If you plan to buy a box clicker, be sure that it has a tab on the end so you can attach a wrist coil or lanyard to it. All i-Clicks have tabs so no worry there.
If possible, you will also need a bait bag. This is a little pouch designed to hold treats. They can be clipped on to your belt or waist band or even your purse. Some bags are worn around your waist like a fanny pack. There are a variety of treat bags on the market so poke around a bit to see what style appeals to you. I have a Trek & Train bait bag from doggonegood.com that works well for me.
http://www.doggonegood.com/index.cfm?fuseaction=product.display&product_id=205 but there are lots of them out there so check it out! If money is tight, you can make due without an bait bag. Using the bags are easy and convenient but not totally essential.
Here are just a few sites that you can visit to explore the various types of clickers, bait bags, training treats, etc.
The sitstay.com web site is a “user friendly” site where you can read descriptions of the various clicker training equipment. It’s located at:
http://www.sitstay.com/store/equip/eq4.shtml cleanrun.com is another site where you can buy clicker training equipment.
http://www.cleanrun.com/index.cfm?fuseaction=category.display&category_ID=57 and clickertraining.com is another site for equipment located at:
http://clickertraining.com/store/?item=traininggear Anyone else out there have a favorite site they want to share? Or how about info on the equipment you may already be using? As we saw with Ruthie and Bonnie, a wrist coil may work well for some and not others! My choice of clicker attachments is the wrist coil. I don’t wear it around my wrist like a bracelet though. I wear it pulled down across the palm of my right hand with my thumb outside of the loop. This allows me to access my clicker instantly and yet gives me the freedom to still use my hand. The clicker dangles right along my finger when I am walking.
Okay, in this message, we’re going to discuss a few exercises that will develop your mechanical skills — mechanics of clicking, timing of clicking, and treat delivery. This will help to sharpen your timing. Then we’ll put it all together to begin teaching your dog! For those who are already using a clicker with your dog, you may still want to try these exercises out. The better we are at timing, the easier it will be for our dogs! For the first part, you will need a clicker and a place to go hide from your dog. Whether your dog already knows the sound of the click or has no clue what the noise means, you want to find a place where your dog will not be able to watch what you are doing. If you have an i-click, use it because it is quieter than the box clicker. If you have a regular box clicker, you can put a few strips of medical tape along the steel mechanism to soften the click sound.
Second, if you have a tape recorder, or something else that records, get that ready too. Okay, here’s a few exercises to get started. If anyone else knows or has an idea for other ways to practice timing, please let us know. The bathroom makes a good place to hide so let’s start there! Turn your recording device on and here goes!
- Listen to a radio, television, talking book or whatever and plan to click every time you hear a certain word. The word you select should be something likely to be repeated several times while you are listening. For example, words like “the” are commonly used and would offer lots of opportunities to click. Check to see how you are doing with your timing by listening to the recording.
- Recruit one of your family members or friends to help with this next exercise. Ask your partner to touch your hand, arm, finger, etc. Ask them to vary the timing so it will be harder for you to predict when you will be touched. . Click each time you feel a touch. Then grab a margarine lid or something similar and hold it in your hand. Now ask your partner to touch the lid, varying the pressure of the touch, the placement of the touch etc. Click each time you feel the touch. Your partner can help critique your timing. Just let them know that your goal is to click as soon as possible after they touch you. Once you have the timing of the click down, we will move on to practicing treat delivery! Good luck and let us all know how you are doing!
I did the first two timing exercises and the results were very interesting to me; not altogether surprising, but interesting to analyze afterward, at any rate. Oh, and quite depressing, but who’s counting?
First, I did the exercise where I dropped the noisy object and clicked on impact. I wasn’t too bad at that. Of about twenty-five drops, I anticipated three by a fraction of a second, was very late once, got very close most of the time-usually during the noise but not the instant it started. I counted four times where I got it dead on (for reference, the sound of clicker and, in this case, TSE bell hitting the floor simultaneously is quite distinctive and easy to tell from ones off by that fraction of a second. The notes blend perfectly when the timing is accurate.)
Incidentally, the recording was extremely useful in both of these Exercises, and I’d highly recommend recording the session. My perceptions about how close or how off my timing was were pretty much dead wrong except for that one way too late click and the four perfect ones. Later, with the second exercise, it was even more important, as that was when I completely missed instances of the word I was listening for and sometimes didn’t even catch them until I listened back to the recording.
Then I switched to the second exercise where I was listening to the radio and clicking for a word. As ginger suggested, I used the word “The.” This is where I got a lot of interesting results.
I missed the word entirely about a third of the times it was said, or realized it had been said far too late to click it.
I anticipated a *lot* and frequently clicked on the words “they” and “these”, when I only intended to click on “the.”
After noticing that several times, I began waiting too long; waiting to Hear the pause that told me, yes, the word was “the” and not “these” or “they.” Of course, this meant that I missed clicking during the word entirely, and frequently, the click came three or four words later. And I still occasionally anticipated and got it wrong altogether.
I found the amount and type of focus I had to use especially interesting.
I kept being distracted by the content and at those times, I totally missed quite a few instances of the word. I never even heard it; it wasn’t just a matter of clicking late or realizing it too late to click. I simply didn’t hear it.
At other times, I was too busy observing– noticing that I’d been wrong or late (and, very occasionally admiring my own timing)– and then I completely missed the next instance or the next two or three instances of my chosen word.
If the word came more than twice in a sentence, I never managed to click For all the instances, and often missed it even when it only came twice.
The ones at the start of sentences were easier because there was that pause ahead of time.
Next, I started using the natural patterns of speech, listening to a Sentence and trying to figure out where it was going. When I did that, the timing of the clicks was much better on the ones I predicted correctly, but I frequently clicked when I thought there should be a “the” and one didn’t appear-in other words, if I expected “Interested in the financial markets” and what was said was actually “interested in financial markets.”, I’d still click. (Can you tell I was using NPR?) Obviously, this produced a lot of false positives. I know I do this when working with the clicker and my dog, and it makes me crazy.
I thought that the exercise would be much easier when the word came as Part of a familiar phrase-say, in the name of a program or mid-way through a familiar sponsoring company’s name. It wasn’t that much easier, as it turned out.
If I broke everything down and listened to each word, one by one, without hearing the context, I’d get one and then miss several in a row.
The best results I got were when I achieved a sort of mid-range awareness of context-I had a sense of the kind of sentence it was, but wasn’t really listening and I wasn’t listening separately to each word, one at a time. It was a tough balance to hit, and it took me a few minutes to get there. And I could only keep it up for a few more minutes after that.
In general, the exercise was easier to do with newscasts than with interviews, where interviewees often spoke in incomplete sentences or ran words together or started to say one thing and hesitated or changed their minds half-way through words.
It took me a couple of minutes to get into the swing of it at all, and When I finally did, my focus didn’t last more than five more minutes, which is a lesson, right there. If I couldn’t focus for more than 5-10 minutes (and the first 2-5 were my getting into the swing of it,) , my dog’s training sessions certainly shouldn’t be longer than that, especially in the beginning of teaching a new behavior when I need to be clearest about which bit is what I want to see again.
For me, the take-home lessons from this are:
- I need to do this exercise a lot more
- It’s easy to miss catching a behavior if it happens multiple times in a row.
- I anticipate. (That one wasn’t actually a surprise-and coupled with a smart dog who also anticipates, this can be courting disaster.)
- My worst responses came at the beginning, right when my dog would be trying hardest to understand me.
- This takes lots of concentration. lots and lots of concentration. I wanted to sit down and rest after only a few minutes. If my dog is thinking and trying half that hard, it’s not surprising he benefits from a little rest afterward before we head out the door and off to do something else. I have a dog-trainer friend (not a guide dog instructor) who has long told me (and it sure seems to be true) that at least half a dog’s learning comes while they lie quietly and absorb what just happened. She likes to work with a dog for fifteen minutes and then pop him back in a crate for half an hour to mull things over before taking him out and working with him for another fifteen minutes. Not necessarily practical for us, but something to keep in mind for times when we can do it that way
- Comparing my experience with the act of picking out a word in a conversation vs. from a more clearly-articulated news report, it looks as if in any given session, my clicks are going to get more accurate as my dog gets better, not vice-versa, as one would hope– with the clicks starting out extremely accurate as the dog tries to grasp the basic idea.
- If simply clicking when I hear a word spoken is that imprecise, it isn’t going to get any easier when it’s something more than listening for words-say trying to click the instant my dog starts to do something while it takes me a moment to be sure what he’s doing is what I wanted and then another moment to click for it. I already knew that one, but this exercise sure brought it home. <sigh> And in all of this, all I was doing was clicking. No leash, no use of Touch to identify correct behavior; no treats; no trying to observe the whole situation and assess what my dog might be thinking, whether the dog is understanding something or what I might need to do to make things clearer or whether I need to back off and return to something easier, r any of those important things to remember during a training session.
<groan!> Oh, and when I’d finished the session, it was really, really hard to keep from listening for the word “the” and to stop twitching every time I heard it, and not only on the radio, but in normal conversation with another person! <grin>
This post is going to talk about practicing praise and treat delivery. All of us have had lots and lots of practice praising our dogs, right? We can still keep praising but now we need to pay attention to the timing of our praise, at least while we are using the clicker. You’ve heard the phrase click and treat…I add one more step in the process…click, praise and treat! Our dogs may not need praise after the click but old habits are hard to break! I am so used to praising my dog when she does something good that I find it difficult to let go of praise altogether. So while we are all in the bathroom practicing clicking with our tape recorders turned on, let’s add praise. The rule of thumb is click than praise. If any part of “good dog” or other words of praise slips out before the click sound, we need to keep practicing! We’ll talk more about the “why” later but for now, rely on the tape recorder or partner to help measure your progress. Oh, and one more thing…while we want to be sure our recording picks up our praise as well as the click sound, we need to take care that our dogs do not hear us practicing! The final step of the process is where the treats come in. The most important thing to know about delivering the treat is that we *must never move our hand toward the treat until we have clicked. Sound easy..it’s not!! Believe me! <sigh> Case in point…I was working on some clicker stuff with my second dog and Lukas Franck was helping me out. Every time I even so much as flinched a muscle to move my hand toward the treat bag without first clicking, Lukas would nail me. Most of the time I wasn’t even aware that I had moved my hand. This is where it comes in really handy to have another person watch our practice sessions. Okay, let’s think about the ways we can practice food delivery. Grab a handful of kibble and put it in the treat bag or in a bowl on the counter. We will also need another bowl or container to drop the treats into. Click, praise and then reach for a piece of the kibble. Drop the kibble into the empty bowl and start all over again. Whether we are sitting, standing or working with a partner, the trick is to sort of set up the situation so it feels like we really are giving our dogs a treat. We want the empty bowl to simulate the dog’s mouth. In other words, the empty bowl should be placed in relationship with where our hand would end up when giving the dog a treat. The whole purpose of this exercise is to improve our timing and observational skills. We need to learn how to respond quickly and almost without thought.
We practiced our click response to sounds, we practice our click response to touch and we began to condition our muscles so that they would be familiar with, and remember, these new kinds of movements. I want to take a moment here and expand on conditioning muscle movement. This will be an important concept to understand since we will later discuss conditioning muscles and how it relates to our dogs when they begin targeting, etc. Remember when you first come to class and begin working with a new dog? You wonder if you will ever get the feel of your new dog. You have to pay attention to the dog’s every movement and learn how it feels to walk with your new partner. Soon before you know it though, you are zipping down the street gracefully weaving in and out of pedestrians and other obstacles. Your movements are no longer choppy or unsure, and all your attention isn’t focused on how to move with your dog. We are trying to achieve the same type of fluency with the clicker, praise and delivery of treats. The other weekend we had fresh string beans. In order to prepare them for cooking, I had to snap off both ends of the beans. I set the bag of beans, an empty bowl and a pot filled with water onto the counter. There were several steps I had to take before I could cook the beans. First, I grabbed a bean out of the bag, then I snapped off the ends and put them into the empty bowl and then I had to put the finished bean into the pot of water. There was a sequence I needed to follow in order to get all the fresh beans into the pot for cooking. There were well over 100 beans to prepare. Each time I did a bean, I got a little faster. When I got to the middle of the bag, I was concentrating less on each individual step. By the time I was nearing the end of the bag, I wasn’t even paying attention to what I was doing. My mind had drifted and I was thinking about dessert! My muscles had repeated the same action so often that I was able to rely on their “memory ” to finish the job. If any of you have ever done piece work in a factory, you know exactly what I am talking about! With lots of practice, we can all learn this sequence of click/praise/treat or click/treat to the point of fluency. Remember, it took more than 1 or 2 tries with the beans before I could achieve fluency. I had to do over fifty beans or more before I really started to get the hang of it. The better we are with the click/praise/treat sequence, the easier it will be for our dogs to understand what we are trying to communicate to them! Good luck and let us all know how you do!
Okay, hopefully everyone has had some time to practice so let’s move on to charging the clicker! The first step is teaching the dog that CLICK equals TREAT. This is referred to as “loading” or “charging” the clicker. You will want to begin in a distraction-free environment. By this I mean you will want to find the most boring place in your home where nothing is happening…no people, no pets, absolutely nothing. You want your dog to have nothing better to do than focus on what you are doing. Get your bowl or bait bag and grab about 15-20 pieces of your dog’s most favorite treat. Each piece should be about the size of a piece of dog food. Have a seat and get comfy, relax. You want to be as nonchalant as possible…ho hum. Arrange yourself so that you have easy access to clicker and treats and you are now ready to start. Click once and give your dog a treat! You want to treat your dog immediately after the click. Repeat until you have used all of the treats. You should click and treat (C/T) as quickly as possible. Your dog must make the association between the click sound and the treat before any other training can begin. You’ll want to “charge” the clicker two or three times a day for a few days until your dog starts to immediately perk up and give you attention at the sound of the click. For now you are not focusing on any one behavior, just on getting the click and treat down.
Tips for Success
- The only noise the dog should hear is the click, don’t give any commands
- Do not click in your dog’s face or right next to its head.
- Remember not to go for the food until after you click.
- Never click without giving a treat, even if you clicked by accident!
- If your dog seems afraid of the click, stop. Get an i-click clicker or muffle the sound of your box clicker by putting a few strips of medical adhesive tape on the metal strip or stuffing the clicker in your pocket. Payton was a little hesitant of the clicker when I first started using it. The first time I clicked, she stepped back a few feet. This wasn’t good so I stuffed my clicker hand under the blankets in order to muffle the next click sound. I just acted like everything was normal, yawn, etc. and held out my hand so she could take the treat. Her stomach rules so she was able to overcome her concerns after a few repetitions. Good luck and for those who have already charged the clicker, please let us know if you have any other tips or experiences to share!
Question: I guess I just don’t get it? When you start this thing, do you just click and treat, when the dog begins to do something you want, like sit or rest or whatever. Do you just start doing it, is that right? How many times would you suggest to treat in a row?
For now you shouldn’t be clicking for any behaviors, just charging the clicker. So, you have your bowl of treats, your dog at your knee and your clicker in hand. Then in rapid succession, click treat, click treat, click treat, until your dog makes the connection between the click sound and the delivery of the treat.
You can’t teach your dog a behavior until it makes this connection in it’s mind.
when you are first getting the dog to understand the association of the click and treat occurring very close to each other, you *are* not looking for a behavior. You just click, treat, click, treat, click, treat to get the dog to understand that the click is going to be followed by the treat. They have to understand those two things first before you can use them on actual tasks with the dog.
During this part of the training, i.e., “Charging the clicker,” you can think of the word, “Charging,” as you would think of it with a battery.
If a battery isn’t charged it won’t run whatever device it’s in. It’s the same with the clicker. Just the sound of the click doesn’t mean anything to your dog in the beginning. What we’re trying to do here is associate the sound of the click with something absolutely fabulous to the dog. That’s why we keep talking about yummy treats. If your dog hears the sound of the click and immediately receives a yummy treat then, in time, the click will become associated in his mind with the yummy treat. The clicker will take on the same association that juno has for the yummy treat. At this point in time we’re looking for absoolutely no particular behavior from Juno. In fact, it doesn’t much matter, just now, what your dog is doing at the moment that you click.
Hi everybody. I’ve charged my clicker 4 times. Holmes loves to eat as much as I do, and I am lucky that I can get away with carrots. he thinks they are grhrhrhrhrhreat! I can’t tell whether he is starting to associate click and treat yet, but he knows i. have the treats and that’s all he cares about. Hardest thing is keeping myself from going for the treat before clicking. In my effort to click then treat as quickly as possible, i.
anticipate myself. And I know that Holmes is watching the hand going for the treat, so do I noticeably wait to start getting the treat until after I click? What is an acceptable delay between the click and treat?
I am really enjoying all of the informative posts from the crew! Keep up the good work. I wish i. would have started this ages ago! It’s a great way to have fun with my pal! So how do i. Know when my little detective has caught onto this? For once it’s Watson throwing clues around!
You do want the clicks and treats to follow each other as quickly as is possible. Don’t worry, even those of us with more experience fall into the same trap (or at least I do) of having one hand heading towards that treat bag or even resting on it (also not good) before I click. What you might do is put the hand that will deliver the treats behind your back. Yes, the deliverly will be slightly delayed at first, but that will begin to help you discipline yourself. Later, you can have the delivery hand at your side. Trust me, it takes time, and you will be rewareded with your efforts.
So, how do you know when Juno has made the connection between the click and the treat? After several repetitions of the twenty or twenty five treats in the bowl exercise, you should be able to click and observe that your dog suddenly attends to you. It should only take a couple of repetitions of this to establish whether or not your dog has come to connect the sound of the click to a very good thing. When I was charging the clicker with Lukas, the puppy, I did the, “Click/treat,” thing two or three times a day for three days. Then, while we were all outside and the dogs were running around I clicked just as Lukas ran by me. There should have been a cartoon showing this dog galloping full speed ahead and then smoke coming out from under his paws as he slammed on the breaks at the sound of the click. Don’t do more than one or two repetitions of this. All you want to establish is that the connection has been made.
Question: One more question to any one who can tell me. When you hold the clicker,do you click with one hand, and treat with the other? Do you hold the treats out of sight, so the dog can’t see them? Also, if the dog starts to look for the treat, what do you do? Thanks,
I click with one hand and treat with the other. I click with my right hand and then reach for the treat with my left. I like using the left hand to treat because it keeps my dog from turning and reaching across in front of me to get to my right hand. It also seems harder for the dog to watch the bait bag and the treat is more likely to seem to come out of nowhere if it’s coming from directly behind him– assuming he’s on my left, that is.
If he’s at my left side, I like to bring the treat down between my body and his head so I present it where it is easy for him to reach without turning around or crossing across in front of me
Obviously, if the dog isn’t on the left as I train, I switch all of this around to the configuration which seems to make for most efficient clicking and treating.
One of the mistakes I’ve made was reaching for the bait bag where I keep the treats before clicking. If I got one bad habit out of my system at the workshop in March, that would be it. It’s better not to even carry your hand near the bag, pocket, bowl or whatever you’re using. The reach for it is as much part of the reinforcement as the treat, itself. And that shouldn’t come until after the click, albeit right away after the click. For me, reaching too soon was a worse problem than taking too long to get the treat out. The dog got fixated on the hand by the bag, rather than what he was doing to earn that click/treat
Now, like many others, my dog knows the clicker is associated with getting a treat, and he’s been known to nudge the clicker, presumably to encourage it to click so a treat will then follow. Again, this is something which isn’t reinforced and eventually extinguishes itself. I think most dogs will start looking for the treat directly. Fine. They don’t get clicked for nosing the bait bag, and they don’t get a treat if there’s been no click. This seems to sort itself out pretty quickly as they come to terms with the idea that no click is free and no treat comes without a click
I still find myself coaching myself “Click, then reach for treat…”
As your dog begins to catch on to the fact that the sound of the click means a treat is coming, there will be the inevitable mugging for the food. When and if this happens you need to be prepared to react to it in a positive manner. The goal here is to teach your dog that you will play the clicker game only if he is not seeking the treats
All of our work in learning to interpret our dog’s movements through touch and sound are going to begin to pay off here. I’ll set up a situation just for an example. Suppose I’m charging the clicker while sitting in a recliner. I have the treats in a bowl and the clicker in my right hand. I place the bowl in the corner of the chair, to the left of and slightly behind my left hip. I’ll be taking the treats with my left hand after I’ve clicked. I’m in this very boring place so that the dog is really paying attention to me. As I begin to click, reach for the treat, and offer it to my dog, the dog becomes even more focused on me. At first, the dog sits quietly in front of me, taking the yummy treats as I click and offer them.
Suddenly I feel the dog’s nose poking around my left hip. I immediately interupt the click and treat that I have been doing. I might have to cover the bowl with my left hand or just move my body so that the bowl is not accessible to the dog. My goal here is to withhold the click and treat until the dog ceases the seeking behavior. As soon as Juno is sitting quietly again, focused on me instead of the food, I begin the click and treat again.
How can you tell when your dog has stopped seeking the food. Well, the most obvious way is that Juno stops trying to take your fingers off. There are other, more subtle ways as well. You can place a bell on your dog’s collar.
When the bell stops ringing you know that your dog is sitting or standing still. If you listen you’ll probably hear the dog sniffing when he is seeking the treats. When this sound stops you know the seeking behavior has been suspended. You can try sitting forward in the chair so that your knees touch the dog’s chest. You’ll be able to feel when the dog stops pressing forward.
As soon as you can tell that your dog has stopped seeking the treats, click and treat immediately. This might take some patience with the more treat obsessed dogs. Be prepared to wait out these dogs. Sit back and look bored. Yawn. Look away. Eventually your dog will cease the seeking behavior because it’s not getting him anywhere.
When you do this click treat thing with your dog in a chair, do you just practice the dog not taking a treat, withoutt teaching a task. Sorry for these questions, but I am very, very new at this clicking thing. Thanks, and take care.
All we’re trying to do at this stage is teach the dog the connection between the sound of the click and the delivery of a treat. We’re not asking anything of the dog yet. The only reason to stop your C/T is if the dog tries to climb in your lap after the treats. If Juno is keeping all four feet on the floor and just taking the treats you offer him just continue with the click and treat until you’ve used up all of the treats.
Remember to start this process with about twenty or twenty-five treats and offer them after each click until they’re all gone
The best possible outcome here would be for Juno to be sharply focused on you and politely taking each offered treat. You may never have to deal with the, “Mugging for food,” behavior from your dog. It’s just that a couple of people have already brought up this question and I’m trying to help folks head off the food dependency scenario.
Some dogs will get it in one session, others will take longer. First, if you’ve noticed that, while charging the clicker, you have a SHARK who likes to bite the food out of your hand, try closing your fist around the nugget of food. Let Juno sniff, lick, bite, whatever, but do not open the hand and give the food until Juno stops this behavior. Sooner or later, Juno will either turn his head or pull back from your hand. This is when you click, praise and deliver the food at first from the open palm of your hand, and gradually work up to finger-held delivery. Practice this several times until Juno consistently waits politely for his treat. We want our dogs to know that they don’t get the treat until we give it to them.
* Again, you may need to go back to teaching your dog good manners from time to time. This exercise should be treated as a separate session and not combined with targeting or any other exercise. For instance, just say you are working on having your dog touch your hand. You notice Juno is snatching the food.
End the touch exercise on a happy note, take a break for a few minutes and start up again but this time only working on the snatching behavior. If you noticed that your dog is only snatching when you have yummy fresh cooked chicken, use chicken to train with. If your dog is only snatching when another dog is close by, practice the exercise with another dog nearby. Later, in another session, you can go back to training touch or any behavior you were working on.
If biting begins again, go back to keeping the food in a closed fist and start the process over. I have found that “shark” behavior tends to crop up when I train right before meal time, when I am using high value treats, or when another dog is around. And yes, I do have a lab/golden cross! <Sigh>
This shepherd is driving me nuts! She still hasn’t or apparently hasn’t connected the click with a treat. I have used cheese, bits of meat, etc.
She is not the stereotypic picky shepherd. She enjoys her food and treats even though she dines slowly, piece by piece while laying down, and she will eat anywhere I take her food.
The problem started out trying to overcome her take it command, of which I have always been very proud. We worked on it until she won’t take anything even from a family member without being told “take it”.
During practice sessions, she would rather sit or stand with her head in my lap as if she just doesn’t understand, or if she is hungry she will take anything I offer, no longer waiting for the take it command.If she isn’t particularly hungry or if she is getting bored, she won’t even bother lifting her head to take the food after I click.
We have been working at it for 2 or 3 times a day since we were told to do so, which was Friday I think.
It’s possible that if you give her the treats (after the clicks) in a different way then you tell her to “take it,” she may start to get it. What I mean is that for example when I hand Brook a treat that she can “take,”
from me, it’s with my palm flat and facing up toward the ceiling. This is a visual cue to her that it’s OK to take it along with the verbal cue. When I am practicing “touch,” or :”leave it,” with her, I my hand is closed with the treat inside of it or it’s between my fingers
Since Camilla is a dog that has been taught not to take food from people, and she’s learned that lesson well, you may have to keep trying until she understands it’s OK to do it when you say it’s OK.
Another possibility is that she is a dog who either is not as food motivated, and maybe another type of reinforcement like her favorite toy might work. Still another possibility is that you haven’t yet found the kind of food that will get her going. Some people use cut up hot dogs, or chicken strips; anything that the dog values above other food.
Great work pulling out all the stops…using high value treats is a good start.
To help Camilla overcome her reluctance to take treats from your hand, you may want to try offering her the treat from her food dish for a while. Practicing these sessions right before meal time may help to heighten her interest as well.
So, after you click, instead of giving her food from your hand, drop the treat into her food dish and let her take it from there. If this seems to work, after a few sessions you can gradually introduce your hand back into the mix by first keeping your hand on the food dish after dropping it in, then holding the treat in your hand while your hand rests in the food dish, and so on.
You may also want to check in with one of our clicker savvy TSE instructors…Lukas, Will or Peggy G. for more ideas if you continue to experience problems.
I am confused now. Not only do we adapt the handling methods, but we use real food too? Not to offend anyone of course… I understand that food is a great motivator for dogs even with the praise. I was just confused by the use of “real foods,” I.e. people food. If your giving your dog food that people eat won’t they sniff more or expect it in restaurants etc? What is to stop the dog from thinking the chicken or hot dog on someone’s plate is there’s if we start food treating our dogs with foods we have previously been told and taught not to give them for health reasons and for working reasons. What are instructors doing regarding this matter?
The answer to your question is a technique called counter conditioning. The reason a handler would use something other than kibble or training treats would be if Juno isn’t motivated by those things, the handler still wants to use clicker training, so they try other forms of food to get the dog interested. By doing exercises that practice “food refusal,” Juno learns that the treats come when he/she has done something to earn the treat. Juno still gets corrected for trying to steal food, and the goal is to learn the time and place when food is acceptable.
Counter conditioning gives the handler a chance to teach Juno that he/she can refuse food on the ground or someone else’s plate, and if he/she refuses, then Juno will be rewarded by being given a treat. I was shown this technique first by Pete Jackson.
An example is; I have two cats. Their food is on the floor in my kitchen.
Periodically, I practice counter conditioning with my dog who loves cat food. I put her leash on and we walk by the cat food and I tell her to “leave it.”
If she does–click praise and treat. If she doesn’t or tries to sniff it, I stop and back up a few steps and start to heal her by it again. What she wants at that point is to keep walking, so I’ve taken away her chance to steal cat food, and her opportunity to keep moving forward. So, the next time we walk by the cat food, she looks straight ahead–click praise and treat. Counter conditioning is something handlers have to continue to work with some dogs on, and others will learn quickly that if they focus their attention on the handler, they will get rewarded with food. I can say that in almost three years of using clicker training with my dog, I haven’t had any experiences where she’s tried to steal food off of someone’s plate. The key to this is Juno learns treat aren’t given indiscriminately, they’re given for something Juno does. Also, if Juno is getting food from the handler, the need to be looking for food in other places actually lessens.
Yes, you are absolutely right. If used incorrectly, food can create lots of undesirable behaviors. When used correctly, however, it can be a very powerful tool. The introduction of the clicker allows us to use food in a way that won’t create the types of problems you have experienced in the past.
Regarding the answer to your question about what TSE trainers are doing: The best way to find out is to call TSE and ask them directly. The toll free number is 800-539-4425. You can ask for Lukas, Will or Peggy G. This may also give you a chance to discuss your concerns about using food with your dog. Sometimes it’s better to talk with someone rather than trying to figure things out through email.
The following excerpt was written by Melissa Alexander, author of Click for Joy. It may help to answer Jon’s question about food dependency and Linda’s question about why not just praise.
Do I have to use food? Why can’t I just praise my dog? Won’t this mean I’ll have to carry around a clicker and treats forever? In order to increase the occurrence of a behavior — which is what we’re trying to do when we train new behaviors — we have to reinforce the behavior.
To do this, you must use a reward that the animal finds reinforcing. That reward can be food, playing with a favorite toy, a belly rub, or, yes, praise.
You must find what motivates your dog to do his very best. It’s a common myth that clicker trainers don’t use praise. Of course we do! I praise my dog frequently, both in and out of training. But when I’m teaching my dog a new behavior, I want him to be as motivated to get it right as I am, so I use a higher value reward, usually food, in addition to (not instead of) praise. It’s another common myth that using food in training will produce a dog that only works for food or that you’ll always have to carry a clicker with you.
You can prevent both of these situations. Reward, don’t bribe. If you’re using a food lure, fade it quickly, then don’t have the food visible when you ask for a behavior. In fact, I like to keep the food in a dish off of my body. The food or toy should be produced only after the dog has performed the behavior. Once the behavior is on cue, and dog will offer it willingly, fade the clicker and use a verbal marker instead. A verbal marker isn’t as precise as a clicker, but at this stage, the dog knows what’s being reinforced. Consider the verbal marker a praise marker, letting the dog know that he did something reinforceable. Once you switch to a verbal marker, begin varying the types of reinforcers. Give him food one time, then play with a toy, then just rub his ears and praise him. Eventually, you can rely on praise more and more often. If you find that the dog becomes frustrated when you begin using other motivators, go back to using a higher ratio of food treats and decrease the ratio of food treats more slowly in the future. You want the absence of the food to motivate the dog to try harder, not to frustrate it into quitting.
Why? An excerpt from an article by Debi Davis, http://www.clickersolutions.com/articles/2002b/targeting.htm
Note: The author refers to a target stick but your hand or a homemade target is what we’re going to use.
“So, your dog knows to touch the target stick with his nose. Have you now raised and lowered the target stick so that he has to take several steps to reach it? Have you moved it closer and closer to the ground, and up high so he has to stand up on hind legs to reach it? Have you moved across the floor and had him follow you, nose to that target, as if his nose was magnetized?
These are all precursors to wonderfully helpful behaviors for every dog to know, whether the dog is an athlete or a couch potato. Dogs who have opportunities to learn new things all the time are happy, healthier, more resilient and stable dogs. With the target stick, you can teach a multitude of new behaviors, keeping his enjoyment of problem solving high.
Though the target stick is one of the most useful tools in teaching new behaviors, it’s also one of the best for getting dogs quickly engaged, focused and
excited about each teaching session.
End of excerpt.
Hopefully, everyone has had a chance to charge the clicker and all of our dogs now know that the click means something wonderful is coming. This next step will begin to teach your dog that he can have some control over this clicking and treating. This is a hugely important concept for your dog to learn. The best part is that it’s also very easily learned.
Place a bunch of treats in a bowl or in your bait bag. Take the bowl, your dog, and the clicker to a quiet place. Hold the clicker in your dominant hand. Wrap your hand right around it so that it’s concealed in your fist.
Then take a treat and hold it in your left hand in exactly the same way.
Hold out both hands side by side. Your dog will most likely begin sniffing at your hands. Hold off the click until the dog actually touches your left hand, the one with the treat in it. As soon as you feel the dog touch your hand, click and open your left hand to give the dog the treat. Do this for many repetitions, about the same number as when you were charging the clicker, twenty to twenty five times. Repeat this a couple of times today and then tell us of your results.
If your dog doesn’t seem to get it that he should be touching your left hand in order to get the click and treat you can try looking at that hand. It’s fine if your dog can see that you pick up the treat and fold it into your left hand. That’s another cue that will get him to touch that hand instead of the hand with the clicker. Sorry for any of you folks who are southpaws.
Just reverse my left’s and right’s.
If your dog gets it right away you can begin to hold your closed hands a little further apart. Keep in mind that all we’re trying to convey here is that the dog has some control over the C/T. He needs to learn that there is some behavior that he can do that will get you to click and treat. Good luck and let us know how it goes.
Louella just loves this click and treat stuff! She figured out very quickly that my left hand holds the treats. The first time we tried it, she sniffed both hands, but after just a couple times she was going for the left hand all of the time. As I moved my hands apart, she continued to go for my left hand almost all of the time. Of course, the times she went for the right hand I did not click. The last set of exercises, she sniffed the right hand only once, after there were no more treats. One thing I’ve noticed: since starting all this she seems even happier than she already is and more attentive to me. I think I probably need to vary where I do this?, because when I call her into the computer room, where it is very quiet, she comes bounding, as if she knows we’re going to click and treat. Of course, we don’t always do that in here, but she’s certainly excited to come in here!
Forgot to mention also, that another thing I’ve noticed with Louella since starting this is that she is now touching my hand with her nose a lot. For example, I just gave her the come command while walking back toward the computer room and as she came up beside me, she touched my hand with her nose.
That’s exactly what we expect to happen. The dog is so thrilled to learn this behavior that earns it clicks and treats that it’s going to throw the behavior at you all the time. Once the behavior is established we’ll give it a name and put it on cue. That should give you control of when your dog touches or does not touch your hand.
Hi. Holmes is having a little trouble because I’ve taught him that he must politely wait for the treat. He just sits and looks at me and waits for the treat. I’ve been getting very close to him and put my closed hands a few inches from his face. Then he will nudge the hand with the treat. I’m so lucky he loves carrots! Is he learning contradictory behavior here?
Mary Anne, I’m not sure if you’re teaching contradictory behavior or not. Just to be sure, though, you could go ahead and move Holmes on to touching a different target. If you rig up a target on which you can attach a bell you can start with that instead of touching your hand. The target that Lukas fixed up for me was the lid to a yogurt container. If you place it on its edge, poke a hole at the top and bottom you can attach something through the top hole that can be used to hang the target and a collar bell to the bottom. A rubber band is fine for the top. It can be used to hang the lid on doorknobs or whatever you want to teach Holmes to touch.
Once you have your target fixed up you can present it to Holmes. He probably won’t touch it right away. Remember not to call him towards you.
At this stage with Kismet I was convinced that she was not the brilliant dog I thought she was. It seemed to take her forever to get it. I wished I had been able to smear something yummy on the yogurt lid to try to help her get the idea. I’ve read that this tactic is very common with clicker trainers.
A little smear of peanut butter or something like that could help Holmes get the idea. If Holmes is totally uninterested try moving around the room. I remember Lukas telling me that I had to be the most interesting thing in the room.
Now, let’s talk about teaching our dogs to target a hand or specific object. Step 1. If using your hand as the target, take a really yummy but small piece of food and wedge it down between your second and third fingers. Slowly extend your hand out and put it one or two inches from Juno’s nose. If Juno looks at your hand, noses it, licks it, nudges it, or bumps it, Click, praise and treat with the same piece of food from your fingers. Repeat 3 or 4 times and then stop putting food between your fingers and just hold out your hand again.
Be ready to click. Juno will most likely continue interacting with your hand, even if it is only to turn his head toward it. Click/praise and treat.
Pay close attention; you don’t want to miss this first interaction after taking away the food lure. It’s going to be fast…and easy to miss if you are not prepared. You can also do this same exercise without food or by using a small plastic lid or some other type of home made target. If you choose to use food, simply smear a tiny tiny bit of peanut butter on the lid, just enough to get the smell but not enough to get a lick. If you find that your dog is licking the target more than twice, omit the click on the next lick and wait for another behavior like a nose bump to click. We don’t want to reinforce licking so we do not reward this behavior with a click. By not clicking, we are telling our dog that this behavior, licking, is not what we want now. We never have to say a word, our clicker will tell the dog what we want it to do. Our dogs need to learn how to experiment to find out what it is that we want. Put yourself in Juno’s shoes (or paws) for a minute. , Juno licks the target, we click, Juno licks the target again, we click again. Juno thinks wow, this is cool, if I lick the target, I get clicked. He licks the target again but wait, my human is not clicking. Here, let me lick again, maybe she missed it, no, she is still not clicking. hmm, well…what if I try something else, how about a little bump with my nose. Yee haw, she clicked! Okay, let me try it again…bump…and yes, click! I can make my human click when I bump my nose on that stupid looking piece of plastic…no problem, piece of cake! This is great fun! So what on earth was all that about? Well, we are changing the rules. We’ve always commanded our dogs to do something and now we are asking them to figure out what we want them to do with no commands. Remember Gary Wilkes comparison to the childhood cold/hot game, this is what we are doing with the clicker.
We click when the dog is getting hotter and don’t click when it is “cold.” We are “shaping” our dog’s behavior. We do this one little step at a time until we get the final behavior. Tips for success:
Start out in a quiet, distraction free location.
Keep training session short, especially when training something new. You can always work for a few minutes, put the clicker gear aside and come back to it in a while for another brief session.
Don’t call, chat, encourage or otherwise “help” the dog to touch your hand. Sit quietly and patiently for about 30 seconds. If your dog isn’t doing anything, stop and try something else. Try moving positions or just try again later.
Do not stare at your dog intensely, or somehow try to “will” it to touch your hand. Relax, smile and have fun. I like to sit on the floor when I first start this exercise. This helps me not to lean over my dog when holding my hand out.
If you click by accident, you must still give your dog a treat! It’s not Juno’s fault you goofed so pay up! <smile>
End each session on a good note and a special word or phrase to let your dog know the game is over for now. I say “all done” in a happy voice right after I give the last treat. If your dog does not offer any clickable behaviors during the session, just ask it to perform a known behavior like sit so you can C/P/T before ending the session!
Use all your senses to help you identify when Juno makes any movement toward touching your hand. If you have no vision, try using sound or the touch technique we discussed in our earlier lessons (let us know if you need a copy to review) . Resist the very strong temptation to “help” move your dog’s head toward the target. You will also need to plan how you are going to hold the clicker. I use the i-click clicker when I am working in a quiet place with Payton.
This way I can hold the clicker in the hand that is touching the dog. Don’t hold the clicker in the hand that your dog is targeting. You can slide the wrist coil up your arm and click by pressing your elbow against your side or the wall or another object, you can rig the clicker up under your foot and so on…. Just be sure to think about how you are going to do this and practice–without the dog, before beginning any training session. We need to plan out the steps that will help our dogs be successful. If we’re fumbling around trying to figure out how to click, treat, touch and all that, we’ll miss a lot of the action and our dogs may lose interest or at least get very confused! We don’t want to miss an opportunity to click so plan ahead! Now for those of you who have some vision, consider using an area that will maximize your vision e.g. use a well lit area, use an area that provides contrast between dog and floor, etc. Again, take the time to plan out your session before getting started.
I want to say that I am very much enjoying this list and I hope that folks won’t mind my observations/questions; I’m very new at this click and treat thing, but am finding it to be lots of fun. I have one of those pups who, though she doesn’t bite,does lick, lick lick and so we’ve been working on her taking treats politely from my hand rather than trying to snuffle them out. <LOL> And actually, I was amazed at how quickly she figured out that she wouldn’t get the treat if she licked my hand. On the
sixth try the first go-round yesterday, she backed away from my hand, I clicked, praised and she took the treat, and thereafter it was really fun to watch her reactions. I could almost hear/feel her doggie mind thinking, “Now what do I do to get this treat?” Maybe I’ll touch the hand with the clicker … “No, that didn’t work, so back to the other hand and I didn’t lick and she clicked and praised and I got the treat …” So today we tried it again. At first she snuffled, but with no response and back to the treat in the fist, and she started doing it right again very quickly. We did this a couple times throughout the day.
So we just tried the touching my hand with her nose exercise, and that was quite interesting. We sat here for a minute or so, then Louella touched my hand with her nose, I clicked praised and gave her a treat. After a little pause, she did it again, with the same result. The same thing a couple more times, and then she started getting all happy as if she was starting to get it, her touching my hand response began to come more quickly, and now we’re finished and she’s running around the house like the complete goofball that she is when she’s happy and/or just so proud of herself.
At any rate, if I am understanding this, the dog is now associating doing the thing I desire for the click and praise and treat. Correct? When the click comes, she knows something else good is about to happen. I want to do this a couple more times with my hand, and then I think I’ll try the plastic lid. That ought to prove to be very interesting.
Following are several comments based on an earlier post. The original post is noted and the further comments are preceded by three asterisks.
Original post: Now, let’s talk about teaching our dogs to target a hand or specific object. Step 1. If using your hand as the target, take a really yummy but small piece of food and wedge it down between your second and third fingers. Slowly extend your hand out and put it one or two inches from Juno’s nose.
This exercise is done with an open hand, not with a closed fist. Original post: If Juno looks at your hand, noses it, licks it, nudges it, or bumps it, Click, praise and treat with the same piece of food from your fingers. Repeat 3 or 4 times and then stop putting food between your fingers and just hold out your hand again.
This technique is called luring. There are several ways to get a behavior and this is but just one of them. Many clicker trainers will use this method to get a behavior started. There is a danger that, if food is used for too many repetitions, the dog may only work for food. The trick is to fade the food out fast. This is why I suggested using food for only 3-4 repetitions. Other trainers may use food for more repetitions. I generally don’t like to use lures because my dog is a bit food obsessive to begin with…sigh. Original post: Be ready to click. Juno will most likely continue interacting with your hand, even if it is only to turn his head toward it. Click/praise and treat.
Pay close attention; you don’t want to miss this first interaction after taking away the food lure. It’s going to be fast…and easy to miss if you are not prepared. You can also do this same exercise without food or by using a small plastic lid or some other type of home made target.
As mentioned above, you can also do this exercise without the use of food. , Again, there are a number of ways to get behavior from your dog. It is up to you to choose the method that best fits your lifestyle, your goals and capabilities. There is no “right” or “wrong” way to do this. If you choose to use a lure, you will most likely get the behavior quicker. If you choose not to use a lure, you will need to focus harder on breaking down the touch behavior into small enough increments so that you can keep up a high rate of reinforcement. One of our members who wrote with concerns about using a luring technique offered a great description of how she set criteria low enough so that she could reinforce her dog’s movements at a high rate. Whether you use a lure or not, this is a great description of how to break down a behavior. I’m pasting it below. “I do not understand why we are not simply offering an open palm at nose level and saying nothing until the dog moves toward the target and clicking each proximity of movement toward the target until the physical touch takes place. This is how I taught XXX “touch” to target my hand and I did not name the behavior until she put muzzle to palm. I used a light lead and in the beginning accepted everything from a head turn in my direction; a step toward the palm; breathing on the palm; a body frush to the palm; and even a lick to the wrist of the palm; and finally a nose to the palm. She did plenty of thinking and it took multiple sessions before she figured it out, but she did figure it out in the end. ” Okay, did you notice how small those steps were leading up to the final touch? Let’s take a closer look:
- a head turn in my direction Human clicks 2-3 times for this movement. If Juno does not progress to another movement that leads him closer to the touch, omit click until another behavior is offered.
- a step toward my palm This is a clickable behavior, Juno is moving closer to achieving desired behavior. Again, do not continue to reinforce this behavior after 2-3 times. You don’t want Juno to get stuck here. We need to encourage him to keep moving closer to the desired behavior. Remember the hot/cold game?
- breathing on the palm Great, Juno is another baby step closer to the touch. C/P/T about 2 times. We are shaping Juno’s behavior. Again, omit the click after a few repetitions so Juno won’t think that all we want him to do is breathe on our hand.
- a body frush to the palm; and even a lick to the wrist of the palm Again, we are clicking and treating for behaviors that are moving in the direction of the final touch. We are gradually increasing our criteria so that each baby step brings Juno closer to the desired behavior. Juno may go back and offer an earlier behavior that he was clicked for in the same session but we simply do not click for those behaviors.
- and finally a nose to the palm Great, jackpot! Some trainers will offer the dog some bonus treats rather than the usual one piece of kibble when they do something really great. Don’t be surprised if Juno goes back to an earlier step any time during this process though. Just because he offerd you the final behavior once or twice doesn’t mean that he totally gets it. It’s very likely that all these steps will take place over a period of several sessions. When beginning your session, don’t start up where you left off, go back a step or two or even from the beginning if you have to. The trick is to set your dog up for success by making it as easy as possible for Juno to earn a treat. The more times you can C/P/T Juno in a single training session, the quicker he will learn.
So we just tried the touching my hand with her nose exercise, and that was quite interesting. We sat here for a minute or so, then Louella touched my hand with her nose, I clicked praised and gave her a treat. After a little pause, she did it again, with the same result. The same thing a couple more times, and then she started getting all happy as if she was starting to get it, her touching my hand response began to come more quickly, and now we’re finished and she’s running around the house like the complete goofball that she is when she’s happy and/or just so proud of herself.
Excellent, sounds like you are both having fun!
The last dog I taught to target took a longer time than the other dogs I’ve worked with and that of your description of Louella’s progress. This dog was young but not very interested in touching a target. I would only get a few clickable behaviors, if that, and then nothing so I made sessions very short, about a minute. I also used a high value treat – fresh cooked chicken pieces cut into pea-size chunks and planned sessions right before meal times. I gradually increased my criteria as each session progressed but it took quite a while, several days actually, to get the dog to touch the target at all, let alone with enthusiasm.
If your dog just keeps sitting and staring for more than a minute, stop and try again later or think about using a different approach. We also need to be sure that we click on even the slightest hint of action toward the hand or target. If you have a sighted person available, you can ask them to help out. Simply explain to them that you want them to click if Juno looks at your hand/target when you extend your hand out. Do this a few times and then hold off the click until Juno makes another movement towards the hand/target. It’s important to give Juno a high rate of reinforcement to keep him going. We can do this by breaking down and rewarding each little step and making sessions short. There are lots of little actions that take place between the time that you hold your hand out to the time the dog actually touches your hand…we need to think about how we can capture these movements with a click.
Although this game is fun for dogs, it can be stressful too. If Juno is going off to lie down or to scratch, etc. in the middle of your training session, stop. He is most likely feeling confused, stressed or too pressured. Remember, this is all new to our dogs and can be very stressful to them if we are pushing our dogs too hard and too fast.
At any rate, if I am understanding this, the dog is now associating doing the thing I desire for the click and praise and treat. Correct? When the click comes, she knows something else good is abot to happen.
Louella is beginning to understand that she is now “in charge” of making you click. She is experimenting to see what behavior will get you to click. She is not waiting for you to command her to do something but rather she is experimenting with or trying different behaviors to see what you want her to do. She is learning that when she hears that click, what she was doing right then is what you wanted her to do. Super job.
I want to do this a couple more times with my hand, and then I think I’ll try the plastic lid. That ought to prove to be very interesting.
Hold on…not so fast! Remember, slow is fast in dog training! <smile> It is so tempting to move ahead too fast when we have a successful behavior and a bright dog!
You will need to begin practicing this exercise in different locations of the house, outside, etc. Try holding your hand further away, up in the air, down on the floor, while you are standing, sitting, kneeling, etc. Think about the ways in which you can generalize this behavior. You will also need to add a cue at some point.
Good luck, it’s so exciting to hear about your progress!
I have a question which may be premature. Do you eventually associate verbal cues to all of these new behaviors? How does Juno know which behavior you want at a later stage? Do you eventually name all of the targets you want Juno to recognize? I know we are a long way from this, so yes or no is fine!
I’m glad to read that it is normal to backup a few steps at the beginning of each session, because that is what I’ve had to do. I’d swear Holmes has it only to find that I need to back up at beginning of session so he can win treats. Of course, I’m not taking all of this too seriously so that it will be fun, and I make a lot of my own mistakes which I am sure is part of the problem. Clicking exactly on time is not always easy to do because I sometimes miss the cue and a split second later figure I’d better click. by Then, I fear I have missed the moment. I feel like an uncoordinated dummy! You know, the thumb gets stuck in the clicker without making a click, or I just drop the thing. I really need to get an I-click which I think would be much easier to manipulate.
All, Reading along, it seems like timing the click with the desired behavior is critical. In the descriptions given in most training articles, this seems to be a totally visual feat. Does anyone have suggestions for a highly accurate, non-visual method for determining when your dog is exhibiting the desired behavior? Most difficult to me are the small, subtle steps you see when trying to shape a new behavior, not the obvious things like sitting, resting, or the like. I have been trying to use my hand to feel when my dog does what I want, but I am concerned that my hand will become part of the cue or a hint as to the behavior I’m looking for. Advice would be appreciated.
Before working on charging the clicker, it’s good to spend some time working with your dog on learning to read him/her with both touch and sound. First-touch Ginger made a point of talking about how to read your dog before it does something rather than after which is how we get signals through the harness and leash, after the behavior has been performed. So for example; you would lightly put your hand or fingers on Juno’s neck and/or shoulders. This is where much of the muscle movement it takes to perform a behavior will begin.
It’s important to do this before charging the clicker and starting to teach a behavior so Juno becomes comfortable with your touch and it doesn’t distract him. If you spend a day or two doing this, you may find you’re learning when your dog is going to do something before he does it.
I try to do several things:
I put the back of my hand lightly against my dog’s ear as she moves toward an object I’ve asked her to find. This doesn’t seem to distract her, and it gives me more information about what she’s doing and when she’s doing it. Sometimes I’ll quickly put my hand under her muzzle as she’s pointing to an object, or I’ll follow her nose which leads me to her line of sight and I can find what she’s looking at like an elevator button.
You can attach a bell to her collar just for the purpose of hearing what she’s doing, or you could attach a bell to an object you want her to touch.
Both of these techniques will provide you with more information. If you do this work before using a clicker with your dog, it will make it easier to use the clicker and follow your dog’s signals.
It does seem a bit discouraging when we read about this new method of training and how vision seems to play such a large role in its success. I suppose it’s just like everything else — there will be things that we just can’t do but that’s life. The good news is that there are a lot of things we *can do and our dogs *can still learn despite some pretty dismal timing.
There are some suggestions in the archives but what we really need to do is jump in there and give it a try. Through our own experience, we will learn what works and what doesn’t.
There have been a number of posts on giving the behavior we’re teaching a name. We’re not quite ready for that yet. Why do we wait until the dog has thoroughly learned the behavior before giving it a name? Keep in mind that we’re teaching Juno a whole different way of thinking here. We’re putting the power in Juno’s paws. It’s up to Juno to figure out what he needs to do in order to get us to click and give him a treat. For most dogs this is such a new concept that it’s going to take quite a lot of concentration.
Yammering away is just going to distract Juno. Later we’ll add a verbal cue that can be used when Juno is trying to figure out something but is on the wrong track. For now, however, the only sounds that should be coming out of your mouth should be words of praise just after you click.
To help keep this in perspective, think about all of the language that Juno hears around him all day long. Only a few of those words have meaning. If I look at my husband and say, “Pinikimunt,” he’s going to look at me like I’ve lost my mind. He has never heard that particular word before and he’s not going to have a clue that I’m telling him to take out the trash. It’s the same with our dogs. If we start saying some word or other that the dog has never heard before he’s not going to associate that word with much of anything. It’s just some more noise. If, on the other hand, we have a nice solid and reliable behavior and then associate a word with that behavior through many repetitions the word will come to be associated with that particular behavior and eventually it will come to be the cue to perform the behavior.
We’re using this nice precise click sound to mark the exact behavior that we want to teach our dogs. Don’t muddy the waters by introducing the cue word before you have the behavior good and solid.
*Excerpt from “Click for Joy,”***A final, finished behavior is often very complex. Not only may be the behavior itself be quite complicated, but there are additional elements such as duration, distance, generalization to different locations, and proofing against a variety of distractions.
Comment: : Okay, let’s look at the behavior we are working on…teaching our dog to touch our hand or a target. The behavior itself is not complicated. Fetch, on the other hand, is an example of a complicated behavior. Simple behaviors like “touch” do get more complicated when we add other elements to the behavior though. In her article, Melissa mentions elements such as generalization to different locations, duration, distance, , and proofing against a variety of distractions. Let’s take a closer look at generalization first. Will Juno touch your hand if it is above his head, below his head, off to the right, off to the left, while you are sitting, lying down or standing up? Will Juno touch your hand in the bathroom, kitchen, living room, in the back yard, on the front step, at the bus stop, at the grocery store, at work, at school, or any other place you visit? Just because Juno may be nailing your hand/target at home in the bathroom doesn’t mean that he will know what you want him to do when you step into another room or otherwise change the location. Likewise, if Juno has only been touching your hand/target in one position, you will want to help him generalize the behavior by holding your hand/target in all different positions, etc. For instance, when I trained Payton to hold her muzzle in my hand, I did several repetitions of hold and then lowered my hand to rest on the floor. She had no idea what I wanted her to do. I had been extending my upturned palm at her chin level so she could rest her muzzle in my hand. When I dropped my hand down two feet, she started offering other behaviors that she already knew…reached for a nearby door handle, kissed my cheek, etc. She knew I wanted her to do something but didn’t know what. Why didn’t Payton know I wanted her to touch my hand? Because I had jumped too far, too fast. I made too big of a step from one point to the next. My hand looked totally different to her when it was resting on the floor. The behavior fell apart so I backed up to holding my palm at her chin level, C/T a few times, and then lowered my hand several inches at a time instead of immediately dropping it all the way to the floor in one giant step. By taking smaller steps, I was able to gradually lower my hand to the floor without losing the behavior in the process. When we break down a behavior into smaller increments, we can C/T more often thereby increasing the rate of reinforcement. This helps to keep our dogs interest in the game. Generalization is just one element, there’s lots more to go but let’s stop here for now and see if folks have any questions or want to share some of their experiences with us.
Hopefully, everyone has their dogs touching their hand or a target now. The next step is to begin to generalize the behavior. This means that we’re going to be asking Juno to perform the touching behavior with variations.
The idea is that we want to get the touching solid enough that Juno will do it no matter where the target is located.
In the same place where you’ve been teaching Juno to touch your hand or the target begin to offer your hand in different locations. Left, right, down, and up. Vary the placement of your hand in every way you can think of. If Juno shows any hesitation at all, go back to holding your hand at the same place at which you’ve been holding it so far. Repeat this through several short training periods spaced out through the day if possible.
Once you have the touching behavior very reliable in the place where you’ve been working up until now you can begin to move around the house. When you move to a different location try to make all of the other elements the same as you did in your first location. For example, if you were sitting in a chair in location 1 do the same in the next location. If you were sitting on the floor, then sit on the floor, etc. Remember that we’re trying to set up our dogs for success. Change only one element of the task at a time and your dog will be more likely to continue to perform the touch behavior reliably.
Proofing against distraction is yet another element that we’ll add later.
For now, just make sure that each new place into which you move is free of distractions.
I just had a very powerful lesson on this subject. I’ve been working with Lukas, the puppy,on touching my hand and keeping his nose on my hand as I walk along. He had gotten pretty darn good at this when I was on the back porch or in the pen. This morning I tried taking him out on the front porch instead. The front porch is a lot bigger and there are lots of herbs, ferns, and other plants. In addition, Jim was out in the yard harvesting some vegetables. I had just had Lukas walking nicely by my side on the back porch so I tried the same thing on the front porch. It was a complete fiasco. Lukas was way more interested in checking out the plants and sticking his head through the railing to see what Jim was doing. So, I had to backtrack and go back to a skill that I was certain he could handle. I sat down on the porch and just played the touching game by holding out my hand several times right in front of him. I did half a dozen rapid c/t with my hand in that position and then began to vary the location of my hand. He did that game perfectly so I called a halt.
I made the mistake of adding two elements at once. I changed locations and there were distractions around. Later in the afternoon I took Lukas back to the front porch and just had him touch my hand in all sorts of positions.
Again, he did fine. Still later, out to the front porch we went again. I asked Lukas to touch my hand a few times just to make sure that he was paying attention. Then I walked back and forth on the porch holding my hand down by my side for Lukas to touch as we walked. He did perfectly fine.
Once I controlled the circumstances better I got exactly the behavior I was looking for.
And, now, a word on naming behaviors. If I had already named this behavior and used the name when he was so distracted that there was no way he was going to respond correctly I would have been undoing the work I had done so far. The more frequently Lukas heard that command, “Heel,” without associating it with the behavior I’m trying to teach, the more difficulty I’m going to have getting a good solid response to the command when I get around to naming it.
Good luck, and let us know how it goes.
FYI. The following post was written by Debi Davis, a clicker trainer who trains service animals. She is responding to a woman whose dog has been traditionally trained and the dog does not want to try anything new unless it is commanded to do so.
Teach your dog to be creative http://www.clickersolutions.com/articles/2002/creativity.htm Teach Your Dog to Be Creative Hope you don’t mind me jumping in and offering an idea to try with Jack, who prefers being lured and isn’t anxious to problem solve. Smart boy! I have a Pap like this, and of course, I realize I created the reliance on my telling him what to do rather than allowing him to solve the problem. It was easier for both of us, I guess. But, in order to surge forward, I had to go back, start at square one and teach him to search out creative options to make me click and treat him. A big help was in finding the absolute BEST reinforcements for this dog, the sumptuous treats he would die to steal off the counter, if only he was tall enough to get there. Then, I picked my times to teach creative problem solving carefully: I wanted him MOTIVATED, and a bit hungry. Early mornings were ideal: he was full of energy, hungry, and willing to push through his frustration walls a bit more than when his tummy wasn’t growling, or when he was more interested in watching the cat scratch in the litter box. Next, I picked the most boring, quiet environment I could find—THE BATHROOM! And I sat on the throne and opened up a book, read a few pages.
The dog circled in boredom, flopped down, perked his ears up each time I turned a page, and tried to stare a hole in my forehead. Yep, it was time to begin. I grabbed my clicker and sumptuous treats: Grueyere cheese covered chicken shreddings, and pitched a hairbrush on the floor. His eyes followed, and at that moment, I clicked and gave a treat. He stood watching me, staring a hole in my head. I picked up the hairbrush from the floor and dropped it again, and this time he bent his head down to sniff at it. I caught the sniff at the moment it happened, clicked and treated. He started to stare a hole in my head, but after 3 seconds of watching my eyes back on the book pages, ignoring him, (but watching CLOSELY from my peripheral vision) he dipped his head and I clicked just as his nose touched the hairbrush, and delivered the treat. He stood staring a hole in my head, but within 2 seconds, dipped his head again, and again, I caught his nose touching the brush, clicked and treated. He immediately dipped his head again and touched his nose to the brush.
CLICK and treat. No staring, just immediate dipping of head right after taking the treat, which I again clicked and treated. Can you see the pace picking up here? He’s figured out what will make me click and treat him, and now he’s eager to do the behavior he suspects will net him the booty. As he swallows the treat this time, his head is already moving down, and I click and treat again, then again, then again, keeping that reinforcement verrrrrry fast and rapid. He has no time to get bored, to get frustrated. I kept the behavior simple, let him choose with just a bit of gentle direction at first (dropping the brush to attract his attention). By the time he figured out the game, he was so engaged—so motivated, it was like watching pistons pump up and down. It was his first experience figuring out how to solve a problem—making me click and feed him. And he decided it was a pretty good game. Each time I made a bathroom trip (I swear, I’m gonna write a book called “Training in the Loo!”) he came with me. I never offered the same interactive item twice. There’s so much in the bathroom to grab at any one time. I ripped a postit note off the pad and stuck it on the wall, very low. He looked at it and I clicked and treated. He looked at it again and I clicked and treated.
Soon we were doing speed drills and he figured out the game was swiveling his head back and forth as fast as he could go. I was thinking a paw thwack, but heck, I got this nice head swivel, so I took it, and continued shaping, going faster and faster so he wouldn’t have a second to get bored. Did it work? Yup! It was the beginning of true problem solving for this boy, and a final break away from that lure that plagued us forever. Also, we had to move past bribery. I’d be so subtle in my bribery, but the more subtle I tried to get, the more discriminating this boy became. I had to get the treats off of me, and introduce other reinforcers he found rewarding: his favorite non-food motivator? Tugging games. The new game of creative problem solving then shifted to “something to do with the tug toy.” I let him choose something creative, which was moving from growling, hard tugs to a release when I pursed my lips in a kiss. He just let go one time accidently when I’d pursed my lips and made a whispered kiss sound, and I clicked at that moment. By then, he knew how to play the game and immediately latched onto this new cue signal for release. The quicker he released, the quicker he got the tug rope back. We kept working rapidly, lots and lots of fast nonstop reinforcements for about 2 minutes. We took the game on the road, into the living room, the bedroom, kitchen and finally outside, with more distractions. We used tons of items, and reinforced anything creative he offered: paw thwacks, lying down on an object, pushing objects with his nose, a paw thwack and a bark, whatever. By keeping the sessions VERRRRRY short, breaking them up all through the day, keeping the reinforcement rate very high, and stopping before he was bored, frustrated or tired–he was able to learn the joy of creative problem solving. I knew he was on the right track when my wheelchair rolled away one day in the house, and I cued him to retrieve it. En route bringing it back to me, a wheel caught on the table leg and he couldn’t move it any further. He became instantly frustrated, went into a down position, turned his head and tried to will me to figure it out for him. I just ate the cheesey chicken and ignored him. He let out a “roooo!” and tried pulling right, left, pushing with his paws, and still the chair remained immobilized. Again, he looked at me, hoping I’d “rescue” him in some way, and again I ate the goodie and ignored him. Finally, he let out a “ROO-ROOO—ROOOOO!” backed up, spun around, ran to the other end of the room, charged toward the chair and jumped up into the seat, effectively dislodging the wheel, and was then able to bring it to me the rest of the way. I gave him all the chicken. It can be a challenging thing to teach a dog used to being directed, to start creative problem solving. Even if they have no background of aversives being used, they do get lazy and find the most efficient way to have their needs met. If walking away and giving up gets them the booty, then of course, they are going to repeat that behavior. They are real efficiency experts, in that way! But, if they learn to find the JOY of problem solving, learning can take a totally new curve, and give them tons more satisfaction than they ever had while waiting for mom to make all the decisions. Go for it, Sandy! Jack will love you for it. And life will never be boring again.