|
Technique Challenge: Sit "Exerpted from
The Good Dog! Training Manual", copyright Janet Smith Training Behaviors Any behavior that your dog can physically perform can be "trained" and put on cue/command. The process is the same no matter what behavior you are trying to teach. The Steps are:
Step one: Get the behavior*. Have a clear picture in you mind of the "finished product". A behavior that occurs naturally...a sit, bark, etc. can be marked and reinforced (click and treat). This is called "capturing behaviors" and is simply a matter of marking and rewarding (c/t) the behavior. More typically behaviors are targeted or lured...for example a cookie held over a dog's head will cause the dog to sit. *at step one there is NO command/cue for the behavior. Step two: In step two you reward the behavior. The click should occur the instant the behavior does...followed by reinforcement/reward. Think of your clicker as a camera photographing the behavior the moment it occurs. Generally one will stay at step two for a minimum of 20 repetitions. prompt behavior-------click----------treat Do not go on to step 3 until the behavior is occuring on a regular basis, every 3 seconds or so, and the dog is not offering any other behaviors. Step Three: Attach verbal cue to the behavior. Choose a cue/command for the behavior (preferably a short single word) say the word just before/as the behavior is occuring. as dog sits-----say sit-----sit-----click------treat Remain at this stage until the verbal cue causes the behavior to occur...without any extra body language or food prompts-you may need to repeat the command at this stage which is fine, you will go to step four when the behavior is occurring on a regular basis with 3 commands or less. Step Four: Once the behavior is occurring on 3 commands or less slowly raise your criteria to 2 commands or less...if the dog does not respond within 2 commands use an LRS (least reinforcing stimulus). A LRS is simply communication which tells the dog "that incorrect/non response will not be reinforced". You move to a slightly different spot and cue/command the behavior again. Correct response (c/t), incorrect response (LRS). *If your dog repeatedly fails to give you the correct response he's telling you that you've raised the criteria too quickly...go back in the training steps to get him successful again. Remain at step four until your dog quickly responds to one command/cue. Step five: is getting the behavior under stimulus control in a variety of settings. You've taken your show "on the road" so to speak. Your dog will respond to the first command no matter what he is doing no matter where he is! How to get stimulus control:
CONGRATULATIONS YOU'VE TRAINED YOUR FIRST BEHAVIOR! Janet Smith
I like to toss the treat away so the puppy has to run to get it. This brings the pup out of the sit, rewards it AND gives it a great opportunity to return and offer another sit. I suggest clients feed a few meals like this... offering a few kibbles for each sit. As the pup gets the idea that sitting is fun I will pause for a bit before the click and tossing the treat... we are now working on a silent sit-stay (extending the time by nanoseconds sometimes :-) I implore clients to refrain from attaching the cue word for as long as they possible can. The longer they wait the more accurate they are at reading when their dog will sit. Their accuracy in predicting the behavior strengthens the word association for the dog. Philosophy: I believe that the one behavior that should be taught better than all the rest is "sit". Puppies and adolescent dogs have a hard time because they do not have the maturity or finesse to engage their brains AND their bodies at the same time. This leaves them in an "either - or" situation. When a dog sits, its body has effectively been disengaged so its brain can now turn "on". I like to teach "sit" as the default position. If my dog can't figure out what to do in a given situation he sits. If we are at a door or a gate I simply wait for the sit before we go through. To get into the car he must sit. To get out of the car he must sit, To get me to lower the food or water bowl to the ground he must sit. Occasionally I'll ask for the behavior but the vast majority of the time I do not. He offers a sit and gets what he wants. If an adolescent dog is about to or has already made a bad choice of behaviors I find that a strong "sit" cue may give him an alternative (and better) behavior which I can reward. Laura Van Dyne
Teaching Sit
Don Hanson
Two methods depending on the situation and the trainer. One, wait for the dog to sit click and treat, repeat. Change positions relative to the dog and areas as the dog progresses. Begin to name it when the dog is consistent. Sounds really easy right? It is. When the dog has been taught a couple of other stunts mix 'em up so that the dog has to pay attention to the cue to win the click. Two, holding food in hand raise it over the dogs eyebrows so that their but drops in order to keep eye contact with the treat, say sit as the dog is in the process of placing bottom on the ground. Repeat until the dog is familiar with what earns the treat. Fade lure into a slight hand gesture, begin to precede the hand gesture with sit cue. If the dog takes too long then oops and move on denying chance to get a treat. Cynthia Wunder
Here is how I am now working to reinforce "SIT". I get the dogs excited initally by chopping up treats and ignoring them. Then I put the treats on a high table and sit in my chair, hands folded. Because my dogs are little, they have to look UP to get my attention. When they do this, they normally sit because of the strain on their necks to keep looking up so high. The moment the dog's rear hits the floor, I click and deliver the treat. I then move my position, move forward a few inches or backwards--making the dog change position and stand again. And again, I sit quietly with my hands folded and when the butt hits the floor, I click and feed. I work hard to keep the momentum up, to keep moving around the floor, in a sort of dance so that the dog has many opportunities for quick reinforcement. I use heavy constant reinforcement, upping the criteria as I go, to new rooms, to the porch, to outside with limited distractions, to outside with mega-distractions. I do not even attempt to put a cue or signal on this behavior until the dog is freely offering it to me every time I sit quietly waiting for a behavior to be offered. I want to establish the "sit" behvior as a default behavior, as I have service dogs. I want them to first offer me a "sit" with attention before any other behavior. I am not critical about the TYPE Of sit I get at this point. But I do start refining which Sits I will reinforce, once the dog has the idea that the act of sitting will be reinforced, and do not keep reinforcing sloppy sits. I have lured the sit, and this has actually worked out splendidly, as well. I'm just finding that I have a bit stronger behavior by capturing than by luring, but it could be that becasue I use so much inadvertant cueing, that my dogs are constantly reading my body language getting mixed signals. I realize I have a weakness in this area, and constantly strive to reduce the unwanted cueing my body seems to do without my bidding! So free shaping seems to work the best for me. I put the cue on the behavior once I'm getting the TYPE of sits I want about 90 percent of the time, and in many environments. I do this by watching my dog carefully, giving off the pre-cue (my sitting quietly waiting), then just as my dog begins to sit, I say the cue word "sit" and click as the rear end hits the floor. But, I also wean them off the verbal cue as soon as possible, because I want them to do automatic sits whenever possible, giving me attention. By linking the attention behavior with capturing the sit behavior, it's made the sit a very nice default to have. Debi Davis
Before I begin to teach the sit I spend a few minutes conditioning the dog to the reward marker -- either the Click! or a verbal "Yes!" I do this by following the Click! or "Yes" with a treat a half-dozen to a dozen times. By doing this I also engage the dog's attention. Once he learns that I'm a little treat machine he usually has little interest in anything else! Most dogs will start offering sits during this conditioning process. I encourage this with my body language -- standing up straight and holding the treat up toward my chest. If the dog is offering sits, I begin clicking only for sits, and continue to reward for every Click! I will back up to encourage the dog to follow me, and click and reward when I stop and he sits. When I know that he will volunteer sits as I move away and stop, I add the verbal cue *as he does the behavior.* That is, just as his rump touches the ground I say "Sit!", then Click! and reward. I will gradually -- over 6 to 24 repetitions -- begin giving the "Sit" cue sooner, until I am asking for the sit *before* he offers the behavior. Then I start to fade my body language; hiding the hand with the treat behind my back as I ask for the "Sit." If necessary I will bring the treat back out to cue the dog to sit if doesn't respond to the verbal cue alone, but this is rarely necessary. A very few dogs need a more deliberate luring of the "Sit." In these cases I will let the dog nibble at the treat, then move it back over his head so that he sits in order to watch the treat more easily as it moves above him. If the dog insists on backing up to watch the treat I will back him into a corner and keep trying to lure the sit. Pat Miller | Training Articles Contents || Site Home | Copyright of all posts is the property of the original author. Please obtain permission from the original author before copying, quoting, or forwarding. List and Site Owner: Melissa Alexander, mca @ clickersolutions.com |