Chapter 25: Reinforcers, Corrections & Reworking Errors
Before we go into corrections let’s look at reinforcers in general.
Positive reinforcers try to set the dog up for success by rewarding increments of behavior, therefore gradually forming a sequence chain or pattern. The trick of course in guide work is where the balance lies. Food is considered aprimary reinforcer to a dog, so it's very high on its list. Primary reinforcers when coupled with secondary reinforcerslike verbal and physical praise can help build a strong response. These secondary reinforcers can then become primary ones as well. This is of great help to the dog and handler during the transition period of one to six months. It helps the dog accept your praise and understand it as being sincere. The food is consistent and clear to the dog and does not change from person to person. Initially when new behaviors are shaped the food is used 100 percent. Once behaviors are reliable then a variable schedule can begin, mixing up physical, verbal and food reward.
Some handlers drop food reward when they leave the school which creates confusion and discomfort for their dog. Remember it is not bribing or taking the place of your physical touch or praise, it actually helps to accelerate and bring it to a maximum level. Once you have been working with a dog for awhile and you as a team are very comfortable, praise and correction is understood from the dog. At that point you may not need food as frequently but it still can help keep your dog fresh.Dogs that are trained through operant conditioning are not afraid to problem solve.If a dog only gets negative reinforcers his repertoire will be somewhat limited. This doesn’t mean to use only positive reinforcers. Use the two together to strike a fair and reasonable balance. The most important thing to ask is “Am I being fair and consistent?”
Remember that event markers let the dog know it did well whether you use your voice or the clicker. The sound of the click or a verbal “yes”paired with food and then followed by supportive praise strengthens the behavior. Food reward is used for hand targeting, helping with fear based issues and sometimes with high levels of distraction.We use food in obedience to strengthen the right behaviors.
When they respond to the first command they receive food.After the food reward is given physical praise maythen be given simultaneously with the word “good boy or good girl.”The word“Good” is linked with petting and verbal praise; not with food.This is done so as to prevent the fixation on the handler's face with an expectation of receiving food every time they hear the word “good”.Having to give the command more than once only gets praise.
Non event markers, like not now, not today, are just neutral phrases to let the dog know to make another choice; they are not a correction or praise. It can be quite effective when searching for a certain door, or just letting the dog know you don't need that particular escalator.
Negative reinforcers are ones that stop when a desired behavior is achieved. If pressure is applied in the collar for the dog to go forward and once he starts moving it’s released, then that would be considered a negative reinforcer. It goes away when the response occurs, and praise is added in the form of verbal or physical or a combination of both.
Generally one should use less negative reinforcement for actual guide work errors.Correctingtoo firmly for guide work errors will diminish the dog's drive and you will have to spend some time building the dog back up again. Physical collar correction should be saved for non negotiable areas. For example; eating pizza off the floor or chasing after an animal is a non negotiable and unwanted behavior.
Corrections should change the dog's behavior not their demeanor. It should give the dog the answer, not that, but this, so you get a dog that is trying versus a dog that shuts down or gives up. Balance is so important! Employees that get even small amounts of praise from their boss will perform at a much higher rate and furthermore will want to continue to perform well.
Corrections for actual guide work errors should remain at the level that reaches the dog.Some dogs require only a light correction in order to be reached; for some a verbal no is enough. Discourage what you don't want, reinforce heavily for what you do want.Correct when needed, not out of frustration and don't correct more then once for a given infraction. Why; because multiple corrections create reactivity in the dog as well as confusion.
Avoid raising your voice with commands as it weakens them and you want your dog to be responsive, not reactive. Use food and praise to create, build, and strengthen behaviors.
You want the dogs to try harder, to work harder but know that a reward is coming.If you reward good performance, you will get good work. If you reward mediocre performance, you will get mediocre work. In the end, being fair and consistent will not only strengthen the bond between you and your dog, but create a good working team.
Corrections in depth:
If you have given a command to your dog and your dog hasn't responded to the command a correction will follow. A correction is language that the dog has been conditioned to and understands and handlers need to use this language when working with a dog. This will insure that your dogs response remains consistent. Remember that corrections may be used in areas that are considered non negotiable. For example chasing after animals and eating food off the floor. In these cases a correction is necessary because we will never ask the dog to chase after animals or eat food off of the floor.
However the handler will always need the dog to get them to a curb, stair or door. Handlers should avoid firmer corrections for stepping slightly over a down curb or lightly brushing an obstacle. Correcting too firmly in these areas will diminish drive in the dog.
A good leash correction has several essential characteristics. It must be given with some slack in the leash initially, so you are able to snap and release. The snap is what makes up a correction; pulling on the leash will accomplish nothing except to make the dog pull harder against you. A correction given properly does not harm the dog in any way, nor does it cause pain to the dog. The amount of force given varies from dog to dog, and even from situation to situation with a single dog. A good rule of thumb is that the correction should be strong enough to create the desired effect the first time it is given, but not so strong that it harms or intimidates the dog.A correction should not be considered a punishment, but an action to turn an incorrect response into a correct one. We will instruct you on the particular type of correction that will be most effective for your particular dog.
Specific corrections for the various non-harness commands are listed below:
•Heel: correction is a snap on the leash toward your left side
•Down: correction is a snap straight down toward the ground
• Sit: correction is a snap straight up from the top of the dog's neck
• Stay: this correction is more complex. If the dog sits up, stands up, or lies down after being told to stay in one position, but does not move from the spot he was told to stay, you should return to the dog and return it to its original position using a verbal command and a leash correction. If the dog breaks position and starts to move away from the spot it was told to stay, you will have to return it to the original place and position
There are also leash corrections for harness commands. They are as follows:
•Forward: correction is a short, light tug forward, given with the right hand.
You should not give this correction unless advised to by your instructor, and should be
very careful about giving it at home, because the dog may be refusing Forward for a good reason.
•Right: correction is a snap to the right, given with the right hand
• Hup-up: correction is a snap in the direction appropriate for the situation, given with the right hand
If, while a dog is working, it becomes so distracted that it fails to respond to one or two right-handed corrections and stops working to try to get to the distraction, you should drop the handle and administer a left-handed leash correction. We will demonstrate other corrections to you on an individual basis.
A harness check is a mild correction given with the harness handle. The correction is given by using your left hand to check or snap back with the handle.
There are several important things to remember when you give a correction:
*Never give a leash correction out of anger or frustration. A leash correction is just a way to show the dog that what it did was incorrect, and to remind it what the correct thing to do is.
*Never shout or raise your voice when giving a correction. Shouting a
command at your dog will not make it obey any faster.
*Never use any form of correction other than those taught here at GEB. It is NEVER acceptable to hit, kick, or scream at your dog.
*Always remember to praise when the dog resumes appropriate behavior after receiving a correction.
*Always remember that people are watching you when you work with your
dog. They get very upset when they see you doing what they may perceive as
abuse. If they see you follow a correction with praise, they are less inclined
to think that you were being abusive to your dog (and less inclined to call
GEB and complain).
*Concentrate on using more wrist action rather than swinging your arm. This keeps the correction quick and low to the body.
Guide work errors are a normal part of the guide work process. It is important to understand thaterrors can occur for a variety of reasons andthat many of them tend to be minor innature. Generally a newer team will experience more errors then an established team,particularly during the acclimation period of six months to a year. Guide work errors areusually a combination of both dog and handler errors as well as environmental influences.
Environments that are complex and have a considerable amount of traffic, pedestriansand loose animals can create problematic areas for guide dog handlers. Errors are morelikely to occur in new areas where neither the dog nor person is familiar or patterned to.Furthermore both members of the team are living, breathing, thinking individuals whichcan contribute to the success of a given workout or progression of workouts.
When guide work errors occur it is important for the handler to recognize them, quicklyidentify what has obstructed their travel and work to remedy it.Generally errors such as running curbs are errors that should be reworked.