Training Techniques: There’s more than one

The following page outlines different possible techniques used in dog training besides the commonly known training methods such as Positive Reinforcement. After each description will be an example and the common uses for that technique.

Probably the most commonly known training method is Positive and Negative Reinforcement, however those are only two amongst many different techniques. One shouldn’t limit themselves to only using one technique or selected few, doing so limits the ability to train a dog as an individual instead of adopting and mending certain techniques to benefit a specific dog. Positive techniques, which is to say methods that don’t involve adverse stimuli, are usually the most effective because they don’t stress the dog which would inhibit his/her learning abilities. This doesn’t mean that techniques using adverse stimuli don’t work or that they should be shunned, in fact some dogs may only learn this way but these kinds of dogs are not as common as many think. Also negative techniques are usually require real expertise lest the action has unwanted effects.

Positive Reinforcement (R+) (Operant)

Definition: Adding a positive stimulus contingent on a behavior in order to increase the probability of that behavior happening again.

Example: A dog sits when the owner gives a cue and when the dog sits the owner gives the dog a treat.

Uses: R+ is usually used after non operant techniques such as luring in order to better reinforce the connection between action and cue.

Negative Reinforcement (R-) (Operant)

Definition: The removal of an adverse stimulus contingent on a behavior in order to increase the probability of that behavior happening again.

Example: In order to teach dumbbell retrieve some people will use a tool called an ear pinch. The pinch is placed on the dog’s ear causing some discomfort and is removed the second the dog grabs the dumbbell, this teaches the dog that he can avoid the adverse stimulus if he retrieves the dumbbell.

Uses: R- is used to teach a dog that he/she can avoid discomfort by practicing certain behaviors.

Positive Punishment (P+) (Operant)

Definition: Adding an adverse stimulus in order to decrease an unwanted behavior.

Example: A dog owners gives a quick leash correction when the dog starts to pull on the lead.

Uses: P+ is often used to teach a dog that a certain behavior is wrong and will be promptly followed by a correction, thus theoretically eliminating the behavior.

Negative Punishment (P-) (Operant)

Definition: The removal of a pleasant or reinforcing stimulus in order to decrease a behavior.

Example: A dog that seeks attention jumps on visitors and the owners when they arrive home, to decrease the behavior the owners don’t give the dog any attention.

Uses: P- is often used to extinguish problem behaviors in an effective and non confrontational manner.

While the above mentioned are the most commonly known training techniques they are quite often misunderstood. They are usually called ‘the four behavior consequence relations’ because they are four possible outcomes to the way a dog is acting. If it hasn’t been noted already, it is important to see that in dog training terminology  the use of the words ‘positive’ and ‘negative’ are used in the sense that something is added or subtracted. Many people erroneously think that positive means something good while negative is something bad. When looking at Positive Reinforcement or Negative Punishment take note of how those techniques simply involves adding or removing a stimulus whether that stimulus is pleasant or adverse. Another important thing to keep in mind is that Reinforcement (whether positive or negative) are done with the goal of increasing a behavior, and that Punishments (whether positive or negative) have the goal of decreasing a behavior.

While the use of negative reinforcement and positive punishment aren’t incorrect to use, they can be particularly hard to implement. Timing is critical so that the negative connection isn’t attributed to something other than the targeted behavior or even worse yet, the owner. It has been proven that the majority of dogs learn best with R+ and P- techniques. Studies indicate that many dogs may react more aggressively if adverse techniques are used on them (Click here to read about the study). This being said, it is best to leave the adverse techniques to advance trainers or professionals.

Luring (Non-Operant)

Definition: The use of an enticing item to get a dog to act or move a certain way and slowly connect that action with a cue.

Example: To teach a dog what the word “lay” means the owner takes a food treat while the dog is sitting down and slowly moves the treat down towards the ground and away from the dog. The natural instinct of the dog is to follow it so it starts to lay down, as he/she does this the owner says “Lay” and praises.

Uses: Luring if used to teach a dog what a specific cue (verbal or hand motion) is asking of him/her.

Modeling (Physical Modeling) (Non-Operant)

Definition: Actually pushing or pulling a part of the dog’s body in order to get him/her to do an action, as the dog does the action the owner gives the cue.

Example: To get a dog to sit, the owner gently pushes down on the dog’s rear to get him/her to sit. Once the dog sits down the owner says “Sit” and praises.

Uses: Modeling is used to teach a dog what a specific cue is asking of him/her, usually used for dogs who don’t do well with luring.

The non operant techniques described above are usually used at the beginning of teaching a new command or action. They are used to communicate what exactly a command is expecting of a dog. A likely comparison to understand how non operant techniques work is if someone were to talk to you in a foreign language asking you to do something. Obviously it’d be hard to know what they want without clues. However it is much different if that same person told you what to do while at the same time doing the action himself or guiding your hands to do it, even if you didn‘t comprehend the language you‘d still get the idea of what the other person wants. Next time you hear the same command you might be able to recognize it and do the action. It is the same with dogs, after all they don’t speak English; this example also helps to illustrate why repetition is very important at the beginning. However non operant techniques don’t take much thinking and are completely useless if the lure or modeling isn’t phased out. An important thing to note about physical modeling is that it has to be gentle to avoid hurting the dog, while also keeping in mind that some dogs don’t react well with modeling because they meet forceful pushing with resistance. Lures shouldn’t be confused with bribes, a lure is a technique used to show a dog what a cue means while a bribe is a treat produced before the behavior is done and usually is the only way to get a dog to do an action (such as coming when called). The biggest difference between the two is that lures should be phased out during the first training session, while bribes are often used long after; in essence a lure may become a bribe if it isn’t phased out early on in training.

Extinction

Definition: Ending a behavior by removing anything that is reinforcing it.

Example: Dog barks for scraps at the table, the owner ignore the barks and the dog learns that barking gets him nothing.

Uses: To eliminate unacceptable behaviors.

Note: Quite often when someone is using P- they are extinguishing a behavior. Extinction is not a technique in itself but rather a term used to label what is done to an unwanted behavior.

Note: Related to extinction is something called an extinction burst which is a period of time when the dog will increase the undesirable behavior in a last attempt to get what it wants. The burst may signify that the unwanted behavior may soon stop. Never give in when the dog is in the extinction burst ‘phase’ because this only teaches him/her that if she tries hard enough he/she will get what they want, making the problem behavior all the harder to extinguish.

Flooding

Definition: Presenting the dog with the fear causing stimulus forcefully so that it no longer elicits a negative reaction.

Example: A dog is extremely scared of toy cars, so in order to stop the fear reaction the dog is put in a room full of toy cars.

Uses: To eliminate aggressive and fearful reactions.

Note: This particular technique has come under much criticism because the fear response doesn’t seem to be solved but rather suppressed which means that the fear reaction may occur again. Techniques like this also puts the dog under a great deal of stress and may cause an even greater fear response in the future.

CounterConditioning

Definition: Eliminating a behavior by getting the dog to obey another command that can’t be done at the same time with the unwanted behavior.

Example: A dog lunges at other dogs so you say “Watch Me!” when walking near another dog. 

Uses: To substitute unruly behavior with an acceptable one or to distract the dog from a stimulus that will elicit an undesirable behavior.

Systematic Desensitization

Definition: Gradually exposing a fear or aggression causing stimulus, always staying below the response threshold, and slowly getting closer to the stimulus without evoking a reaction from the dog.

Example: A dog is scared of thunder. The owner buys a recording of thunder and plays it at low volumes, he/she rewards the dog for good behavior, slowly the volume is turned up and if the dog continues to act well the owner continues praising. If at any time when the owner turns up the volumes slight, the dog reacts adversely then they’ve reached the fear threshold and the owner turns down the volume a little. They continue until the dog no longer reacts fearfully even though the volume of the thunder is turned up really loud.

Uses: Desensitizes dogs so that they no longer react in a bad way to certain stimuli. This technique could be used with things that cause aggression, fear, prey drive, etc.

Note: Systematic Desensitization is usually used in conjunction with counterconditioning to gain the best results; the use of both techniques is then denoted CCDS.

Note: Sometimes CCDS might not work because the stimulus used to desensitize the dog (e.g. the thunder recording) may not be the same as the real thing to the dog. Should this be the case, a myriad of solutions may be needed to successfully cure the problem behavior such as stress jackets or medication implemented while using CCDS.

The techniques mentioned above are some of the more commonly known techniques yet there may be more that use different terms but still stem from the original operant/non-operant principles. The best way to train a dog is to assess its particular needs and mode of learning. Being a good trainer doesn’t mean sticking to a certain technique religiously, but rather  being able to determine which techniques may be used to help the individual dog learn in the most efficient manner.

-Steven L.

If you have any questions or comments about the content above or would like to discuss other dog/pet related material, feel free to visit K9domain's K9Academy Dog Message Board, registration is free and easy! There you will find knowledgeable members that are willing to help you with any problem you may have.

References and Further Reading

Horwitz, Debra F. Counterconditioning and Desensitization.< http://www.vin.com/VINDBPub/SearchPB/Proceedings/PR05000/PR00312.htm >

Plonsky, Mark. Confusing Consequences: A Brief Introduction to Operant Conditioning.< http://www.uwsp.edu/psych/dog/LA/DrP2.htm >

Pryor, Karen. The Poisoned Cue: Positive and Negative Discriminative Stimuli.< http://www.clickertraining.com/node/164 >

Robin. Capturing, Shaping, Modeling, Luring.< http://ccscdogtalk.blogspot.com/2008/01/capturing-shaping-modeling-luring.html >

Toft, Catherine. Examples -- applying operant training.< http://r-plusdogtraining.info/examples.htm >