Tag Archives: learning theory

Why Is Punishment So Hard To Get Right In Dog Training?

First off, The DogSmith does not advocate punishment, not when there are so many other more pleasant ways to decrease the intensity, duration or frequency of a problematic behavior.

In the scientific world a ‘punisher’ has to be an event that decreases the strength or likelihood of a behavior. So punishment does not have to be scary or intrusive or aversive. BUT punishment is so difficult to get right consistently and at the correct intensity that pet owning parents are best to stay away from using punishment to help them train their dogs.

If you are punishing your dog and after a couple of occasions the behavior has not improved then stop what you are doing and consult a professional dog trainer. If you continue on ineffectively applying punishment you will make the behavior worse and create other problematic behaviors as a result of your dog now being fearful or anxious about your actions towards them.

Punishment is very difficult to get right. Why? Well because:

  1. The effects of punishment are dependent on many variables. Like reinforcement, the degree of contingency between the punishing event and the behavior are an important variable. The level of behavior suppression achieved is determined by the degree of contingency.
  2. Contiguity, the interval between the behavior and the punishing consequence, is also a variable that affects punishment. The longer the delay is between the behavior and the punishment the slower the learning. A delayed punisher may not only suppress the desired behavior but also affect other behaviors. (Chance 2008 p213).
  3. Another variable that affects punishment is the intensity of the punisher. It is difficult and often impossible to know at what level the punishment will be effective (Chance 2008 p 213).
  4. Another variable that affects punishment is the presence of reinforcements that have maintained the behavior. The behavior will not be suppressed if the animal still gains reinforcements.  For punishment to be successful it helps if the animal can find an alternate way of obtaining the reinforcers (Chance 2008 p216).
  5. Other variables that affect punishment are an animal’s deprivation level and the qualitative features of a punisher.   Punishment has less affect when an animal is hungry if the reinforcement available is food. The consequence to the behavior has to be punishing for the animal, it must decrease the strength of the behavior (Chance 2008 p208).

So as you can see there are so many variables that affect punishment. In a real life situation and not in a laboratory it is very difficult to get each of these components right without creating ‘fall out’ behaviors. Badly timed punishment is ineffective and the wrong intensity of punishment can be harmful and very invasive to your pet.

You also need to consider before you use punishment;

Do  YOU do not want to be perceived as unsafe to your dog? Punishment will negatively impact your  dog-human relationship and bond.

Consult a professional dog trainer or dog behavior counselor if you have a dog behavior or training problem.


Operant Conditioning and Respondent Conditioning for Dog Trainers – Certified DogSmiths Have Science On Their Side.

Abstract

Certified and professional dog trainers have science on their side. Dog training methods used by educated and well versed dog trainers are based on sound scientific principles that really do work. A good people trainer can take the complex world of learning theory and put it into a methodology that a pet dog owner can use and achieve effective results from. The DogSmith is an organization that not only understands the principles of learning theory but has the ability to teach pet dog owners so they too can take advantage of a professional dog trainer’s tool-kit. This promotes effective, efficient and fun dog training in an environment free of harsh methods and aversives  and many other fall outs associated with harsh and in-humane training methods.

Is a hug fun for the dog?

At the DogSmith we accept that dogs learn in two key ways.

Task-conscious or acquisition learning.  The Pet Dog’s Environment

Acquisition learning is seen as going on all the time. It is ‘concrete, immediate and confined to a specific activity; it is not concerned with general principles’. Examples include much of the learning involved in and around the home.  Some have referred to this kind of learning as unconscious or implicit. Rogers (2003: 21), however, suggests that it might be better to speak of it as having a consciousness of the task. In other words the learner may not be conscious of learning (Rogers 2003).

Learning-conscious or formalized learning.  The Training Class

Formalized learning arises from the process of facilitating learning. It is ‘educative learning’ rather than the accumulation of experience. To this extent there is a consciousness of learning – subjects are aware that the task they are engaged in entails learning. ‘Learning itself is the task. What formalized learning does is to make learning more conscious in order to enhance it’ (Rogers 2003: 27). It involves guided episodes of learning.

When approached in this way it becomes clear that these contrasting ways of learning can appear in the same context. Both are present in the subjects living environment.  Because we, The DogSmith believe dog’s learn in two ways we found it necessary to develop our MTR approach to Dog Training supported by our ARRF philosophy.

ARRF © methodology

A = Active Involvement. Active involvement in the learning process is critical. When the student is actively participating, rather than passively observing, greater learning takes place.  This applies to both the dog and its owner.

R = Repetition. Newly acquired skills need to be repeated frequently in a variety of contexts to ensure they are robust. This means the skills you and your dog learn n will be effective in and around your home and out and about around town. Frequent re petition in various scenarios ensures the skill is truly owned so the student can not only generalize its behaviors in new situations but can also discriminate when appropriate.

R = Reward. Positive reinforcements, in the form of rewards for accomplishing skills successfully, are far more effective to ensure learning takes place. Rewards such as food and toys are quickly re placed by life rewards, such as attention and petting, when behaviors are under cue control.

F = Finite Objectives. Clearly defined and attainable objectives make it clear to student and instructor what is to be learned and taught. With clear objectives the student and instructor can easily recognize when a particular skill has been mastered and during the process we train – test -train to ensure our objectives are met.

ARRF evolved from our understanding and knowledge of respondent and operant conditioning and how we can utilize this scientific approach to training dogs. MTR, Management, Training and Relationship are the components that a dog owner has to understand to ensure their dog is receptive to learning, able to learn and managed throughout the learning process.

How Does Respondent Conditioning Take Place

Within an organism there are two types of reflexes, unconditioned reflexes and conditioned reflexes. An unconditioned reflex (UR) is unlearned and occurs unconditionally, whereas a conditioned reflex (CR) is acquired and considered impermanent.

An unconditioned reflex consists of an unconditioned stimulus (US) and an unconditioned response (UR).  An unconditioned stimulus is something that when presented evokes a natural, unconditioned, response,  such as blinking when air is pushed towards the eyelid or sweating when stressed or scared. Unconditioned reflexes are important for an animal’s survival.  Freeze dried liver offered to a dog is an example of a US and the dog drooling is an example of the resulting UR.

A conditioned reflex occurs when a conditioned stimulus (CS) creates a conditioned response (CR).  This is a learned response to a given set of conditions occurring in the environment.  Pavlov recognized that any stimulus could become a conditioned stimulus when paired repeatedly with an unconditioned stimulus.

Respondent conditioning takes place when an unconditioned stimulus that elicits an unconditioned response is repeatedly paired with a neutral stimulus. As a result of conditioning, the neutral stimulus becomes a conditioned stimulus that reliably elicits a conditioned response. Each single pairing is considered a trial. With respondent conditioning the presentation of the two stimuli, neutral and unconditioned, are presented regardless of the behavior the individual is exhibiting. The behavior elicited is a reflex response.

High order conditioning takes place when a well established conditioned stimulus is paired with a neutral stimulus to elicit a conditioned response. High order conditioning takes place in the absence of an unconditioned stimulus. With high order conditioning many more stimuli can come to elicit conditional responses not just those paired with an unconditioned stimulus, thus enhancing the ability of the animal to adapt and survive.  But high order conditioning also affects and influences many emotional reactions such as fear.

Operant Conditioning

There are four types of operant learning, defined as such because the behavior operates on the environment.  Two of the quadrants of operant conditioning strengthen behaviors, referred to as reinforcements. The other two of the operant conditioning quadrants weaken behavior, referred to as punishments. The quadrants are referred to as a negative reinforcement, positive reinforcement, negative punishment and positive punishment.  The terms positive and negative do not describe the consequence, they indicate whether a stimulus, has been added (positive) or subtracted (negative) to increase or weaken the preceding behavior.

Both positive and negative reinforcement increase the strength of the behavior due to its consequence.  With positive reinforcement the behavior is followed by the appearance of or an increase in the intensity of a stimulus. The stimulus is called a positive reinforcement as it is something the subject seeks out therefore it reinforcers the behavior that precedes it.  With negative reinforcement the behavior is strengthened by the subject’s ability to avoid or escape an aversive stimulus, thus negative reinforcement is sometimes referred to as escape-avoidance learning. An experience must have three characteristics to qualify as reinforcement.  The behavior must have a consequence, the behavior must increase in strength and the increase in strength must be a result of the consequence.

As behavior is the function of its consequences and whereas reinforcement strengthens the likelihood of a behavior then punishments reduce the strength of the behavior. Punishers are aversive are something a subject works to avoid. When an aversive event is added to a situation then positive punishment has taken place. Negative punishment subtracts something from the situation, like privileges, and is sometimes called penalty training.  Experiences must have three characteristics to qualify as punishment. First, the behavior must have a consequence, second the behavior must decrease in strength and finally the reduction in strength must be a result of the consequence.

To schedule an appointment with a DogSmith click here

Bibliography

Chance, P. (2008) Learning and Behavior, Wadsworth Cengage Learning


What is a Positive Dog Trainer – What does it mean?

Does The DogSmith say they are Positive dog trainers?

As it is used in dog training, the term “positive” can be very misleading and confusing to the pet owner. In the context of learning theory, the science behind how we all learn, the term positive means the addition of something. Positive reinforcement means you are giving the dog something it likes whereas Positive Punishment means you are giving the dog something it does not like. At The DogSmith we use the least aversive and minimally intrusive training methods possible. Dogs do not seek to avoid or escape the methods we use. So if you call that positive then yes, we are positive trainers. We NEVER inflict mental or physical harm on any dog. The DogSmith methods are educational, instructional, effective and efficient and most importantly they are user friendly. Friendly for the dog and friendly for the owner.