INTRODUCTION
"Principles of Conditioning" as they
appear in this material are not new, unique, nor secret techniques. These
conditioning principles are based on behavioral modification procedures used
not only by dog trainers, but trainers of other animals, and psychologist.
Failure to properly follow the conditioning principles outlined in this chapter
will prevent you from following a logical progression of training and problem
solving.
The conditioning principles outlined can be used for
patrol training, tracking, obedience, or any other task you may wish to train
the dog to perform. They are also used to eliminate unwanted behaviors as well. Novice
handlers and trainers should seek the help of experienced trainers or animal
behaviorist before attempting to correct serious or dangerous behaviors found
in some dogs.
To insure the team maintains proficiency and
reliability, supervisors should be trained in their responsibilities to the
team and how to support and maintain the team handlers can't do it all, nor
should the handler be expected to do it all. A handler should not always set up his or her training problems.
Likewise, the handler should not be the
one judging how well the dog performs. During the recording keeping section
of this course you will be instructed how to maintain proper records that will
allow you or anyone else to make an objective assessment of your performance.
MOTIVATION AND SENSATION
Motivation and sensation play a vital role in a
dog's behavior. A dog perceives its environment through its senses. Its eyes
cannot choose to do anything other than see; its ears cannot choose to do
anything other than hear; and its nose cannot choose to anything but smell.
After a sense organ is stimulated, a dog may be motivated to act A dog’s motivation is based on two concepts: needs and
drives.
Needs
The concept of need refers to an internal imbalance
that occurs when a dog is deprived, of something. A dog experiences needs at
all times and in varying intensities. The need for praise is a social need. A
dog's needs are most compelling when they are not fulfilled, causing a state of
deprivation.
Drives
The concept of drive refers to the attempt of a dog
to restore its state of balance by satisfying certain need. Drives arise from
needs.
BASIC NEEDS (DOG)
A dog responds to its environment to satisfy its
physiological (bodily) and psychological (mental) needs. Before a dog is able
to learn or perform a task, it is necessary for certain needs to be met
Oxygen
Breathing, to obtain oxygen, is the strongest need
for the dog. Dogs are panting animals. Understanding this need and what the dog
will do to meet this need plays an important role in the training and handling
of the dog.It is also considered a dog's strongest drive.
Water
Water must be provided to the dog in adequate
quantities to prevent thirst from interfering with learning or task
performance. It can never be used as a reward.
Food
Food must be supplied in adequate qualities to
prevent hunger from interfering with task performance. Food or treats can be
used as a rewards or motivators.
Prey-Kill
This need incorporates chase with manipulation. Manipulation, in the case of the dog, is biting and tearing.
Social Needs
Dogs have a need to be with a pack. A pack can
consist of either dogs or humans. A period of socialization between the dog and
handler is required in order for his/her praise to be meaningful to the dog as
a reward. This socialization process must be continued for praise to be
effective in maintaining the desired behavior.
Pain
The dog can detect different levels of physical and
emotional pain. The other basic senses are to alert the dog to the impending
onset of pain, which enables it to avoid pain. This is the method used to
control the dog by use of verbal and or physical correction in avoidance
training and the pressure of the choke chain in escape training. By combining
physical correction with vocal correction, the vocal correction becomes
emotionally painful to the dog.
BASIC NEEDS (STUDENT)
Humans have needs as well as the dog. We as
instructors are aware of these needs and take them into consideration during
this course. Students who understand these needs are less likely to experience
frustration during training which will help improve the quality of instruction.
Physiological Needs
Need to satisfy hunger and thirst: These needs are
met by allowing you to take breaks.Need to feel the
world is organized and predictable, safe, secure, and stable: These needs will
be met toward the end of your first week of training or the beginning of your
second week.
Psychological Needs
A sense of belonging and esteem are your next
important needs. Your need to belong has most likely been met. You belong to a
group called a class and this need should have been satisfied shortly after
meeting your classmates and instructors.
Your esteem needs come a little slower. This is new
to you and you are likely to question your abilities and well as the dog's. It is not uncommon for these feelings to last well
into the second week of training. About the middle of the second week of
training you will begin to move into the next step.
Self-Actualization Needs
Need for self-fulfillment: As you and your dog begin
to function as a team instead of individuals you will see yourself
accomplishing a goal.
SENSORY SYSTEM
A dog perceives its environment through the use of
its basic senses. The dog senses its environment in essentially the same way
that a human does. It is generally the human's lack of knowledge of the dog's
sensory system that cause much of the
misunderstandings in dog training and utilization. In the following information
we wi1l discuss working dogs capabilities and limitations. This, possibly more
than any other "training literature," will be invaluable to anyone
handling, training or supervising working dog activities.
Sensory Spectrum
To be knowledgeable of the sensory system of a
working dog you must first understand the sensory spectrum and how it works.
This spectrum applies to all senses in the sensory system.
Threshold
This is the intensity level at which a stimulus is
great enough to detect or create a "just noticeable difference" or
"JND." In the context of the measurement of sensation, the term
generally refers to the intensity level required to activate a particular sense
organ. Thresholds differ from sense to sense, situation to situation and dog to
dog.
Absolute Threshold (AT)
This is the minimal amount of stimulus, which
is needed by the dog to lead to an observable response (OR Orienting Response). This absolute
threshold is a statistical average of a stimulus that can be detected fifty
percent 50% of the time. This can be measured only if all outside conditions
are controlled (i.e., heat, humidity, etc.). Absolute threshold will vary from
dog to dog depending on the capabilities of each dog. It will also vary in the
same dog from one measurement to the next and from day to day.
Just Noticeable Difference (JND)/ Difference Threshold (DT)
JND is an internal preception of a change in a stimilus. It is the animal's ability to detect slight changes
or differences of stimuli pertaining to a specific sense. This ability to
detect slight changes is also known as a Difference Threshold (DT). The
illustration on the right shows how there is nothing for the dog to detect.
Then it detects something. Depending on how it perceives the location the dog may move into
more or it may move into less.
Orienting Response
Responses an organism uses to orients
itself toward the source of a stimulus. These orienting responses follow the
detection of a stimulus. They are often referred to as behavior changes. There
are a host of stimuli that may cause orienting responses it is for this reason most courts
have ruled orienting responses or changers of behavior are insufficient to
provide probable cause. Follow the above pictures you will see orienting
responses made by the dog as it locates source and goes to its final response.
Saturation Point
The dog can no longer distinguish a DT at the
saturation point. When the dog reaches the saturation point, it knows the
stimulus is present and should make a final response. However, the dog may not
pinpoint the source because of its inability to detect a higher DT.
Source
This is the point of origin of the stimulus. The dog
finds source by determining an absolute threshold and using a difference
threshold or just noticeable difference to reach it. When the dog encounters a
stimulus at its AT, he will locate a higher DT and
then still a higher one until he locates source. When the dog reaches source,
it will make the final response it was trained to perform.
FACTORS THAT AFFECT THE SENSORY SPECTURM
Every time a dog uses its senses to perform a task,
it must use the sensory spectrum. It must reach the absolute threshold of a
stimulus and follow the difference thresholds to the source or saturation
point. However, there are instances when the sensory spectrum will be affected
while the dog is being utilized.
Weather
Weather affects the spectrum because in working dogs
the weather conditions cannot be controlled. An increase or a decrease in
temperature, humidity, atmospheric pressure, wind etc., will cause a change in
the difference threshold. Remember, the absolute threshold is a statistical
average measured under laboratory conditions.
Threshold Shifts
Threshold shifts occur when the sensory system is
activated. This means that once the dog's sense has been activated a certain
recovery period is required before the dog will be able to detect a smaller amount of
odor, light, sound etc. The greater the amount of stimulus that the dog is
exposed to, the longer the recovery time is needed before smaller amounts can
again be detected.
Quantity Of Odor
When conducting training on tasks that use the dog's
sense of smell, you must vary the quantity of the odors during training.
Initially, you start with large quantities then reduce the amount.
Physical Barriers
Physical barriers sometimes prevent the dog from
working the Difference Threshold to source. The dog has the ability to detect a
higher concentration of a given stimulus but a barrier or object prevents him
from going to source.
SENSES
A dog perceives its environment using its basic
senses. Dogs sense and perceive their environment differently than humans.
Smell (Olfaction)
Dogs have the ability to identify one odor from
another. The concentration of odor gives information about the source of the
odor.
Hearing (Audition)
Dogs can determine the direction, origin and
sometimes source of sound. Dogs use level
changes and movement of source of sound to identify its origin.
Sight
The dog learns to react to or ignore visual cues in
the environment. Cues such as brightness, shape, size, and motion give the dog
needed information to learn required tasks.
Detector dogs often use sight to identify shapes and
objects where it has found the target odor in the past. This behavior can be
useful as a dog will automatically slow and examine an area more intently when
it encounters it. The use of sight can also be frustrating. Dogs will sometimes
try to break search pattern to go to where the target odor has been placed
before. Experience with the dog will help you learn when the dog breaks search
pattern due to odor or sight.
Touch
The amount of pressure in the handler’s touch can be
of great importance to the dog. Trainers use touch to soothe an excited dog, to
encourage a dog, or as a form of secondary reinforcement. Dogs use touch as
well. They gather information about an object by sniffing and touching it with
their nose. Likewise, they will use touch to interact with the handler.
Other basic senses warn the dog of the impending
onset of pain which enables it to avoid it or (if it can’t be avoided) to
minimize it. We use the dog’s sense of pain when we jerk on the choke chain. By
combining verbal corrections with physical corrections the verbal corrections
will become emotionally painful to the dog.
The dog senses small amounts of pressure through
sensitive hairs called tactile hairs. These hairs also enable the dog to sense
changes in wind. Light pressure, such as
petting praise, can be used as a reward. Receptors in the skin enable the dog
to sense heavy pressure such as tightening the choke chain. The speed and
amount of heavy pressure is a cue for the dog in escape or avoidance training.
Heavy pressure usually indicates a correction such as a jerk on the leash.
Equilibrium
The dog's sense of balance keeps it in an up-right
position. Dogs learn to ignore the sensory input to do certain task such as
rolling over.
This sense enables the dog to be aware of its body
position and movement at all times. Receptors within the muscles, tendons,
ligaments and joints send information to the dog’s brain. This sense is
important in training because the dog connects a reward with the last body
movement made before the reward is given.
Vibratory
The dog can detect vibrations by way of this sense.
Dogs may shy from unfamiliar vibrations. Gradual exposure to an environment,
which includes vibrations, allows the dog to ignore this input.
Remember that the dog locates the target odor
(drugs/ explosives/etc.) while it is in its prey or hunting drive. If the dog
becomes frightened, it will not hunt. Under such conditions detection work is
not possible.
Internal Receptors
Internal receptors tell the dog when it is hungry,
thirsty, or when it needs to eliminate waste. The receptors which tell a dog it
is thirsty or needs to eliminate must be satisfied before learning can take
place. However, the receptors that tell the dog it is hungry can be used
effectively to control the dog's behavior.
Factors Affecting Basic Senses
There are several factors that affect sensory
abilities. As a dog grows older, the visual and auditory systems become less efficient.
Various diseases also affect the dog's sensory ability.
Distractions
Distractions in an environment may confuse the dog
or cause the dog to ignore cues given by the handler. These could include
people, animals or vehicle traffic in the area while the team is working.
The working dog is constantly receiving information
concerning its environment with the sensing systems just described. If at all
possible, these distractions should be eliminated. Due to the working dog's
unique abilities, handlers, trainer and supervisors must be aware of the
limitations and capabilities of the detector dog.
Noxious Odors
Odors, such as ammonia, JP-4 and decaying flesh can
distract a dog in any situation and could in fact cause the dog to avoid areas
containing them. These odors primarily affect the olfactory system. Some odors
considered noxious to humans actually attract dogs.
Fatigue
Fatigue decreases the efficiency of all the senses,
especially when a sense is used for a prolonged period of time. A relatively
short period of recovery time is required to bring the efficiency back up.
Additionally, if the dog does not get adequate rest the basic senses would also
be adversely affected. Conversely, fatigue can be put to good use in controlled
aggression tasks with some dogs.
Working dog handlers should consider having a quiet
place for the dog to rest four to six hours before going on duty. Since most
police dogs are in the home environment a crate in a dark, cool room will offer
an excellent opportunity to rest. Family members should respect the quiet time,
allowing the dog to rest.
Age and Disease
As a dog grows older, the visual and auditory system
becomes less efficient. Various diseases also decrease the dog’s sensory
abilities. These two factors are considered together due to the fact that as a
dog ages, it becomes more susceptible to disease. These factors are of a more
serious nature than others because the effects are more pronounced and most
likely permanent.
DIFFERENCE BETWEEN LEARNING AND TRAINING
Learning is a relatively
permanent change in behavior as a consequence of the dog's experience. Training
is the implementation of principles of conditioning through repetition to teach
a dog a task.
ESSENTIALS FOR DOG TRAINING
Knowledge
You must know the principles
or conditioning. Understanding the principles of conditioning allows the
trainer to make informed and intelligent training decisions.
Practice
(Repetition)
The phrase "practice
makes perfect” applies very much to dog training. You must practice intelligently;
If you practice mistakes you will develop perfect
mistakes.
Patience
As a handler or trainer you must have patience.
Patience allows time for the animal's learning process to work. Remember it
takes many trials for the dog to learn a task.
Objective Training
Training will not be productive or successful
without purpose. Each training session must be designed to improve the team by
identifying strong and weak points, correcting deficiencies, or testing
corrective training to ensure that it was effective.
REQUIRMENTS FOR LEARNING
For learning to take place three requirements must
be met:
·
The final response must be defined.
·
The dog's behavior must be shaped through the method of successive
approximation.
·
Elimination of incorrect response with extinction training.
Defining the Final Response
The defined final response separates all other
behaviors associated with the task being performed. All training from that
point is geared to achieve the final response. In drug detector dogs, most
courts only recognize the dog's final response as sufficient evidence to
provide probable cause.
Method of Successive Approximation (Shaping)
In shaping, we reinforce behaviors that lead to the
final response. By gradually building on existing behaviors that are close to
the desired behavior, we shape the dog into the desired behavior. Simply put,
it is taking one step at a time.
Elimination of Incorrect Response with Extinction Training
During the shaping process, cues are often built in
to aid the learning process. These cues
must be gradually eliminated through a process known as extinction training.
TWO FORMS OF ASSOCIATIVE LEARNING
Two forms of learning used by dog trainers are
operant and classical conditioning.
These forms of learning and their application will be discussed in the
following sections.
CLASSICAL CONDITIONING
Classical conditioning occurs when a neutral
stimulus is paired with an unconditioned stimulus (UCS) to elicit a response.
Once the neutral stimulus starts to elicit a response, it becomes a conditioned
stimulus (CS) and the response is now a conditioned response (CR). Classically
conditioned responses receive no reinforcement after the behavior is exhibited.
Conditioned Stimulus (CS)
CS is neutral and will not elicit a response unless
a sufficient number of trials are conducted in which it is paired with an UCS
(Unconditioned Stimulus)
Unconditioned Stimulus (UCS)
An UCS will elicit a response automatically. The dog
does not learn to respond to this stimulus. It will consistently elicit an
automatic response in that species. Pain, pressure and discomfort are UCS used
for initial training of working dogs.
Unconditioned Response (UCR)
Any response that is reliably elicited from a dog
without a particular unconditioned stimulus is an UCR. A dog that is fatigued
or bored will sit or lie down on its own.
This is an example of an UCR.
Conditioned Response (CR)
This response is elicited by a conditioned stimulus.
This is the result of repeated pairing of a CS, which is neutral in and of
itself, with an UCS until the dog responds consistently to the CS alone.
OPERANT CONDITIONING
Dogs used for service are trained primarily using
operant conditioning. Responses that receive positive reinforcement are likely
to be repeated, and those that receive negative reinforcement or no
reinforcement at all are extinguished. Note the difference between classical
and operant conditioning. Operant conditioning receives some form of
reinforcement after the behavior has been exhibited.
Within operant conditioning, there are a host of
terms and applications. Handlers, trainers, and supervisors should familiarize
themselves with these concepts and obtain a working knowledge of their affects.
Operant conditioning tells us behaviors that are
positively reinforced will reoccur. It is the principle that reminds us that
handlers in detector dog work should never reinforce a response in field
conditions. Handlers may turn the situation into training and reinforce the
dog's response only after the presence of substance is confirmed. During the
schedules of reinforcement, we will learn that not every correct response needs
to be reinforced. But a strict contingency of reinforcement must be followed.
Reinforcement Contingency
To train an animal using operant conditioning, a
reinforcement contingency must be established. Simply put, correct behaviors
will receive positive reinforcement and incorrect behaviors will receive either
negative reinforcement or will be ignored. Failure to follow a reinforcement contingency
will result in an untrained and confused animal. For example, a detector dog
receives positive reinforcement for exhibiting its defined final response when
it encounters the target odor. If a detector dog receives positive
reinforcement for exhibiting its defined final response when the target odor is
not present, a violation of the reinforcement contingency has occurred.
Violation of the reinforcement contingency can become adventitious
reinforcement. Adventitious
reinforcement leads to superstitious behavior.
Adventitious Reinforcement
Reinforcement received without a contingency for
reinforcement is adventitious reinforcement. An example of adventitious
reinforcement is a detector dog receiving positive reinforcement for exhibiting
its final response where the target odor is not present. Another example of
adventitious reinforcement is a dog responding to the exterior of a vehicle and
a search of the vehicle produces one or two marijuana seeds. The handler
believes that the dog responded to the seeds themselves.
Superstitious Behavior
An untrue belief often occurs as a direct result of
adventitious reinforcement.
The Learning Model (SD-R-Rf)
This model is used in training a dog to learn a
task.
·
SD refers to the discriminative stimulus
·
R refers to the response
·
Rf refers to the reinforcement
|
(SD )
Discriminative Stimulus |
(R) Response |
(Rf) Reward |
|
Marijuana |
Sit |
Kong |
REINFORCERS
The term reinforcer can have a host of meanings. The
following sections will discuss different types of reinforcers and their use in
training.
Primary Reinforcer
This reinforcer earns its name as it fills a
psychological or biological need of the animal.
Things such as food and tangible rewards (kong or ball) are examples of primary reinforcers.
Most detector dogs are reinforced using prey-kill stimulation as the primary
reinforcer.
Secondary Reinforcer
Secondary reinforcers are learned and can be
anything. Learned or conditioned reinforcers are both positive and negative
reinforcers. They start as a neutral stimulus. Through association with the
primary reinforcer, they too become reinforcing.
POSITIVE REINFORCEMENT and REWARD
This is a training method using positive reinforcers
after the completion of a task. The receiving of the reward is based on making
the correct response, and increases the likelihood that the response will be
repeated.
Positive
Reinforcer
Any event that, when presented immediately following
a behavior, causes the behavior to increase in frequency. The reinforcer can be
either primary or secondary. A positive reinforcer to an animal does not always
equal reward.
Reward
The receiving of a positive reinforcer after the
completion of a task is a reward. A reward is the highest positive reinforcer.
It has more value to the animal than other positive reinforcers. Positive
reinforcers may not always be a reward, but a reward is always a positive
reinforcer.
Reinforcer Value /Hierarchy of Reinforcement
The dog receiving the reinforcer determines the
value of a reinforcer. Strength of the reinforcer can be measured by the manner
in which it maintains the desired behavior. For example, a strong reward will
maintain a high response rate for those behaviors that operate under its
control. When performance is reduced
without any noticeable cause the reinforcer for the behavior should be
examined.
Reinforcers have hierarchies. These hierarchies vary
from dog to dog. It often varies in the same dog from task to task.
Satiation
Satiation is when a reinforcer losses its power to
maintain a behavior. Simply put, the dog becomes tired of the reinforcer, such
as a ball, etc. Once satiation occurs, the handler or trainer must find a
replacement reinforcer.
Often a decrease in performance of a task is the
result of satiation. Handlers who
maintain proficiency training and still record a decrease in performance should
consider changing the dog's reward, schedule of reinforcement or both.
Immediacy
Maximum effectiveness of a reinforcer is achieved
when the reinforcer is given immediately after the behavior being reinforced.
During initial training stages, when the dog is learning a task, a reinforcer
is likely to have little or no effect if the reinforcer is received more than
10 seconds after the desired behavior occurs.
Adverse Effects of Positive Reinforcement
Positive reinforcement can be used to train
desirable behaviors, but it can also be used to accidentally teach unwanted
behaviors. In detector dogs many non-productive responses are due to handlers
rewarding the dog in field conditions before verification of the target odor is
established. Once the target is verified, the handler may turn the find into a
training scenario and then reward the dog (Ref: 'Adventitious Reinforcement'
and 'Superstitious Behavior').
Effective Application of Positive Reinforcement
When applying reinforcement, the handler or trainer
must first select the behaviors to be reinforced. He/she must also select a reinforcer that can
be controlled and maintains the desired behavior. Contingency of reinforcement must be followed.
NEGATIVE REINFORCEMENT and PUNISHMENT
Training methods that use negative reinforcers or
punishment is negative reinforcement. The removal of an adverse stimulus after
the dog has performed the correct response and the removal once the response
has been made increases the likelihood that the response will be repeated.
Negative Reinforcer
The removal of positive reinforcers can take the
form of punishment. This removal is negative reinforcement. Negative
reinforcers can be used to increase or decrease behaviors.
Punishment
Administrating an adverse stimulus in the face of a
particular response or failure to respond is punishment.
Titration
When applying punishment, the handler must be aware
of the dog's titration level. Titration is the amount of correction needed to
stop an unwanted behavior. Each dog has a different titration level. Caution
must be made not to over-correct a dog. Start with the lowest and escalate to
the level needed to accomplish the task. Should you exceed a dog's upper
titration level, you risk either retaliation or
causing the dog to go into avoidance.
Escape Training
Escape Training is simply the conditioning of the
animal by use of pressure. As with other techniques, escape training has
positive and negative effects on learning. Generally, dogs will yield to
lighter forms of pressure and resist heavier pressures. Escape training can be
used to show the dog what response you want. A good example is teaching a dog
to sit. Through use of pressure, the dog learns that the means of escaping the
pressure being applied on its neck and hindquarters is to sit.
Avoidance Training
In avoidance training the dog learns not to make an
incorrect response to avoid punishment. During avoidance training the dog
receives punishment for each incorrect response or failure to respond.
Punishment is either verbal alone or physical and verbal.
TIME FACTOR FOR POSITIVE AND NEGATIVE REINFORCEMENT
The most effective period between a response and the
reinforcer is + or - 1/2 seconds. However, the dog must be given its positive
or negative reinforcer immediately upon completion of the task for learning to
take place.
Care must be taken so the dog knows which behavior
brought the positive or negative reinforcer. If the reinforcer is not paired
with the desired or undesired behavior, the dog may not associate the
reinforcer with the correct behavior.
SCHEDULES OF REINFORCEMENT
Reward schedules are discussed in this section.
Reward schedules provide information of when reinforcement will be received.
Continuous Reinforcement Schedule (CR)
Positive reinforcement is given for every positive
response and every incorrect response receives negative reinforcement.
Intermittent Schedule of Reinforcement
Intermittent reinforcement is when correct responses
sometimes do not receive a reinforcer. The intermittent schedule of
reinforcement is the most powerful schedule of reinforcement. It is less
resistant to extinction, more resistant to satiation, and maintains the
behavior more consistently that any other schedule of positive reinforcement.
This is the desired schedule of reinforcement. The following schedules of
reinforcement are forms of intermittent reinforcement:
Fixed Ratio Reward Schedule (FR)
Reinforcement is given to the dog after it makes two
or more correct responses. The responses do not necessarily have to be the same
responses.
To start a dog on this schedule, every second
correct response is rewarded. When the dogs consistently
makes two responses to obtain a reward, three responses can now be
required. By increasing the number of responses one at a time, allowing the dog
to perform at each level with 100% proficiency, a high fixed ratio reward
schedule can be attained.
If proficiency is inadequate at any time, decrease
the number of responses required to obtain a reward. Then proceed as before,
adding one response at a time. The dog must perform at each level with 100%
accuracy.
Fixed Interval Reward Schedule (FI)
A reward is given when the dog is required to
respond for a fixed period of time or distance. In initial training a short
time period must be selected. If the dog does not respond correctly, a shorter
period must be selected until the dog responds correctly to obtain a reward. As
in the Fixed Ratio Reward Schedule, short periods are added to the interval and
the dog is required to attain 100% at each interval. If the dog fails to
respond correctly for the required length of time, adjust the time requirements
of a response to a lower time requirement.
Variable Ratio Reward Schedule
Once the dog has learned to perform on a high fixed
ratio reward schedule, the Variable Ratio Reward Schedule is used. The maximum
number of responses required by this schedule must have been learned by a dog
on a fixed Ratio Schedule. Select a range learned by a dog on a Fixed Ratio
Schedule. A range, for example, of 5 to 10 responses may be required. During
this range, the dog is randomly rewarded when a correct behavior is exhibited.
This reward schedule, once attained, gives you greater control of your dog's
behavior than the continuous or fixed ratio reward schedules. It should be
noted that behaviors
(good or bad) learned and rewarded on a variable ratio reward schedule are
harder to extinguish.
Variable Interval Reward Schedule (VI)
Once the dog learns to perform any task for a
specified period on a Fixed Interval Reward Schedule, the Variable Interval
Reward Schedule is used.
GENERALIZATION
generalization can occur when
a behavior or responses learned are acquired with a different triggering or
controlling stimulus that is similar to the one being presented. The fact that
a dog can generalize is important to a trainer and handler. Some
generalizations are positive and reduce training and learning time while others
are negative and impede training and learning.
Stimulus Generalization
Two or more stimuli can so closely resemble the
target stimulus that it causes the organism to produce the conditioned response
(defined final response) for the target stimulus. If the handler used the word
'drown' in the presence of the dog and the dog goes into the 'down' position,
the dog has generalized the sound of the two words.
Response Generalization
Response generalization occurs when a behavior
becomes more probable in the presence of a stimulus or situation as a result of
a similar behavior having been strengthened in the presence of that stimulus or
situation.
This is sometimes seen in detector dog when they
down instead of sit without the down having been reinforced in the presence of
a target odor.
Positive Transfer
Transfer in which knowledge
or skills acquired in one context or on one task result in increased
performance in another context or on another task.
Negative Transfer
Transfer in which knowledge
or skills acquired in one context or on one task result in decreased
performance in another context or on another task.
EXTINCTION TRAINING (ELIMINATION of INCORRECT RESPONSES)
Extinction training is training to remove unwanted
behavior or incorrect responses. Before an incorrect response can be
eliminated, its cause must first be identified. Then the reinforcer that is
maintaining the unwanted behavior must be removed. Once these are identified,
extinction training can begin. When conducting extinction training, handlers,
trainers and supervisors must beware of what may occur during extinction
training.
Extinction Burst
Extinction burst occurs when a sudden increase in
the incorrect response occurs. This occurrence is common. Many times those not
familiar with this behavior will stop effective extinction training believing
that the process is making the behavior worse.
What actually happens is that the problem temporarily gets worse before
it gets better.
Extinction Aggression
Aggression associated with the extinction process is
extinction aggression. During the extinction period, the reinforcer that
maintains the behavior are withheld causing
frustration.
Spontaneous Recovery
Spontaneous recovery is the reoccurrence of a
behavior that has already undergone extinction training. Once extinction
training has been completed, the dog should receive intermittent extinction
training to help prevent the incorrect response from reoccurring.
Stimulus Discrimination
Discrimination of one stimulus in face of all other
stimuli is stimulus discrimination. To obtain discrimination, extinction
training must be conducted. During detection training it is important that the
dog does not generalize to other similar odors.
Extinction Latent
Preventing a response in the
presence of a triggering stimulus or stimuli. For example, placing an
odor that is known to induce an incorrect response in detector dogs and
preventing the dog from responding to the odor. This form of extinction is very
effective. It is for this reason that handlers must never pull or force their
dog from investigating odor unless the handler knows for a fact that the target
odor is not present.
Causes of Incorrect Responses
Incorrect responses can arise at anytime and without
the handler's, trainer's, or supervisor's knowledge.
Incorrect responses greatly affect the overall performance of all police dog
teams. Every dog will from time to time make incorrect responses. It is
imperative that K9 handlers, trainers, and supervisors immediately address the
impact of an incorrect response. The occurrence of an incorrect response does
not mean a dog is unreliable, or unsafe, but failure to immediately address an
incorrect response with corrective training will lead to an unreliable and
unsafe dog.
Elimination of incorrect responses can be time consuming.
The time needed to identify the causes, and design extinction training to
eliminate the incorrect response can be greatly reduced with the maintenance of
accurate and complete training and use records.
Before an incorrect response can be prevented or
eliminated, an understanding of what can cause an incorrect response must be
reached.
Odor Association
Odors that are commonly associated with the target
odor are associated odors. In drug detection it is rare for drugs not to have
an associated odor either in training or actual searches. Odors of packaging
materials, masking agents, and cutting agents are a few of the odors associated
with drug dog training.
Odor association can also impair a dog's response.
Patrol dogs that always encounter a hidden helper wearing a protective sleeve
may not respond when the sleeve is not present.
Placement Association
Places where targets are commonly hidden can cause
incorrect responses. This applies to all detection work.
Handler Cues / Clever Hans
Handler cues usually occur from the handler knowing
where the target is hidden and exhibiting vocal cues and body language not
exhibited when the handler does not know where the target is located.
Handlers can also cue their dogs during actual
searches because of a preconceived idea that the target is present in a
particular location.
Schedule of Reinforcement
Should the dog be conditioned to a fixed interval,
it will respond approximately at that interval even if the target odor is not
present. In detector work some handlers train to find a set number of aids in a
set time frame because that is what they are tested on during certification.
Failure to Conduct Fading Exercises
Fading is the gradual change, on successive trials,
of a stimulus or stimuli that controls a response, so that the response
eventually occurs to a particular changed or completely new stimulus.
Failure to Conduct Testing
Testing can be used to determine if an incorrect
response problem is present. Failure to conduct this testing will greatly reduce
the dog's ability to provide probable cause or decrease the safety of the team
in patrol or explosive detection.
Adventitious Reinforcement
Adventitious reinforcement is not based on a
prescribed or specific response or behavior contingency. In detection work,
both the dog and handler can be adventitiously reinforced. The most common
example of adventitious reinforcement causing incorrect responses is a handler
rewarding a dog in field conditions before confirmation of the triggering
stimulus is made.
Handers can also become
adventitiously reinforced when the dog responds incorrectly, but drugs are
found.
PREVENTION OF INCORRECT RESPONSES
Most incorrect responses can be prevented with
on-going training if the handler, trainer or supervisor has a working knowledge
of the causes of incorrect response.
Prevention can usually be obtained by reviewing
training and use records. Long before the problem becomes apparent, it is
developing. It is imperative records not only be maintained, but the proper
data recorded and reviewed. Identifying trends and conducting extinction
training against those trends will go a long way in preventing incorrect
responses.