Learning Theory Flashcards
Classical Conditioning
the learned association between two events where a neutral stimulus elicits an unconditioned response. Ian Pavlov’s dogs drooling when assistants entered the lab in anticipation of food.
Conditioned Emotional Response (CER)
any negative emotional response, typically fear or anxiety, that becomes associated with a neutral stimulus as a result of classical conditioning. Also known as Conditioned Fear Response (CFR) Classical conditioning influences emotional states. It is possible to counter condition an emotional response but can be time consuming.
Operant Conditioning
Learning by association to the consequence that follows the behavior. Behavior is a function of the consequences. B.F. Skinner
Difference between operant and classical conditioning
classical forms an association between 2 stimuli, operant forms and association between a behavior and a consequence.
Law of effect
Edward Thorndike, responses that are made just prior to a pleasant/satisfying event are more likely to be repeated.
The 4 Stages of Learning
- Acquisition (acquiring)
- Fluency (automatic)
- Generalization (application)
- Maintenance (always)
The 4 D’s
Distance
duration
different environment
delivery of reward.
The Law of Parsimony
When 2 competing theories make the same predictions, the one that is simpler is more likely to be accurate. Also known as Occam’s razor.
Behavior modification
the use of techniques to decrease or increase certain behaviors of the animal.
Pam Reid’s Categories of Canine Behavior
Useful, Critical, Nuisance, Dangerous
Antecedent
any stimulus present in the environment before the behavior occurs. Can be used to tell the dog to do a behavior.
Behavior
anything the dog does.
Consequence
anything that happens to the dog as a result of the behavior. Can be good, bad, or nothing at all.
The behavior you are training should be
Observable, measurable, and reproducible
Reinforcement
Increases the likelihood of a behavior happening again. Anything the dog finds rewarding. Treats, toys, tug, praise, etc.
Punishment
Decreases the likelihood of a behavior happening again. Anything the dog finds aversive. Collar correction, time out, verbal correction, special pressure, removal of reinforcer, etc.
How to assess a behavior
Predictability, frequency, and intensity
Differential Reinforcement of Other Behavior DRO
Reinforcing a behavior that is anything but the particular behavior you want to decrease.
Differential Reinforcement of Incompatible Behavior DRI
Reinforcing a behavior that is incompatible with the behavior you want to decrease.
Luring
showing the animal something it wants, usually food, and using it to encourage the animal to move in the desired way.
Shaping
the reinforcement of successive approximations of a desired behavior. Thus, my training with George consisted of marking and reinforcing first walking closer and closer to the hoop, then raising up his head, then touching the hoop with his beak, then holding onto the hoop for a micro second, etc.
Chaining behaviors
a behavior is broken down and each component is taught separately in small, sequential steps. Each behavior/step performed serves as a cue for the next behavior.
Backward Chaining
complex behavior broken down in steps and is taught by starting with the last step and working backward from there. The learning is often easier because the newly acquired step is immediately followed by an already learned and comfortable step. An application of the Premack Principle.
Premack Principle
A high probability behavior (something the dog loves to do) can be used as a reward for a low probability behavior. Example: Back chaining
Cue
antecedent, stimulus, command, signal, or request. Something that comes before a behavior and tell the dog to do the behavior.
Discriminative Stimulus
a conditioned stimulus or antecedent event which signals a certain response will be reinforced.
Overshadowing
when the most important detail in the environment (from the animal’s point of view) covers up other details that are present.
Blocking
a previously learned cue conflicts with a newly learned cue. If they are presented simultaneously the previously learned cue is said to be ______ the new cue.
Salience
something that has noticeable significance to the animal
Stimulus Control
a dog responding correctly to a cue when asked. Not doing the behavior if not cued. Not doing the behavior in response to another cue. Not offering a different behavior in response to the cue.
Discrimination
being able to tell relevant stimuli from all other stimuli
Generalization
Being able to perform a behavior when any element of the context changes.
Superstitious Behavior
coincidental learning of some irrelevant behavior along with a desired one. Usually as a result of accidental reinforcement.
Targeting
teaching your dog to touch a designated body part to a designated location. Nose targeting is most commonly taught, but it can also be trained with a front or hind paw, a hip or shoulder, even an ear or tail. Targeting often starts with nose targeting which a trainer can then use to teach general targeting. This will allow you to teach behaviors by using your hand as a target instead of using luring. Behaviors such as loose leash walking and even recall can be obtained with targeting. Another popular technique is using a mat or platform for full body targeting, often referred to as “place.”
Capturing
observing the animal and rewarding a spontaneously offered/emitted behavior.
Molding/Modeling
physically manipulation of the dog into the desired position. Opposition reflex of the dog may work against you. Physical touch can also distract the dog from what it is supposed to be learning.
Prompting
similar to luring, except the animal can’t see the food. For example, I held my hand (the one that always delivered food) above the hoop to encourage George’s head to move upward toward the hoop. I marked each approximation with “Good,” and then fed from the hand that did the prompting. It is more effective to move one’s body in a way to encourage the desired action, rather than let the food itself be the focus. That way the animal pays more attention to what he is doing in relation to the environment.
Fading a lure or prompt
One method for fading a lure is to remove the lure in small steps, while keeping in mind not to make things harder and harder for the dog as he succeeds. To do this, lure the behavior 3-5 times with a treat (a nice smelly one) in your hand. Then pretend to get a treat but keep your hand empty. Lure the behavior with your “pretend treat” and click or say “yes” as the dog does the behavior. Next, open your hand to reveal that no treat was inside and then as quickly as possible get a treat for your dog and feed it to your dog. The fact that he was following an empty hand may come as a shock to your dog, so the next time you lure the behavior you should go back a step and lower criteria by using the food lure again. You don’t want to raise criteria too quickly or your dog could become frustrated. Continue to lure your dog with a treat, and then without a treat, and also with the smell of the treat in your hand so that your dog doesn’t get confused and stop following your luring hand gesture. Once you’ve trained about 10 behaviors using this method, your dog should start to become an expert in weaning off the lure.
A benefit of using luring to train behaviors is that you can turn the luring gesture into the hand signal that you will ultimately use to cue the behavior. For example, to teach the cue for spin, you can start out by moving your hand in a big circle with no treat, then in a circle slightly higher from the ground, then a smaller circle while bending over less. Finally, you can fade the gesture into a tiny finger-flick or shoulder-movement in the direction you want the dog to turn. You can then add a verbal cue for the behavior, if you wish. When you are changing the luring gesture into the hand signal, make sure you don’t just keep increasing difficulty, especially if your dog is new to this process. Instead, lower criteria and raise criteria on the average, over time. For example, you could do a big hand gesture, then 3 smaller ones, then a big one again, then a small one, then a smaller one, then a bigger one. This helps the dog feel that things aren’t just getting harder and harder. In other words, if the exercise is always getting harder and more confusing the dog can lose interest.
Flooding
to expose the dog to the stimulus that triggers the unwanted reaction at a close proximity and through a prolonged time of exposure. High risk of fallout and generally not ethical or least intrusive minimally aversive. Example: Take a gun shy dog to a gun range.
Positive Reinforcement
Adding something good. Increases likelihood of behavior happening again. Example: Dog sits, gets tasty treat.
Negative Reinforcement
Removing something, the removal is reinforcing to the animal. Increases the likelihood of behavior happening again. Example: Upward leash pressure applied, when dog sits the pressure is released.
HH Definition: Contingently withdraw an aversive antecedent stimulus to increase the probability that the right behavior will occur.
Positive Punishment
Adding something unpleasant. Decreases the likelihood of behavior happening again. Example: Dog jumps up on person and a collar correction is given.
HH Definition: Contingently deliver an aversive consequence to reduce the probability that the problem behavior will occur.
Negative Punishment
Removing something, the removal is punishing to the animal. Decreases the likelihood of behavior happening again. Example: Dog jumps on person and person leaves the room.
HH Definition: Contingently withdraw a positive reinforcer to reduce the probability that the problem behavior will occur.
Extinction
In Operant Conditioning it is the procedure of withholding the reinforcers that maintain behavior after the behavior has already been taught and regularly reinforced. In classical conditioning it is the procedure of repeatedly presenting a conditioned stimulus without the unconditioned stimulus so the learned stimulus loses the power to elict the conditioned response.
HH Definition: Permanently remove the maintaining reinforcer to suppress the behavior or reduce it to baseline levels.
Extinction Burst
the phenomenon of behavior temporarily or briefly getting worse, not better, when a previously rewarded behavior is not rewarded.
Spontaneous Recovery
behavior affected by extinction recurring when the trigger is presented again. Also know as transient increase in behavior.
Counter conditioning
Process where normal defense reactions elicited by aversive stimulus, are modified by association with a positive reinforcer. Also known as active desensitization.
Desensitization
Exposing an animal to a stimulus using time or experience to drive the stimulus value toward neutral. A process of changing an animal’s perception of an event to a neutral perception.
Habituation
the lessening or disappearance of a response with repeated presentations of the stimulus. Also known as passive desensitization.
Sensitization
intensifying of an animal’s response to stimuli over repeated presentations of the stimulus.
Continuous Reinforcement Schedule
the desired behavior is reinforced every single time it occurs. This schedule is best used during the initial stages of learning to create a strong association between the behavior and response and is most effective when trying to teach a new behavior.
Fixed Ratio Reinforcement Schedule-
a schedule of reinforcement where a response is reinforced only after a specified number of responses. Essentially, the subject provides a set number of responses and then the trainer offers a reward. One advantage of this type of schedule is that it produces a high, steady rate of responding with only a brief pause after the delivery of the reinforcer. For example, it might be the delivery of a reward for every fifth response. After the subject responds to the stimulus five times, a reward is delivered.
Intermittent (Variable) Ratio Reinforcement Schedule
reinforcement is delivered after a varying number of correct responses.
Fixed Interval Schedule of Reinforcement
reinforcing a behavior after a certain interval of time has passed. The interval of time is always the same. In a 3-second schedule, the first response after three seconds have passed will be reinforced, but no response made before the three seconds have passed will be reinforced.
Intermittent (Variable) Interval Reinforcement Schedule
reinforcing a behavior after a variable interval of time has passed. The interval of time is not always the same but centers around some average length of time. Example: In a 3-second schedule, the time of each interval changes but still ends up at a 3 second average. After an animal learns the schedule, the rate of behavior tends to be steadier than with a fixed interval schedule.
Differential Reinforcement Schedule
type of variable reinforcement schedule where only the “best” responses of the behavior are reinforced.
Limited Hold
variation of differential reinforcement. The reward is based on the speed/latency of the animal’s response. Example: If the dog does not sit within 3 seconds, reinforcement is withheld.
Jackpot
an unusually large or valuable unexpected reward
Secondary Reinforcer/Conditioned Reinforcer
event that marks a behavior as rewardable and promises the delivery of a reward in the near future. Example: Clicker, marker word
Deprivation
Reducing availability of or access to a reinforcer. Can cause anxiety, frustration, and/or aggression.
Differential Reinforcement of Alternative Behavior
Reinforce an acceptable replacement behavior and remove the maintaining reinforcer for the problem behavior.
