Learning Theories
HUMAN BEHAVIOR
LISA J. MERLO, PH.D.
Objectives
2
Understand theory of Classical Conditioning Applications to clinical practice
Understand theory of Operant Conditioning Applications to clinical practice Understand theory of Social Learning Applications to clinical practice
Classical Conditioning
Learning what events signal:
4
Lightning thunder Sour-smelling milk bad taste Unexpected loud noise startle response Chopping onions tears
The scientific study of simple associations led to the development of classical conditioning theory
Ivan Pavlov
5
Russian physiologist studying
the digestive system
Studied the role of saliva in
the digestive process
Developed method to
measure saliva production
Salivary Conditioning Apparatus
6
Gave dog food and recorded salivation while it ate
Pavlovs Saliva Research
7
Pavlovs Problem: experienced dogs salivated
before the food was presented
Pavlovs Theory: Some stimulus that preceded
the food presentation had acquired capacity to elicit the response of salivation
Dogs were exhibiting a simple type of learning
This type of learning is the foundation for
Classical Conditioning
Pavlovs Experiments
8
Pavlov conducted experiments to test his new theory
of learning
Studied the effect of different stimuli on the behavior
of interest (AKA, response)
Animals automatically demonstrated the response after exposure to some stimuli Animals did not demonstrate the response after exposure to other stimuli Animals could be taught to demonstrate the response to some stimuli by pairing them with others
Paradigm of Classical Conditioning
9
1st: Select a stimulus that automatically elicits a
characteristic response
Stimulus =Unconditioned stimulus (US) Response = Unconditioned response (UR)
Unconditioned means the stimulus-response connection is innate
2nd: Select a Stimulus for Conditioning (CS) Conditioned Stimulus (CS) Can be any reasonable stimulus that does not initially evoke the UR
Conditioned means the stimulus-response connection occurs only AFTER the conditioning procedure takes place
Classical Conditioning & Pavlovs Dogs
10
Key Definitions
11
Unconditioned Stimulus (US) - stimulus which
naturally triggers a response [food]
Unconditioned Response (UR) - unlearned, natural
response to the US [salivating]
Conditioned Stimulus (CS) - previously neutral
stimulus, which eventually triggers a response [bell]
Conditioned Response (CR) - learned response to a
previously neutral stimulus [salivating]
Cognitive Factors Are Ignored
12
Factors Affecting Classical Conditioning
13
The CS should be presented BEFORE the US The US should follow the CS closely in time
Short delay = best Trace (moderate delay) = more difficult to achieve CR Long delay = difficult to achieve CR Simultaneous = much weaker conditioning than short delay Backward (i.e., US then CS) = usually not effective
The CS should UNIQUELY predict the US It should provide new information suggesting that the US is about to occur
Higher Ordering Conditioning
14
Second Order Conditioning: CR transferred from one
CS to another CS
Example
Food (US) paired with metronome ticking (CS) to establish firstorder conditioning
Food (US) Salivation (UR) Eventually, ticking (CS) Salivation (CR)
Ticking (CS) paired with black square (2nd CS) to establish secondorder conditioning
Black Square Ticking (CS) Salivation (CR) Eventually, black Square (2nd CS) Salivation (CR) --- despite never being paired with food (US) Can also be done (less successfully) with 3rd order
conditioning
Key point: US does not always need to be present to establish association; Allows more flexibility in learning opportunities
Second-Order Classical Conditioning
15
Little Albert & Stimulus Generalization
16
Little Albert (11 months) was afraid of
nothing, except a loud sound made by striking a steel bar. This made him cry. By striking the bar at the same time that Albert touched a white rat, the fear transferred to the white rat (classical conditioning). After 7 combined stimulations (rat + sound), Albert remained greatly disturbed at the sight of a rat, but the fear also spread to include: a white rabbit, cotton, wool, a fur coat, a dog, a Santa Clause mask, and the experimenters hair. The fear did not transfer to his wooden blocks or other objects dissimilar to the rat.
Generalization and Discrimination
17
Generalization: Transfer of the effects of
conditioning to similar stimuli
In general, the more similar a stimulus is to the training stimulus, the greater its capacity to elicit a conditioned response
Discrimination: Subject responds to a new
stimulus differently than he responds to an established conditioned stimulus
Extinction
18
Conditioning is not necessarily permanent Extinction occurs when the CS is presented
repeatedly without the US (e.g., bell without food)
CS no longer signals the US, so conditioned
responding gradually diminishes
But CS-US association is not erased completely Spontaneous recovery Rapid Reacquisition
Spontaneous Recovery
19
Spontaneous recovery = reappearance of an
extinguished response
Example: following one day of extinction trails (i.e.,
presenting CS without US until no CRs are present), the CR will be present again the following day
Time is a key factor The more time between 1st and 2nd extinction sessions, the more spontaneous recovery is observed
Extinction & Spontaneous Recovery
20
Rapid Reacquisition
21
Conditioning phase extinction phase re-
conditioning phase (i.e. reacquisition phase)
Rate of learning is faster the second time Continues to get faster with repeated cycles
Clinical Applications: Classical Conditioning
22
Pair unconditioned stimulus with undesirable behavior
to change the response
E.g., treating nocturnal enuresis with Mower & Mowers Bell & Pad method
Bell (US) Wake-up and tighten muscles to prevent urination (UR) Bodily sensations before urinating (CS) alarm bell (US) Wake-up and tighten muscles to prevent urination (CR)
Use extinction to reduce unwanted response to
conditioned stimulus
E.g., decreasing fear response to object of phobia with extinction Repeatedly present the rat (CS) without the loud noise (US), until the CS no longer results in the CR
Operant Conditioning
Learning About the Consequences of Behavior
24
Operant conditioning = a procedure for studying
how organisms learn about the consequences of their own voluntary behavior
Realization that our ACTIONS (rather than
conditioned stimuli) lead to outcomes results in operant conditioning
By operating on your environment, you can
produce a positive or negative consequence
Edward Lee Thorndike
25
Studied animal intelligence and learning Placed a cat in the puzzle box. It was
encourage to escape to reach a scrap of fish Thorndike timed how long it took to escape. The cats experimented with different ways to escape the puzzle box and eventually would stumble upon the lever which opened the cage. When the cat had escaped, it was put back in the box and timed as it tried to escape. In successive trials, the cats would learn that pressing the lever would have favorable consequences, so they became increasingly quick at pressing the lever
Thorndikes Law of Effect
26
If a response to a particular situation is followed by a
satisfying or pleasant consequence, it will be strengthened.
If a response to a particular situation is followed by
an unsatisfying or unpleasant consequence, it will be weakened.
B.F. Skinner
27
Original research Rats in Skinner boxes Rats rewarded with food pellet for engaging in specific behaviors Basic assumption Voluntary responses are strengthened or weakened as a result of their consequences
Discriminative Stimulus: the
stimulus situation that sets the occasion for a response to be followed by reinforcement or punishment
Learning the Association
28
Classical vs. Operant Conditioning
29
The Nature of Reinforcement
30
Reinforcement = consequences that INCREASE
the likelihood of the behavior occurring again
Positive reinforcement = a consequence that, when presented after a behavior, increases the likelihood of that behavior occurring again (e.g., getting an ice cream cone as a reward for earning an A on a test)
Negative reinforcement = a consequence that, when removed after a behavior, increases the likelihood of that behavior occurring again (e.g., getting out of chores as a reward for earning an A on a test)
The Nature of Punishment
31
Punishment = consequences that DECREASE the
likelihood of a behavior occurring again
Positive Punishment: a consequence that, when presented after a behavior, decreases the likelihood of that behavior occurring again (e.g., getting a spanking after hitting your little sister)
Negative Punishment: a consequences that, when removed after a behavior, decreases the likelihood of that behavior occurring again (e.g., getting time out from the fun activity after hitting your little sister )
Schedules of Reinforcement
32
Schedule of reinforcement = the rule used by the
experimenter to determine when particular responses will be enforced (in order to teach the organism that its behavior uniquely predicts the consequence)
Continuous reinforcement schedule = consequence delivered EVERY time behavior occurs
Partial Reinforcement schedule = consequence delivered only some of the time after the response has occurred
Partial Reinforcement Schedules
33
Fixed ratio vs. Variable ratio: (deals with number
of responses before consequence presented)
Fixed Ratio = the number of responses required for delivery of the consequence does not change (is fixed) across trials Variable Ratio = the number of responses required for delivery of the consequence changes (varies) across trials
Fixed interval vs. Variable interval: (deals with
amount of time before consequence presented)
Fixed interval = consequence delivered for the first response that occurs following an unchanging (fixed) amount of time Variable interval = the allotted time before a response will yield a consequence changes (varies) across trials
Comparing Reinforcement Schedules
34
Variable Ratio Schedule Hardest to Extinguish
35
Effects on Behavior
36
Continuous reinforcement fastest acquisition of
response (learning); fastest extinction of response (unlearning) Fixed ratio fast acquisition; fast extinction Variable ratio consistent acquisition; slowest extinction Fixed interval quickly learns to adapt behavior to timing; fast extinction Variable interval consistent acquisition; slower extinction
Operant Conditioning Concepts
37
Extinction: When a previously reinforced behavior
decreases in frequency, and eventually ceases altogether, because reinforcement is withheld Spontaneous recovery: When an extinguished behavior reappears without having been reinforced Generalization: When an individual learns to make a particular response to a particular stimulus, and then makes the same response in a slightly different situation Discrimination: When an individual learns to notice the unique aspects of seemingly similar situations, and thus develops different ways of responding
** Names are the same as classical conditioning, but concepts are slightly different
Clinical Application: Operant Conditioning
38
Implement continuous reinforcement schedule (in order to
facilitate acquisition of desired behavior), then switch to variable ratio schedule (in order to protect against extinction)
Parent Training teach parents how to use operant
principles to shape childrens behavior E.g., children with disruptive behavior disorders Praise/reward, Token economy, Time Out, Extinction, Punishment, Contingency management
Exposure / Response Prevention use negative
reinforcement to change patients behavior E.g., patients with OCD & other anxiety disorders
Social Learning
Learning From Others
40
Social (Observational) Learning: Occurs when an
organisms responding (learning) is influenced by the observation of others, who are called models
Allows organisms to learn, without requiring their own trial
and error experiences
Modeling occurs when an organism imitates the behavior
of others
Albert Bandura
41
Conducted the Bobo Doll Experiment to
study vicarious learning Showed nursery school children a film that portrayed a woman striking, punching, and kicking a large, inflatable Bobo doll When later placed in a room full of toys (including the Bobo doll), many children imitated the woman and violently attacked the doll This was more common when children viewed a film in which the woman was praised (e.g., Youre the champion) for attacking Bobo
The Model and the Imitators
42
Factors Affecting Social Learning
43
Models are more effective when they are: Attractive, honest, competent, and have elevated status More likely to imitate models who are successful and/or rewarded for their behavior Vicarious reinforcement and vicarious
punishment have effects on behavior that parallel Thorndikes Law of Effect
Clinical Applications: Social Learning
44
Skills Training Modeling adaptive behavior as a parent, therapist, etc. E.g., teach shy children positive social skills Vicarious exposure to adaptive behaviors Show videos or models engaging in desired behavior E.g., show child with a dog phobia video of similar children playing with dogs and having fun Limiting exposure to maladaptive behaviors Prevent children from seeing others engaging in undesired behavior E.g., censoring TV shows, movies, video games
Questions?
45 [email protected]