Learning

advertisement
Learning
Classical Conditioning
The Fathers of Behaviorism
(1st half of 20th century)
What is learning?
 Long-lasting
change in behavior resulting
from experience. (behavior as
measurement of learning)
Classical Conditioning

Ivan Pavlov

1849-1936
 30+ years in research
on Learning
 Tripped upon theory
of learning while
studying digestion
 Classical Conditioning
Pavlov’s Experiment
Classical Conditioning
 Learning
to associate a neutral stimuli with
another stimuli that produces reflexive,
involuntary responses. (Bell with food…)
Key principles
 Unconditioned

Something that elicits a natural, reflexive
response (food)
 Unconditioned

stimulus (NS)
NS (bell) paired with the US (food) to form
association between US and CS
 Conditioned

response: (UR=unlearned!)
natural, involuntary response (salivation)
 Neutral

stimulus: US
Stimulus: (bell)
When CS elicits CR (NS becomes CS)
Conditioning Processes
 Acquisition

Learning: when animal responds to CS
without US (bell = salivation)
 Strength
of conditioning affected by order
and timing of US and CS
 So, what’s the most effective method of
conditioning? (In Pavlov’s experiment)

Ring bell, while it is still ringing, present food
(Delayed conditioning)
 What

if you ring bell after the food?
Backward conditioning = ineffective
Extinction
 To
unlearn a behavior (suppressed)
Qualifies when CS no longer elicits the CR
 How


is it done?
Present CS without US
(Bell without food)
Spontaneous Recovery
 Sudden
reappearance of a CR upon
presentation of CS (after extinction)
 Renewal
Effect: sudden reappearance of
CR after extinction when return to
environment where acquisition took place
Generalization / Discrimination
 Generalization

When a stimulus similar to CS will elicit a CR.
(similar to bell sound- tapping glass with
spoon)
 Discrimination

To distinguish between various stimuli
(between animals, objects. Sounds etc.)
John Watson
“Tabula Rasa”

1913: Publication in Psychology Review

“behaviorist manifesto”
1) Psychology’s content should be behavior
2) Method should be objective rather than introspective
3) Its goal should be the “prediction of and control of
behavior” rather than the fundamental understanding of
mental events
Age 30: chair of John Hopkins Psychology department
1915: president of American Psychological Association
John Watson: Aversive Conditioning







Little Albert
Experiment (11 months)
Aversive conditioning
(conditioning for
negative response)
US = rat
UR = liking of rat
CS =gong behind boy
CR= fear of rat
Life examples?
Second-order Conditioning
 Second-Order,
or higher order conditioning
 Once a CS elicits a CR, the CS can be used
(as a US) to condition a response to a new
stimulus
 Example:
 Dog
salivates to bell (first order conditioning)
 Light is paired with bell (second-order)
 Light = salivation
Biology and Classical Conditioning

Animals / humans are biologically wired to make
certain associations more easily than others

Examples:
 Learned taste aversions (adaptive response)
 Disgust reactions (Rozin and Colleagues)



Fudge: shaped in squares & dog feces
Bottle: labeled sucrose & cyanide
Classical conditioning, but biologically predisposed
Garcia and Koelling’s Experiment
 Biological
Preparedness in Classical
Conditioning
 CS
US
Learned Response
Loud noise
shock
fear
Loud noise
radiation (nausea) nothing
Sweet water
Shock
nothing
Sweet water
radiation (nausea) avoid water
What conclusions can be drawn from this?
Results appear adaptive. (each animal has different
biological predispositions to learning that enhance survival)
Thus, significance?

Classical conditioning is a critical way in which
all organisms learn to adapt to their environment
 Classical Conditioning in today’s world? Provide
an example for each of the following:




Drug addiction / quitting
politicians
Advertising
Fears / phobias
Operant Conditioning

Learning based consequences…


Association made between consequences and one’s
behavior
Thorndike’s Law of Effect:
(Early pioneer of Operant Learning)



Positive consequences results in strong stimulusresponse connection = increased behavior
Negative consequence results in weakened stimulusresponse connection = decreased behavior
Instrumental Learning: Consequence shapes
behaviors
Thorndike’s Cat and Puzzle Box Study
 Cat’s
learn gradually, not by insight.
 Suggests stimulus-response process
Classical vs. Operant
 Remember…
 Classical
= stimuli
 Operant = consequences
Introducing B.F. Skinner….

“Cognitive science is the creationism of psychology.”

“External influences, not internal thoughts feelings,
influence behavior.”

“Recognizing that behavior is shaped by its
consequences is the first step in taking control of the
environment and ensuring that it delivers consequences
promoting desirable behavior.”

“The movement toward a better society demands giving
up the belief in dignity.” (Dignity is an illusion..)
The Skinner Box
Skinner’s Reinforcement
 Reinforcement
: A consequence that
increases likelihood of behavior
Types
Effects
Examples?
Positive R
Add something pleasant
Negative R
Remove something unpleasant
Skinner’s Punishment
 Punishment: A consequence
that
decreases likelihood of behavior
Types
Effects
Examples?
Positive Punishment
Adds something negative
Negative Punishment
(omission training)
Removes something pleasant
Application:
Reinforcement or Punishment?

Indicate for the following which type of
Reinforcement or Punishment applies.
1.
Taking aspirin for a headache.
Water boarding to force a confession
Running home to get out of the cold.
Having your license revoked for speeding.
Spanking a child for a tantrum.
$25 for each A you make…. (Insane!)
2.
3.
4.
5.
6.
Learning by Operant Conditioning
 Shaping:

Reinforcing the steps used to reach a desired
behavior. (single behavior: Press bar for
food)
 Chaining:

Reinforcing a number of separate behaviors
for a more complex activity. (Obstacle
course)
Operant Conditioning

1.
2.
3.
4.
5.
The following terms can also be applied
to Operant Conditioning. Explain an
example for each using the Skinner Box.
Acquisition
Extinction
Spontaneous recovery
Generalization
Discrimination
A Quick Review!!!

1.
2.
3.
4.
5.
Brain defrost……
Pavlov is to _________ , as Skinner is to
_________ .
Who coined the phrase behaviorism? (Baby
Albert and aversive conditioning…)
Pavlov’s classical conditioning involves
natural, or uncontrolled responses (T-F)
Identify Skinner’s four major consequences of
operant learning and give an example for
each…
Classical conditioning is shaped by ________ ,
while operant is shaped by _________ .
Types of Reinforcers
 Primary

Reinforcers
Natural reinforcers: Examples?
• Food, water, rest (innately pleasing)
 Secondary

Reinforcers
Things we’ve learned to value : Examples?
• Praise, treasure box, to play video games
Money as a Reinforcer
 Money

Can be used for anything
 Token



= generalized reinforcer
Economy
Tokens as positive reinforcement
Cash in for other reinforcers
Sound familiar?
• Schools, mental institutions, prisons
Challenges of Parenthood
What’s the Best Consequence…?
For each of the following, choose only one of
Skinner’s 4 consequences and how you would
specifically enforce it.
1.
2.
3.
4.
Your ten year old, in a fit of anger, accidentally
breaks the living room window.
Your 11th grade teenager tells you she is going to
a movie and heads to a party.
Your four year old kicks you in the shin.
Your middle school daughter makes straight A’s.
Questions for Consideration…
 Is
spanking an advisable punishment for
children?
 To what extent should we reinforce our kids?
For each of the following provide a specific
reinforcement, if one is necessary.


Making good grades on your report card
Completing your chores all week
Is buying a car for your 16 year old reinforcement?
Biology and Operant Conditioning
Reinforcers: Effects can Vary
 Effect of reinforcer can vary depending on
animal, its instincts, and situation


Instinctive Drift: ignore rewards to follow natural
(instinctive) behavior
Premack Principle


If two activities- the one preferred can be used to
reinforce the one not preferred.
Example?
• Eat your lima beans = “may be excused”
Reinforcement Schedules
 Reinforcement
Schedules = pattern of
reinforcing behavior

Administered in 2 ways
• Ratio = number of responses made
• Interval= passage of time
Reinforcement Schedules
4




major reinforcement schedules
FR (Fixed ratio) = Reinforcement after set
number of responses- FR-5
VR (Variable ratio) = Reinforcement after
varied number of responses (average number
of responses set- VR-5)
FI (Fixed Interval) =Fixed amount of time set
before reward for behavior- FI 3
VI (variable interval) =varied amount of time
before reward (average time set- VI-3)
Learning and Extinction
 FR
and FI = faster acquisition (learning),
but faster extinction
 VR and VI: Slower learning but slower
extinction
 Which of these four yields the highest rate
of response?

FR
 Why
slower extinction with VR and VI?
• Noticing a break in pattern is more difficult
• “always that chance”

Activity: Identifying reinforcement schedules..
Learning with Punishment
 Escape

learning
To terminate an aversive stimulus: Example?
• To disrupt English class so as to “get out”
 Avoidance

learning
To avoid stimulus all together: Example?
• Cut English class
Pitfalls of Punishment…
 According
to behaviorists, what are the
potential pitfalls of punishment?




Tells only what not to do, not what to do…
Creates anxiety which interferes with learning
Only suppresses behavior, doesn’t eliminate
(discrimination)
Physical punishment = aggressive behavior
(correlation, not causation)
Classical v. Operant
 Compare
and contrast Classical and Operant
Conditioning.
Similarities



Both forms of associative learning
Both involve acquisition, extinction, spontaneous
recovery, generalization, discrimination
Both influenced by biology, cognition
Differences
 Classical
Response = automatic
Reward independent of action
learning = autonomic response
Operant
Response = voluntary
Reward contingent on action
learning = voluntary behavior
Behavioral Legacy

Pavlov:



Skinner



Classical Conditioning- how all organisms learn to
adapt to their environment
Practical applications for fears, phobias, etc.
Definitive insight into learned behavior
Practical applications abound
Both asserted that learning occurs without
thought (cognition)
 Focused only on observable behavior
Cognitive Learning
 How
could cognitive theorists argue that
cognition is influential in both classical and
operant conditioning?


Classical: CS triggers anticipation of US
Operant: awareness that responses =
consequences and thus act to maximize
reinforcement (minimize punishment)
Cognitive Learning

Observational Learning




AKA Modeling
Observation / imitation
Mirror neurons (frontal lobe / neural basis for
observational learning)
Albert Banduras: Bobo Doll Experiment
• Social learning theory (species specific)


Prosocial behavior (role modeling)
Antisocial behavior (Bobo Doll Experiment)
• Implications for television and youth?



By age 75 in U.S. 9 yrs of T.V.! (9 of 10 homes)
World Pop Culture (billion tv sets)
MTV 17 languages / CNN 150 countries
Cognitive Learning
Latent Learning (“hidden”)
 Learning that is not directly observable
 Tolman’s Rat maze study






Group 1: reward every time reached goal
Group 2: no reward when reached goal
Group 3: no reward 1st 10 days, reward on 11th
Finding: Latent learning (3rd group had learned
cognitive map in 1st trial, but didn’t show it until
reward)
Thus learning takes place without reinforcement
Cognitive Learning
Learning (“aha!”)
 Sudden grasp of problem
 Wolfgang Kohler: Chimpanzees
 sudden insight, not gradual strengthening
of S-R association
 Insight
Cognitive Learning
 Abstract

Learning
Higher order thinking (inferring relationships,
complex problem solving)
Download