Transcript Chapter 5

Chapter 9
Observational Learning
Of Octopuses and Crabs
•
•
•
•
•
Octopus
Crab
Put crab in jar
Octopus opens jar
http://www.youtube.co
m/watch?v=ocWF6d0
nelY
Uh-oh!
Time to open jar
Results
Trials
Observation
• Second octopus watches training of first
octopus
Time to open jar
First
(model)
Second
(observer)
Trials
Observational Learning
• Also called “Vicarious Learning”
• A change in behaviour due to the experience
of observing a model
Early Work on Observational
Learning
• Anecdotal evidence
• Thorndike
– Puzzle box
– Experienced model, naïve learner
– Animals don’t learn by observation
• Successes and failures
Imitative Behaviour
• Not necessarily the same as observational
learning
• Do same thing as a model
• May indicate a lack of learning about a
situation
– E.g., doing a behaviour that leads to an aversive
outcome
Human Infants & Imitation
• At what age can humans imitate/learn
vicariously?
• Metzolff & Moore (1977)
– 12 to 21 day old infants
– Facial gestures: tongue protrusion, mouth open,
lip protrusion
Experiment
•
•
•
•
•
Experimenter: 90 sec passive face (baseline)
Infant shown gesture four times
Imitation-test period
Video taped and scored by “blind” judges
Supports imitation
% of mouth open response
% of tongue-out response
Results
50
25
baseline
tongue-out mouth open
Experimenter’s Gesture
10
5
baseline
tongue-out mouth open
Experimenter’s Gesture
Infant Research
• Replication difficult
– Tongue protrusion elicited by other means
– Infants’ attention: general arousal?
• Evolutionary advantage
• Observing and copying behaviour of
parents, siblings, etc. may bring more
attention
What can you Learn with
Observational Learning?
• Classical conditioning: no
• Operant conditioning:yes
• Observe model’s outcome (appetitive or
aversive)
Epstein (1984)
• Model bird (MB)
– Experienced
• Observer bird (OB)
– Naïve
• Experiment
– Model trained to use
object for food
– Ball (B), switch (S),
key (K); five phases
Five Phases
•
•
•
•
•
1. OB right; B, S, K in left
baseline
2. OB right; B, S, K in right
3. MB in left, OB in right; B, S, K in left
4. As above, but B, S, K in right
5. OB right; B, S, K in right
Results
• Phase 4: observer imitates model
• Phase 5: observer uses objects without
model to guide actions
• Learning due to observation of a model and
expectation of reinforcement (motivation;
latent learning)
• Not due to observer’s prior experience or
direct reinforcement
Levy, McClinton, Rabinowitz &
Wolkin (1974)
• Children observed model look at paired
pictures, indicating preferences
• Model received approval, disapproval, or
neutral consequences
• Observers’ subsequent preferences were for
the pictures that the models received
approval for selecting
Generalized Imitation
• Don’t see the outcome of the model’s
behaviour, but the observer imitates anyway
• Prior experience with observational learning
• We learn that imitating others’ behaviour
may provide reinforcement
• We generalize from one condition to
another
Generalized Imitation
• Also an issue of discrimination as applied
to:
• Discriminative stimulus
– Who to imitate
• Response
– When to imitate
• Outcome
– What to expect
Miller-Dollard Reinforcement
Theory
• Observational learning as a subset of operant
conditioning
• Observer’s behaviour changes due to
consequences of observer’s behaviour, not
the model’s
Process
• Three steps:
– 1. Learner observes behaviour of model
– 2. Learner copies response
– 3. Learner receives reinforcement --> continue
imitating
• Model rat knows maze; observer allowed to
follow
– Group 1: both model & observer reinforced at end;
observer will run maze on own
– Group 2: only model reinforced at end; observer will
not run maze (latent learning)
Interesting Issues
• When to delay before imitating
– Learning and telling a joke
– Limited time offer (e.g., trick-or-treating)
• Lack of reinforcement
– May not actually observe model’s outcome
– Model’s outcome may be on PRE
– Imitation may not lead to immediately
appetitive outcome
– Generalized imitation
Bandura’s Studies
• Bobo the Clown experiments
• Children as observers
• Watched various models demonstrating
behaviours
• Consequences of model’s behaviour
• Availability of reward/punishment
Model
Observers
Bandura’s Social Cognitive
Theory
•
•
•
•
Attentional processes
Retentional processes
Motor reproductive processes
Incentive and motivational processes
Characteristics of the Model
•
•
•
•
•
Rewardingness
Authority
Dominance
Similarity
Sincerity
Characteristics of the Learner
• Uncertainty
• Sex
• Age
Characteristics of the Situation
• Task uncertainty
• Task difficulty
• Presentation of model
Comparison
• Miller-Dollard Reinforcement Theory
– Behavioural; operant conditioning
– Past experiences
• Bandura’s Social Learning Theory
– Innate processes, learning, and cognition
– Future expectations
Television Violence
•
•
•
•
National Television Violence Study (1998)
8000 hours of programming
7 days/week, 6AM-11PM for 3 years
60% of programs contained violence
– Less than 4% contained anti-violence message
• By age 12, average child has seen 8,000
murders and over 100,000 other acts of
violence on TV
TV Violence & Aggression
• Strong correlation
• Adapted from Bushman & Anderson (2000)
Smoking & lung cancer
TV violence & aggression
Condom use and HIV
Secondary smoking & lung cancer
Lead exposure and child IQ
Calcium intake & bone mass
Homework & academic achievement
-0.2
0
0.2
Correlation
0.4
Issues for Studies
• Correlational findings
– Do not imply causation
– Third factor?
• Longitudinal studies
– Follow subjects over time
• Experimental studies
– Control and experimental groups
– Contrived?
– Brief time period
Phobias
• Development
– Vicarious acquisition
– Prevalence in humans?
• Treatment
– Flooding
– Systematic desensitization
Observational Learning
Treatments
• Modeling
– Used with the very young
– Sometimes more rapid
– May be better at generalization
• Three types of phobia reduction modeling
– 1. Graduated modeling
– 2. Participant modeling
– 3. Symbolic modeling
Modeling in Behaviour Therapy
•
•
•
•
A model can influence an observer
Facilitate known response
Teach new behaviours
Reduce/eliminate undesirable behaviours
Mirror Neurons
• Scattered throughout premotor cortex,
centres for language, empathy, pain
• Fire when certain actions are preformed by
or observed in someone else
• “Mental imitation” of witnessed (or heard)
actions
Discovery
• Giacomo Rizzolatti, Vitorio Gallese, &
Leonardo Fogassi
• “Raisin incident”
• Macaque monkey with electrodes in
premotor cortex
• Published in 1996
Locations in Humans
• More mirror neurons in more places than in
monkeys
• Premotor cortex (movement)
• Inferior parietal areas (perception)
• Posterior parietal lobe, superior temporal
sulcus, & insula (comprehend another’s
feelings, understand intention, and use
language)
Role
• Learning through observation
• Understanding meaning or intention of
action
• E.g., become better at golf by watching golf
• Not limited to motor responses
Gallese, Rizolatti, et al. (2005)
• Subjects listened to sentences describing
actions
• Same mirror neurons fired as would have if
subjects had done the action or seen the
action performed
• Mirror neurons responded to abstract
representation (i.e., language)