Transcript Operant or Instrumental Conditioning

Operant or Instrumental
Conditioning
Psychology 3306
Introduction



Thorndike and his
puzzle boxes
Guthrie and Horton
Superstitious
Behaviour




Interim
Terminal
Adjunctive
Not exactly
superstitious or
random
Shaping

Successive approximations


Secondary reinforcers


Get closer and closer to behaviour
Feeder click for example
Behaviour modification
Freddie!





Coined the phrase
‘Operant conditioning’
The animal operates on
the environment
Unlike ‘respondent
conditioning’ (Pavlovian)
Pioneered the use of free
operants
Pioneered the use of
respone rate
The Skinner Box


Basically this allowed
the researcher to walk
away
Allowed for a
dependent variable
that could be easily
measured and
compared across
species too
Criticisms of the Skinner box
Is it artificial?
 Well duh…
 But
 Many species can be tested
 Real world applications
 Therapy
 Who cares?

Key concepts and terms
Discriminative stimulus
 Three term contingency
 Acquisition
 Extinction
 Spontaneous recovery
 Generalization
 Conditioned reinforcement
 Response chains

Constraints
Instinctive drift and the Brelands
 Autoshaping (Brown and Jenkins, 1968)

Superstitious behaviour?
 Form of response depends on reinforcer
(Jenkins and Moore, 1973)
 Wasserman’s chicks (1973)
 Timberlake’s behaviour systems approach

Schedules of Reinforcement
You could give a reinforcement after each
behaviour you are interested in
 This is called CRF or Continuous
reinforcement
 However this is rarely used
 Does not maintain behaviour very well

Schedules of Reinforcement

Fixed Interval



First response after a
given interval is
rewarded
FI Scallop
Variable Interval


Like FI but varies with
a given average
Scallop disappears
Schedules of Reinforcement

Fixed Ratio



Reinforcement is given
after a given number
of responses
A little less smooth
Variable Ratio

After a varying number
of responses
Schedules and their properties
Variable schedules are more robust
 PREE, Partial reinforcement extinction
effect


Harder to extinguish responding on VI, FI, VR
and FR than on CRF
DRL, Differential reinforcement for low
rates of responding
 DRH, High rates

Schedule this….
Concurrent and chained schedules
 Behaviour follows the schedule in effect at
the time
 Allowed people to determine that the post
reinforcement pause in FR schedules is
due to the present schedule and not the
previous one

Applications

Work with autistic kids
Prompts
 Fading
 Secondary reinforcers

Token economies
 I/O applications
 Behaviour therapy

These ideas are nothing new…
Most folks are unaware of schedules and
contingencies
 Systematic application thereof
 Who cares?

Punishment and avoidance
Behavior
increases
Stimulus
Presented
Stimulus
Removed or
omitted
decreases
Positive
reinforcement
punishment
Negative
reinforcement
Omission
Avoidance
Shuttle box
 Go from escape to avoidance
 Avoidance paradox
 Two factor theory
 Avoid by escaping CS
 Animals will avoid a CS that predicts
shock in another context

But….
Does the CS induce fear?
 Equivocal at best
 Maybe avoidance itself is reinforcing
 This is the one factor theory
 The Sidman test shows that this is true,
avoidance itself is reinforcing
 But, temporal conditioning

Cognitive theories
Selligman and Johnston
 Expectations
 Animal expects:

No shock if it responds
 Shock if it does not respond

This explains the slow extinction
 Shock avoidance response blocking,
remove the ability to escape, you get
extinction

More on avoidance
Bob Bolle’s idea about SSDRs
 Learned helplessness
 Is it depression?


Suggestive, but not quite I don’t think
Punishment
Opposite of reinforcement?
 Sorta
 But, to be effective it must be:

Introduced at full intensity
 Given immediately
 After every behaviour

Motivational effects
 Other contingencies and behaviours

Bad boys bad boys, watcha gonna
do?
Maybe a punisher is an SD?
 All that said punishment CAN control
behaviour
 So, what’s the down side?

punishment
Fear and anger are bad for learning
 General suppression
 Constant monitoring needed
 Avoidance
 Reluctance to use it
 Bad consequences
 It is just plain mean

Omission
The avoidance of punishment
 Easily learned
 With all this stuff on punishment,
remember morality and data are two
different things
