Unit 4: Operant Conditioning Flashcards
What’s the difference between instrumental and classical conditioning?
classical: presence or absence of stimulus causes response
operant: behaviour causes presence or absence of stimulus (consequence)
Who did the study of instrumental conditioning start with and what was he interested in?
Edward Thorndike
interested in animal intelligence
How were Thorndike’s experiments generally structured?
hungry animals placed in puzzle boxes
food outside of boxes but in view
-> animals had to learn how to escape the box to obtain food
How did the animal’s behaviours change in the puzzle box?
initially unable to escape
slow to make right response
continued to practice until latencies become shorter
How did the animals learn how to solve the puzzle box?
trial and error to discover behaviour required to escape
successful behaviours retained
useless behaviours eliminated
How did Thorndike label the animals ability to learn how to escape a puzzle box?
animal intelligence
Why is animal intelligence not an accurate term?
many behaviours seem unintelligent
initial presence of various responses typical for confined animal, with some leading to a desirable result
consequences reinforce the action
->cat doesnt understand how levers work, but presses it because it thinks it will be rewarded for it
Law of Effect
If R in presence of S is followed by positive event -> becomes strengthened
If it isn’t followed by a positive event -> S-R association becomes weakened
What do we measure in discrete-trial procedures?
rat runs down a maze to get reward
measures response latency (time it takes for the rat to leave the start of the box) and running speed (how fast it reaches the end)
What’s a T maze trial?
type of discrete-trial procedure
allows us to measure percentage of correct choices
What are trials?
specific periods of time during which the animal can show instrumental responses
set by the experimenter
Why didn’t Skinner use discrete-trial procedures?
behaviour is continuous (one leads to the next)
-> trials more natural if animals aren’t removed
behaviour can be broken down into measurable units: operants
What is magazine training?
US paired with CS via classical conditioning
sound elicits sign-tracking response
How does response shaping work? (example: rat in Skinner box)
after magazine training, rat can learn operant response
1. food given if rat goes on hind legs anywhere in chamber
2. food given if rat leans over lever
3. food given if rat goes up on hind legs and presses lever
=> sequence called shaping/ reinforcement of successive approximations
What are operant responses in free-operant procedures measured as?
rates
Response rate
frequency of instrumental behaviours occurring
high: high probability of behaviour occurring
low: low probability of behaviour occurring
How can we differentiate outcomes?
appetitive vs aversive
positive vs negative
Which components do all instrumental conditioning procedures involve?
instrumental response
outcome (reinforcement, punishment)
stimulus
association between response and outcome
positive reinforcement
behaviour produces (adds) appetitive outcome
negative reinforcement
behaviour produces absence of aversive stimulus
positive punishment
behaviour produces aversive outcome
negative punishment
behaviour produces absence of appetitive outcome
Can a behaviour always be reinforced (by anything)?
no, only if behaviour is naturally linked to reinforcement
e.g. cant reinforce yawning in cats with opening box, because yawning isn’t naturally linked with release from confinement
What does the presence of a stimulus activate?
behaviour system related to that stimulus
e.g. hunger (S) causes hamsters to start digging and scrabbling (behavioural system linked to hunger), while stopping self-care behaviour (behaviour doesn’t address hunger)
What does instrumental conditioning depend on with regards to the reinforcement?
quality and quantity of reinforcement
nature of reinforcement
previous reinforcements for same instrumental behaviour
Behavioural contrast effect
big reward perceived as especially good after small reward and vice versa
Which types of relationships between response and reinforcement are there?
temporal relationship: contiguity
causal relationship: contingency
Are temporal and causal factors dependent on each other?
no
What can we say about temporal relations?
immediate reinforcement is preferable to delayed reinforcement
credit assignment
if too much time passes, we won’t be able to link specific behaviours to the reinforcement
(credit assignment is the reason for this)
What’s more important to create associations, contingency or contiguity?
contiguity
The fact that a behaviour occurred just before the reinforcement was more important than whether it caused the reinforcement. What is this kind of reinforcement called?
adventitious/ accidental reinforcement
learned helplessness
when experiencing a tense state repeatedly
-> feeling of being incapable to change the situation