Lecture 14: Goal-Directed Behaviour Flashcards
Describe Thorndike’s puzzle box
Cat was put in confined space, and it would eventually realise that by pressing the pedal, would result in the door opening and it being able to escape. They would learn the positive outcome of pressing the pedal and also avoid the negative consequence of staying confined in the box.
What are some of Thorndike’s laws of learning
Law of Effect - behaviour that leads to a positive outcome is more likely to occur in the future.
* Law of Exercise – connections between responses and outcomes are strengthened by repetition.
* Law of Readiness – learning is motivated by an internal state.
Describe Operant/instrumental conditioning
Associations between response/behaviour and
outcome/consequences (Response-Outcome learning)
* Learning of adaptive behaviour
– Through experience of success, failure (trial and error
learning)
* Organism operates on the environment
– Behaviour changes the environment
* Behaviour is instrumental
– Obtains desired effect (animal obtains goal)
– This type of learning facilitates goal-directed learning
Describe Skinner’s box
Animal like rat or bird in a box, accidentally presses lever and receives food so associates pressing the lever with a reward.
Define reinforcer
Stimulus/event that increases the likelihood of the preceding behaviour to occur. e.g. food
Can be:
- Positive reinforcer: Stimulus (usually positive) produced by the behaviour that increases the likelihood of the preceding behaviour to occur.
Negative reinforcer = Stimulus (usually negative) eliminated by the behaviour that increases the likelihood of the preceding behaviour to occur. It is not punishment. e.g. electric shock is avoided so positive outcome.
Define punishment & Omission
Punishment = Negative stimulus/event that decreases the likelihood of the preceding behaviour to occur. e.g. pressing the lever leads to shock so they wouldn’t press it.
Omission = Elimination of positive reinforcer decreases the likelihood of preceding behaviour. If the animal doesn’t receive a reward it wouldn’t press the lever.
Describe the types and schedules of reinforment used in the Skinner box
- Continuous reinforcement = each behavioural response is reinforced.
- Partial Reinforcement = behaviour is reinforced only part of the time.
1)Ratio schedules = reinforcement given after every nth response
* Fixed= response requirement always constant (e.g., FR1, FR10)
* Variable= response requirement varies around average (e.g. VR5)
2)Interval schedules = reinforcement given after certain amount of time
* Fixed = reward intervals constant (FI10)
* Variable= reward interval varies around mean time (e.g., VI10)
What psychological approach is conditioning part of
behaviourism - Skinner did not believe in free will
Give a modern example of operant conditioning being used today
In token economies or contingency management widely used in substance treatment to reduce relapse.
Compare respondent conditioning vs. operant conditioning
1)Learning associations between
stimuli in the environment
(Stimulus-Stimulus Learning)
- Reinforcement not contingent on action/behaviour/response
- Behaviour is elicited in response to a stimulus
(CS»respondent)
2) Operant (Skinnerian) conditioning
- Learning associations between actions and stimuli/outcomes (A-O learning)
- Reinforcement contingent on action/behaviour/response
- Behaviour is emitted/produced by the individual (operant)
Give a dual-process approach example: avoidance learning
Rat placed in chamber with 2 compartments
*Speaker delivers auditory stimulus (would-be
CS)
*Grid floor delivers mild footshock (US)
*Barrier for escape/avoidance of footshock
*Initially, rat escapes following US (i.e.
footshock)
*Eventually learns to escape following CS (i.e.
auditory stimulus)
Classical conditioning – Tone (CS) leads to shock (US)
Operant conditioning – Escape/avoidance leads to safety.
The animal learns to associate the sound with the electric shock, the animal then learns there’s a hole on the wall so that the animal can escape/avoid the electric shock.
Give the principles of associative learning
Learning through reinforcement
* Association by contiguity
-Co-occurrence in space and time
* Arbitrariness
– We can learn associations between any stimuli and between any response and outcome
* Empty organism
– Organism is black box – collection of associations
* Passive organism
– Learning happens TO the organism
Explain the taste aversion learning experiment
1) Compound CS – Bright, noisy & sweetened water. All rats conditioned to associate sweet taste with a click and a light.
Group 1 of rats: CS + US - Mild footshock
(immediate discomfort) when licking nozzle/bottle w sweet taste
Group 2: CS + X-rays (delayed illness) when licking it. Symptoms perceived later on than shock.
There isn’t continuity between illness and the conditioned stimulus.
Preference test of conditioned aversion: choice between bright/noise flavourless water and sweet water - split CS
Group 1: avoids the bright/noisy water.
Group 2: Avoided the sweet water.
Suggests continuity is not necessary and that we can associate something with making us ill even if it happens later in time.
How does the taste aversion experiment abide to the principles of associative learning?
Association by contiguity?
But little contiguity between taste and illness
* Arbitrariness?
No learning between taste CS and shock, or between click/light CS and illness
* Empty organism?
Certain associations are easier established than others
(“Preparedness principle”; Seligman 1970; evolutionary basis to learning what makes us ill)
Explain Edward Tolman’s latent learning effect
Mice had to learn to navigate around a maze in order to be rewarded at the end - reinforcement.
Group 1: Always reinforced
Group 2: Never reinforced
The ones that were always reinforced made significantly less errors compared to the others.
Group 3: started off never reinforced but became always reinforced after day 10. Steep drop in number of errors = suggests latent learning is possible and reinforcement is not always necessary but speeds up the process. Dormant learning is present.
How does Edward Tolman’s latent learning effect compare to the principles of associative learning?
Association by contiguity
But little contiguity between taste and illness
* Arbitrariness
No learning between taste and shock or click/light and illness
* Empty organism
Certain associations are easier established than others
(Preparedness principle (Seligman 1970))
* Learning through reinforcement
But latent learning in absence of reinforcement – reinforcement speeds up learning
