Unit 8 Learning
Classical
- Also known as learning by association
- Ivan Pavlov, John Watson
Procedure
- Neutral Stimulus (already gets a response) is paired with an Unconditioned Stimulus (no response trained yet)
- Learning is already present/achieved
Result
- Neutral Stimulus becomes a Conditioned Stimulus and elicits a Conditioned Response
- Neutral = conditioned
Examples
- Dogs already salivated at the food, so the bell was introduced as a Neutral Stimulus, which became the Conditioned Stimulus, prompting the Conditioned Response of salivatio
Little Albert
- John Watson did a lot of studies on orphans
- Little Albert was the son of his employee who he threatened to fire if he couldn’t use her son
- Would cause a loud and startling sound every time Albert reached out for a rat
- Albert associates the rat with the startle response
- Also became afraid of Watson and white stuffed animals
Notes
- NS = Neutral Stimulus
- UcS = Unconditioned Stimulus
- UcR = Unconditioned Response
- CS = Conditioned Stimulus
- CR = Conditioned Response
- NS and US pair
- US and UR are always linked
- After NS links to US and therefore UR, NS is now CS and causes CR
- CR is not always the same as UR
- Acquisition
- Generally, the optimal acquisition schedule is trace conditioning, or delayed
- Acquiring, ‘getting,’ the training
- Extinction
- When the conditioned reaction no longer exists in association to the stimulus
- If there is no reinforcement or wasn’t strong enough in the first place, it can go extinct
- Spontaneous recovery
- The training suddenly comes back
- Usually not as strong as the initial pairing
- (Stimulus) generalization
- Other similar stimulus gets the same conditioned response
- (Stimulus) discrimination
- Having a conditioned response to the conditioned stimulus and only that stimulus
- Higher-order conditioning
- Adding a second set of classical conditions to an existing classical conditioned response
- The NS in the first becomes the US in the second one
Operant
- Learning by consequences
- B.F. Skinner, Edward Thorndike (Instrumental learning, law of effect)
Procedure
- Behavior is followed by a consequence of either reinforcement or punishment
- ABC method
- There is a motivation, the Antecedent
- This prompts trial-and-error Behavior
- The behavior receives a Consequence
Result
- The behavior increases or decreases in frequency
Examples
- Skinner’s pigeons
Notes
Reinforcement
- Always perceived as good
- Will increase the frequency of or the strength of a behavior
Punishment
- Always perceived as bad
- Will decrease or weaken a behavior
Comparison
| Classical | Operant | |
|---|---|---|
| Acquisition | (Acquisition phase first) Association is made between stimuli; called Respondent Behavior | (Shaping first) Learning that behaviors produce consequences; called Operant Behavior |
| Extinction | Association is broken (CR no longer occurs) | Responding decreases when reinforcement stops |
| Spontaneous Recovery | Association (CR) returns | Consequence regains meaning (Behavior frequency returns) |
| Generalization | Similar stimuli produce similar response | Response to similar stimuli also reinforced |
| Discrimination | Ability to distinguish between similar stimuli (Do not produce the same response) | Learning that some responses will be reinforced while others will not - Discriminative Stimulus |
| Higher-Order Conditioning vs. Chaining | Higher (Second) Order Conditioning; Connecting one association onto another, using an existing learned association | Chaining; Increasing the complexity of behaviors through reinforcement |
Reinforcement vs. Punishment
Positive Reinforcement
- Something perceived as ‘good’ is presented to strengthen behavior
- Sometimes the reward is biologically recognized as good because of its necessity
- Food, shelter
- Other rewards are conditioned to be seen as ‘good’
- Most material goods
Negative Reinforcement
- Removal of something perceived as ‘bad’ to strengthen behavior
- Removing a threat or stressor
Positive Punishment
- Something perceived as ‘bad’ is presented to weaken behavior
Negative Punishment
- Taking away something perceived as ‘good’ to weaken a behavior
- Omission Training
Escape and Avoidance Learning/Learned Helplessness
- Learning to escape or avoid something because of punishment
- Martin Seligman shocked dogs no matter what they did
- They couldn’t ever figure out the solution so they just gave up
- He also determined that if a creature can learn helplessness, they can also learn optimism
Behavior Modification
- Making and breaking habits
- Token economy
- Meaningless ‘token’ is assigned value through what it can buy/achieve
- Indoctrinating capitalism into children and the mentally ill, yay!
- Premack principle
- Probability of behavior
- Doing things you don’t want to do
- Why? Because we know what comes after
- Either we’re punished if we don’t or are rewarded if we do
- Impulse control
- Understanding delay of gratification
- Saving for something bigger vs. little pleasures
Social (Observational) Learning/Modeling
- Albert Bandura
Procedure
- Learning by watching
- Primates do this, some young animals mimic their parents, but mainly a human thing
- Observer watches Model to learn a behavior
Result
- The Observer learns the behavior and is able to perform it without watching anymore
Examples
Bandura’s Bobo Doll
- Called this the social cognitive theory
- He believed it wasn’t passive, that there was cognition/thinking going on in the Observer
- Also showed that aggressive actions doesn’t air out frustration, it created more
- Repeated exposure to aggressive modeling leads to higher instances of aggression
- Attention: One must be attentive to what action is being done
- The children had to be attentive to the video
- Retention: One must be able to recall what was observed in order to imitate it
- Reproduction: One must have the physical and mental capacity to do the activity to be imitated
- Motivation: One must have an incentive to do the behavior
Chameleon Effect and Prosocial Behavior
- The chameleon effect is essentially conformity
- It is not a bad thing to learn from models
- We tend to adopt postures, gestures, and mannerisms of interaction patterns
Insight/Cognitive Learning
- Wolfgang Kohler
Procedure
- Learning suddenly after no consequence, association, modeling, etc.
- Just a burst of cognitive processing
Result
- Quick and immediate solutions without strategy-based solving
Examples
- Chimps and pigeons
- Edward Tolman and the Cognitive Map
- Rats needed a reason to display what they have learned
- There’s ‘latent learning,’ learning that occurred previously but has no necessity to be shown yet
- Creatures create cognitive maps of their spatial environment