Chapter 8 Flashcards
The Adaptive Mind: Learning
the three categories of behaviours
- Reflexes are inevitable, involuntary responses to stimuli that are controlled by nervous system circuits located in the spinal cord and brainstem. They are beneficial by promoting welfare (e.g. pulling away when touching a hot stovetop), but have the disadvantage of being inflexible.
- Instincts (a.k.a. fixed action patterns) are inborn patterns of behaviour elicited by environmental stimuli. They share the reliability and lack of dependence on experience to reflexes, but are more complex; requiring much more neurons.
- Learning is a relatively permanent change in behaviour or the capacity for behaviour due to experience. Our behaviour changes as we mature.
Most human behaviour results from…
learning
he believed that human behaviour simply appears more complex and thoughtful because we often face the need to choose among competing instincts
William James (1887)
imprinting
when a young organism bonds with adults (e.g. baby swans follow their parents)
three categories of learning
- Associative learning is the formation of associations, or connections, among stimuli and behaviours; “If A happens, B is likely to follow.”
- Non-associative learning involves changes in the magnitude of responses to stimuli, rather than the formation of connections between them.
- Observational learning (a.k.a. social learning/modelling) occurs when one organism watches the actions of another organism, and provides the advantage of transmitting information across generations within families and cultures.
two types of associative learning
- Classical conditioning (a.k.a. Pavlovian conditioning) is when associations are formed between two stimuli that occur sequentially in time (e.g. if a child sees a bee and gets stung, the child forms a connection between seeing bees and the pain of bee stings; leading to them feeling frightened the next time they see a bee).
- Operant conditioning (a.k.a. instrumental conditioning) is when associations are formed between behaviours and their outcomes (e.g. if you park illegally and get a hefty ticket, you are likely to pay more attention the next time you park).
two types of non-associative learning
- Habituation is when reactions to repeated, unchanging, harmless stimuli decrease (e.g. someone who is not accustomed to young children may find it alarming to witness two kids who are screaming and rolling on the ground, whereas their mother (who is accustomed to the behaviour) wouldn’t bat an eye).
- Sensitization is the increased reaction to many stimuli following exposure to one strong stimulus (e.g. those who experience an earthquake experience exaggerated responses to movement, light, or noise).
Habituation ensures precious resources aren’t wasted when monitoring low-priority stimuli, whereas sensitization is useful in detecting high-priority stimuli.
implicit vs. explicit learning
Implicit learning occurs in the absence of conscious awareness, whereas explicit learning (what typically comes to mind when we think of “learning”) involves conscious awareness.
stimuli related to classical conditioning
Conditioned stimuli (CS) refer to environmental events whose significance is learned through classical conditioning, whereas unconditioned stimuli (UCS) elicit a response without prior experience, due to an innate biological significance.
In Pavlov’s experiment, dogs heard the sound of a ticking metronome (CS) before food (UCS) appeared.
responses related to classical conditioning
Conditioned responses (CRs) are learned through classical conditioning, whereas unconditioned responses (UCRs) require no previous experience.
acquisition
the development of a CR; it must be contiguous (closeness in time) and contigent (correlation between the CS and UCS)
CONTIGUITY:
- If the CS occured long before the UCS, the organism may not view the two as related.
- If the CS occured simultaneously or after the UCS, it would not be a useful signal.
CONTINGENCY:
- Learning about a reliable signal is easier than learning about a signal that occurs only sometimes.
extinction
the reduction of a learned response
- in classical conditioning, extinction occurs when the association between the CS and the UCS is broken
- in operant conditioning, extinction occurs when the consequence no longer follows the learned behaviour
spontaneous recovery
during extinction training, the reappearance of CRs after periods of rest; it takes time for the new learning to replace the old
inhibition
when a CS predicts the non-occurrence of an UCS
e.g. A predator at a watering hole in hunting mode is an excitatory CS eliciting fear, but a predator relaxing after a recent kill and calmly drinking water is an inhibitory CS that tells the animal that the predator is unlikely to attack soon.
stimuli generalizations
the tendency to respond to stimuli that are similar to an original CS
e.g. If our ancestors had one bad experience with a lion, it would make sense to avoid all lions, and animals with lion-like characteristics.
discrimination
a learned ability to distinguish between stimuli
e.g. If you present food following a high tone, but never a low tone, a dog will initially learn to salivate to both pitched tones, but as learning progresses, will eventually learn to differentiate between the abilities of the two stimuli to predict food.
higher-order conditioning
learning in which stimuli associated with a CS elicit CRs
e.g. A person who was bitten by a dog might show fear the next time the dog is seen (CS). The sight of the dog could subsequently act more like an UCR, and produce fear in response to other stimuli, like hearing the dog bark, that might signal the dog’s appearance.
latent inhibition
the slower learning that occurs when a CS is already familiar, compared to when the CS is unfamiliar
e.g. If you’ve eaten lots of pizzas over time and get sick after eating one, you’d probably not associate the pizza with sickness. However, if you got sick the first time you ate chocolate-covered ants, you’d associate the two.
the Rescorla-Wagner model
Early in training, learning proceeds more rapidly when the association (between the CS and UCS) is new and surprising. However, later in training, learning is slower as the association is now familiar and no longer surprising.
taste aversion
when the sight, smell, or flavour of a food (CS) has been paired in the past with illness (UCS), resulting in the dislike of said food (CR)
Garcia and Koelling’s rat experiment
Rats were given either sweetened water (“tasty water”), or normal water that triggered a light and clicking sound when drank (“bright-noisy water”). The rats were then either injected with lithium chloride to induce nausea, or were shocked. Rats that drank tasty water made quicker associations with the nausea, and rats that drank bright-noisy water made quicker associations with the shock.
This experiment had profound implications on classical conditioning, and behaviourism as a whole, as it challenged Pavlov’s views that stimuli could be easily interchangeable and prompted a renewed interest in the biological preparedness to learn things.
Albert’s white rat experiment, and its significance
In 1920, John B. Watson and Rosalie Raynor conducted an experiment with 9-month-old Albert. When Albert would play with a white lab rat (CS), Watson and Raynor made a loud noise (UCS) by hitting a steel bar with a hammer. Albert would then get frightened by the noise (UCR). A week later, when Albert was again offered the rat, he was afraid (CR); his fear generalized to other white, furry objects, like rabbits, dogs, and fur coats.
This experiment demonstrated that fears could potentially result from classical conditioning.
exposure therapy (flooding)
treating phobias by exposing people to fear-producing stimuli in a manner that is safe, until they no longer respond (i.e. extinction)
counter-conditioning, and Peter’s leporiphobia
the substitution of one CR for another opposite response
e.g. In a 1924 experiment, Mary Cover Jones treated Peter, a 3-year old with a phobia of rabbits, by associating food (the new UCS) with the presence of a rabbit (CS). Eventually, Peter was able to pet the rabbit while eating.
two applications of counter-conditioning
- Aversion therapy is when a CS formerly paired with a pleasurable UCS, is instead paired with an unpleasant UCS (e.g. a compound containing silver interacts with nicotine to make tobacco cigarettes taste bad, and make smokers quit more easily).
- Systematic desensitization is when people relax while being exposed to stimuli that elicit fear. Once the person is relaxed, the fear stimulus is gradually introduced.
addiction and classical conditioning
A major challenge for those recovering from addiction is that environmental cues (CSs) associated with the effects of substance use (UCSs) can elicite cravings (CRs) (e.g. a visit to an alcoholic’s favourite bar can lead to relapse).
latent inhibition and prejudice
If children grow up with little exposure to people outside of their own race, they will associate their own (familiar) race with other attributes much slower, but will quickly form stronger associations with other (unfamiliar) races much quicker (i.e. news reports about crime).
Increased creativity is associated with a ________ latent inhibition.
reduced
Skinner box
A specially adapted cage, named after it’s creator B. F. Skinner, that allows researchers to investigate the results of reinforcement and punishment on the likelihood that the rat will press a bar.
the four classes of consequences
- positive reinforcement
- negative reinforcement
- positive punishment
- negative punishment
reinforcement vs. punishment
Reinforcement increases the frequency of a behaviour, whereas punishment decreases the frequency of a behaviour.
positive reinforcement
increases the frequency of an associated behaviour by providing a desired outcome
e.g. Candy is given when trying to teach kids with autism how to speak.
the Premack principle
whatever behaviour an organism spends the most time and energy doing is likely to be important to that organism
two types of reinforcers
- Primary reinforcers have natural roles in survival (e.g. food).
- Conditioned reinforcers (a.k.a. secondary reinforcers) gain value from being associated with other things that are valued (e.g. if you always say “good dog” before giving your dog a treat, the use of the phrase “good dog” can reinforce the dog’s voluntary behaviour).
negative reinforcement
increases the frequency of behaviours that allow an organism to avoid unpleasant consequences (i.e. escape/avoidance behaviours)
e.g. We open umbrellas to avoid getting wet.
positive punishment
eliminates or decreases the frequency of behaviours by applying an aversive stimulus
e.g. A rat will stop pressing a bar in a Skinner box if each press results in an electric shock.
negative punishment
eliminates or decreases the frequency of behaviours by removing something desirable whenever the target behaviour occurs
e.g. If food is made available in a Skinner box as long as the rat doesn’t press the bar, the rat will stop pressing the bar.
the three conditions that must be met for punishment to have observable effects on behaviour
- significance: if a punisher is too mild for a particular individual, there is little incentive for that person to change behaviour
- immediacy: immediate punishments are more effective than delayed punishments
- consistency: effective punishment must be uniform
Skinner’s proposed alternative to punishment
extinction; learned behaviours stop when they are no longer followed by a reinforced consequence
continuous reinforcement
reinforcing a behaviour every time it occurs
drawback to continuous reinforcement
After a new behaviour is being learned, it’s inconvenient to continue with continuous reinforcement.
e.g. If dog owners want their dogs to walk with a loose leash, once the skill is learned, it doesn’t make sense to continue reinforcing the skill with treats.
partial reinforcement
the reinforcement of a desired behaviour on some occasions, but not others
ratio schedules vs. interval schedules
Ratio schedules depend on the number of responses produced, whereas interval schedules depend on the passage of time.
fixed ratio (FR) schedule
a reinforcement schedule in which reinforcement occurs following a set number of behaviours (e.g. restaurants where you must make a certain number of purchases before getting a reward)
- continuous reinforcement has an FR of 1 (one behaviour per reinforcer)
- as the amount of work for each reward is increases, responding becomes slower
variable ratio (VR) schedule
a reinforcement schedule in which reinforcement occurs following a variable number of behaviours (e.g. slot machines)
fixed interval (FI) schedule
a reinforcement schedule in which the time that must pass before reinforcement becomes available following a single response is set to a certain amount
e.g. If a class has a test every three weeks, students are likely to increase their studying right before the test, rather than studying consistently throughout the term.
variable interval (VI) schedule
a reinforcement schedule in which the interval of time between reinforcements is allowed to fluctuate around some average amount over the course of a session (e.g. random pop quizzes)
the partial reinforcement effect in extinction
extinction occurs more rapidly following continuous reinforcement, compared to that following partial reinforcement
the two factors that influence the partial reinforcement effect in extinction
- the transition from a continuous schedule to extinction is more obvious than the transition from a partial schedule to extinction (e.g. if you are accustomed to being paid for a babysitting job everytime you work, you’ll notice when you haven’t gotten paid)
- partial schedules teach organisms to persist in the face of non-reinforcement (e.g. when working regularly on a large project despite a much-delayed grade, persistence is key)
the matching law
when exposed to 2+ reinforcement schedules simultaneously, preference is shown to the behavior that achieves the highest amount of reinforcement
e.g. People spend so much time playing video games, rather than spending time in real-life socializing, as they find it rewarding and satisfying.
the method of successive approximations (shaping)
a method for increasing the frequency of behaviours that never or rarely occur
Shaping begins with the reinforcement of spontaneous behaviours that are similar to the target behaviour we want to train. As training continues, stricter requirements are given for reinforcement until the exact behaviour occurs.
An example of shaping with table manners:
1. “Thank you for picking up the spoon.”
2. “Thank you for putting the spoon in the food.”
3. Thank you for using the spoon to eat your soup.”
shaping and the right amount of reinforcement
- With too much reinforcement, learning stops because there is no incentive for change.
- With too little reinforcement, the learner becomes discouraged.
chaining
- breaking down a complex behaviour into manageable steps
- can be done forwards or backwards
- useful for teaching new skills to children with special needs
latent learning
learning that occurs in the absence of reinforcement; there is a difference between what has been learned and what is performed (e.g. doing badly on a test despite learning a lot of material)
Tolman’s Maze, and his rebuttal against behaviourism
Edward Tolman placed food in the goal boxes of mazes for rats to find, and then blocked the path in a maze to find food, but allowed the rats to choose from a number of additional paths.
Early behaviourists believe that the rats learned a simple operant “turn here for food” association, and would hence choose paths most similar to the trained one. However, the rats didn’t do that, and instead showed evidence of forming cognitive maps; they were willing to turn in a different direction if that path led to food.
social facilitation
the phenomenon in which for simple tasks (e.g. pedalling a bike), the presence of others makes us perform better
token economy
an application of operant conditioning in which tokens that can be exchanged for other reinforcers are used to increase the frequency of desirable behaviours (e.g. money as tokens)
behaviour therapy
the use of operant conditioning concepts (i.e. extinction, reward) to treat psychological and psychiatric disorders
imitation
copying behaviour that is unlikely to occur naturally and spontaneously
the four necessary cognitive processes in the modeling of others’ behaviour
- Models that get our attention are more likely to elicit imitation.
- There must be retention of what the model did.
- Reproduction of the behaviour must be possible.
- A person must have a motivation for imitating the behaviour.
memes
the basic unit of cultural transmission