Mod 5 - Learning

LEARNING Summary 1: Introduction to Learning learning – The modification through experience of pre-existing behaviour and understanding. Learning happens when we: •

Experiencing events,

•

Observing relationships between those events, and

•

Noting regularity in the world around them.

Three main questions guiding psychological research: 1. Which events and relationships do people learn about? 2. What circumstances determine whether and how people learn? 3. Is learning a slow process requiring lots of practice, or does it involve sudden flashes of insight? novel stimuli – stimuli we have not experienced before sensitization – novel stimuli attracts exaggerated responses habituation – responses to unchanging stimuli decreases over time (eg. Ticking of the clock) dishabituation – the reappearance of original responses when the stimuli changes (eg. When the clock stops) Sensitization and habituation are non-associatative learning – learning results from the impact of one particular stimulus, where the person/animal does not associate one stimuli with another. Opponent-process theory – new stimulus events (esp those that arouse strong +ve or –ve emotions) disrupt the individual’s physiological state of equilibrium. But this disruption triggers an opposite/opponent process that counteracts the disruption and eventually restores equilibrium. If the arousing event occurs repeatedly, this opponent process get stronger and more rapid, eventually suppress the initial response, creating habituation. Eg. Sky diving, drugs addition. Law of association by contiguity (closeness in space or time) – this law of association says that stimuli or events that occur closely together in time become associated, such that in the future when one event occurs, the other will also come to mind. Behaviourism

Radical behaviourism – a kind of behaviourism describing the learning process in terms of the relationship b/w a stimulus and response – the S-R approach. It is sometimes called the black-box as the internal mental process cannot be seen. Watson’s contribution •

“given the response the stimuli can be predicted, and given the stimuli the response can be predicted”

•

Behaviourism differences among people stem mainly from their varying experiences, mediated by learning

Skinner’s contribution – 2 separate learning processes

•

Classical conditioning is a procedure in which a neutral stimulus is repeatedly paired with a stimulus that elicits a reflex or other response until the neutral stimulus alone comes to elicit a similar response.

•

Operant conditioning is a process by which the consequences of a response increase/decrease the likelihood that the response will occur again. i.e. reward and punishments shape/change behaviour

Cognitive perspective Watson and Skinner avoided explanation in terms of unseen events in the mind/brain. Contemporary psychologists introduced the S-O-R approach. Stimulus (environment)  Organism  Response (behaviour) This is called cognitive psychology – looks at how people represents, store, and use information. Cognitive psychology – the study of the mental processes by which information from the environment is modified, made meaningful, stored, retrieved, used, and communicated to others. Ecological perspective Ecological approach – An approach to perception maintaining that humans and other species are so well adapted to their natural environment that many aspects of the world are perceived without requiring higher-level analysis and inferences

Summary 2: Introduction to Classical Conditioning Learning by observing relations or associations b/w events in the world.

•

Pavlov’s (Russian physiologist) digestive processes of dogs

o unconditioned response (UCR) to food, it is a unconditioned stimulus (UCS). it involves an innate, reflexive response

o bell starts as neutral stimulus as dog does not respond to it

o then dog learnt the bell reliably predicts the arrival of food – the bell become a conditioned stimulus (CS) producing a conditioned response (CR).

•

Extinction – the gradual disappearance of a CR when a CS no longer predicts the appearance of an US.

•

Reconditioning – the quick relearning of a CR following extinction.

•

Spontaneous recovery – the reappearance of the CR after extinction and without further pairings of the conditioned and unconditioned stimuli.

•

Generalisation – a phenomenon in which a CR is elicited by stimuli that are similar but not identical to the CS. Eg. Sour milk and other dairy with similar odor

•

Discrimination – a process through which individuals learn to differentiate among similar stimuli and respond appropriately to each other. Eg. Other babies whimpering will not trigger a CR.

•

Signalling of Significant Events

o Timing – Forward, Backward and Simultaneous. The temporal spacing of the UCS and CS

o Predictability – conditioned stimulus reliability o

Signal Strength

o

Attention

o The blocking effect – an old stimuli blocking the learning of new stimuli o Second-Order Conditioning – eg. white coat hypertensive A phenomenon in which a CS acts like and UCS, creating CS out of events associated with it.

o Bio-preparedness – eg. conditioned taste aversion; nausea is more likely to be associated with food/drink then light/noise. Make sense in evolutionary terms. Animals & humans are innately prepared to learn aversions to certain stimuli. Summary 3: Applications and theories of Classical Conditioning Basis of some behaviour therapy techniques. •

Watson and Rayner – little Albert fears the rat (CS) because of the relationship it learnt b/w a loud noise (UCS). Generalise with rabbit and soft fluffy toy also.

•

Nowadays used frequently on advertising. eg. Beer and car commercials

•

Pavlov experiment to condition a dog to drug reaction to a non-drug stimulus

•

Recent research conditioned effects of stimuli paired with some drugs are the OPPOSITE of the most prominent unconditioned effects of the drugs. Eg. Morphine reduces sensitivity to pain, but it triggers the direct reflex, in turn, activates a reflective physiological response that counters the direct effects, ie increase sensitivities. Eg. Drug overdose when the amount use is actually normal, besides extinction, the usual drug-taking environment etc can have an impact to the sensitivity to the drug.

•

Operant response – actions a subject operate on the world to produce some effect. Eg pushing a level, bitten open a bag (for food), saying “please” to get things done. This is also called instrumental responses because they function like instruments, or tools.

•

Operant/instrumental conditioning – a process through which an organism learns to respond to the environment in a way that produces positive consequences and avoids negative ones.

•

Application

o Phobias are extreme fears of objects or situations that either are not objectively dangerous or are less dangerous than the phobic person’s reaction suggests. Classical conditioning can help treat phobias, even PTSD (posttraumatic stress disorder).

o Predator control – lithium laced muttons fed to wolves and coyotes so they do not eat the sheep anymore. o

Predicting Alzheimer’s Disease – air puff directed to the eye.

o Behaviour therapy – treatments that use classical conditioning principles to change behaviour. ie. Psychological problems are learnt behaviours that can be changed by taking actions to learn new ones. Behaviour approach & social-cognitive approach.

o Behaviour modification – treatments that use operant conditioning methods to change behaviour. Commonly used techniques:



Systematic desensitisation – a behavioural treatment for anxiety in which client visualise a graduated series of anxietyprovoking stimuli while remaining relaxed. Joseph Wolpe (1958) often used progressive relaxation training to prevent anxiety/phobias. Then, while relaxed, client is asked to imagine/presented an item from a desensitisation hierarchy.



Modelling – demonstrating desirable behaviours as a way of teaching them to clients. The combination of live modelling with gradual practice is called participant model. Assertiveness training and social skills training are methods for teaching clients how to interact with others more comfortably and effectively.



Positive Reinforcement – a therapy method that uses rewards to strengthen desirable behaviours. Using operant conditioning principles, they setup contingencies (rules).



In operant conditioning, extinction is the process of removing the reinforcers that normally follow a particular response – when behaviour doesn’t “pay off”, people usually stop it eg. Calling a phone number) Extinction is also the basis of flooding – a procedure for reducing anxiety that involves keeping a person in a feared, but harmless, situation. This prevents them from engaging in the normal rewarding pattern of escape. This is also called the exposure technique.



Aversion therapy/conditioning – a method that uses classical conditioning to create a negative response to a

particular stimulus. May not working for all clients and the effects are often temporary.

•



Punishment – an operant conditioning technique that weakens undesirable behaviours by following it with an unpleasant stimulus. It presents the unpleasant stimulus AFTER the undesirable response occurs.



Counter conditioning – Mary Cover-Jones (1924) introduces another stimulus the produces a response that is incompatible with the conditioned fear response.

Summary of views on CC

o Watson – strict behaviourist, CC is a simple linkage b/w stimuli and response (S-R)

o Pavlov – association b/w two stimuli is learnt (S-S). CS elicits a mental representation of UCS which produce CR.

o Experiments favour S-S (Anderson 2000). Eg. Robert Rescorla 1973 – rat fears light  mental re of loud sound  freezing

o Cognitive view – Rescorla – bell  expectation of food  tail wagging, food begging, salivation, etc. The expectancy theory. 

The CS must precede the UCS



The CS must signal heightened probability of occurrence of the UCS



Conditioning is ineffective when the animal already has a good predictor.

Summary 4: Introduction to Operant Conditioning Behaviour is controlled by its consequences.

•

Edward L. Thorndike 1874 – 1949 (1898)

o Puzzle box (eg. Hungry cat push a pedal to solve the puzzle that open the door to food. Learning by trial and error.

o The cat’s learning is governed by law of effect – a law stating that if a response made in the presence of a particular stimulus is followed by satisfaction, that response is more likely to be made the next time the stimulus is encountered. •

Skinner (1938) o

Emphasises that organism learns a response by operating on the environment, hence the process of learning these responses is called operant conditioning.

o Skinner box – the puzzle box measures learning in terms of whether an animal successfully completed trial and how long it took; the Skinner box measures learning in terms of how often an animal responds during a specified period of time, ie the rate of responding.

•

Operant – a response that has some effect on the world. Eg. When dog salivated, it has no effect on the buzzer or on whether food is presented. If a child says “I’m hungry” and then is fed, child has made an operate response that influences when food will appear.

•

Reinforcer – a stimulus event that increases the probability that the response that immediately preceded it will occur. o

Positive reinforcers strengthen a response if they are experienced after that response occurs. Eg. Food, smiles, money and other desirable outcomes.

o

Negative reinforcers are the removal of unpleasant stimuli such as pain, noise, threats.

o Punishment – presentation of an aversive stimulus or the removal of a pleasant stimulus. This reduces the frequency.

•

Escape/avoidance conditioning – a type of learning in which an organism learns to make a particular response in order to terminate an aversive stimulus. Techniques used for negative reinforcement.

o In escape conditioning the learnt response stops and aversive stimulus. eg. Take pills to stop pain, mute button for tv commercials

o In avoidance conditioning the learnt response prevents the aversive stimulus from occurring in the first place. Eg. Child apologies as soon as parents frown. This involves a marriage of classical and operant conditioning.

•

Discriminative stimuli – stimuli that signal whether reinforcement is available if a certain response is made.

•

Forming and strengthening operant behaviour

o Shaping – the process of reinforcing responses that come successively closer to the desired response. Eg. Dog tricks

o Secondary reinforcer – a reward that people or animal learnt to like. Eg. Money, praise, smile and other form of social approval (social reinforcers) Primary reinforcer – reinforcer that meet an organism’s basic needs such as food and water. Chaining – any stimuli can become a reinforcer.

o Delay and size of reinforcement – drinking: the good feelings are immediate, the hangover is delayed.

o Schedules of reinforcement (Skinner) 

Continuous reinforcement schedule – a pattern in which a reinforcer is delivered every time a particular response occurs. Eg. light switch



Partial reinforcement schedule – a pattern in which a reinforcer is administered only some of the time after a particular response occurs. Also called an intermittent reinforcement schedule. Eg. Workers’ salary



Fixed-ratio (FR) schedules – a partial reinforcement schedule that provides reinforcement following a fixed number of responses. Eg. Every 10 press on the level in Skinner box; Per piece paid worker. Characterised by strong “bursts” of responding.



Variable-ratio (VR) schedules – a partial reinforcement schedule that provides reinforcement after a varying number of responses. Eg. Gambling machine with average return set Leads to higher levels of responding than the FR schedule, and a post reinforcement pause. High resistance to extinction.



Fixed-interval (FI) schedule – a partial reinforcement schedule that provides reinforcement for the first response that occurs after some fixed time has passed since the last reward. Eg. Study cram. Leads to clustering of responses at about the time the reward is due.



Variable-interval (VI) schedule – a partial reinforcement schedule that provides reinforcement for the first response after varying periods of time. Leads to a steady rate of responding with no post-reinforcement pause. Strong resistant to extinction.



Partial reinforcement extinction effect – a phenomenon in which behaviours learned under a partial reinforcement schedule are more difficult to extinguish than behaviours learned on a continuous reinforcement schedule.

Summary 5: Applications and theories of Operant Conditioning Eg. “civilised”, gender roles, rewards and punishments, mental patients, drug addicts, etc 

Theoretical views of OC – contrast 2 viewpoints: the simple behaviourist account and a cognitive perspective.



1st ½ of the 20th century, in Nth America focus on behaviourism. Learning as resulting from automatic, unconscious formation or modification of associations.



Cognitive psychologist argues both CC & OC help organisms to detect causality, ie. Not just from automatic associations, but also complex mental process that underlie our adaptation to + understanding of the world around us.

o Learnt helplessness – learning that responses do not affect consequences, resulting in failure to try to exert control over the environment. In another word, learnt to give up.

o Means-End Analysis – a problem solving strategies that involves continuously asking where you are in relation to the final goal, and then deciding on the means by which can get one step closer. Using decomposition (breaking in smaller subgoals)

o Edward Tolman describes the learning as a means-end relationship. Dickinson and Dawson’s experiments supported the same view

o Reward contrast effects – shift in response rate when the size of the reward changes. Negative effects – less response when moved from large reward to small reward. Positive effects – increased response when moved from small reward to large reward. 

Organisms learnt to expect



Able to compare actual vs expectation



Evidence in mammals and birds but not fish and reptiles



Tolman’s cognitive maps – 3 identical rat mazes with varying reinforcements – the learning is not affected, the change is in performance. Latent Learning – learning that is not demonstrated at the time it occurs. Cognitive map – a mental representation of the environment. Tolman concluded that these maps developed naturally through experience, no overt response or reinforcement. Also the ability to reroute when encounters a block.



Insight – a sudden understanding about what is required to solve a problem. Maybe resulted from a “mental trial and error”



Observational learning – learning how to perform new behaviours by watching others. (learning by imitation); Vicarious learning – learning conditioned responses by watching what happens to others. Powerful source of socilaisation.



Great deals of learning occur, but only translate to behaviour when a reward is available.

Summary 6: Applications Observational learning and the ecological perspective



Albert Bandura and the “Bobo doll” – children not only learnt ways of being aggressive but also the general message.



Ecological perspective – learning must be understood in relation to the natural environment, or ecology, in which species evolved. Some behaviour must be learnt for survival need. Ie. in relation to natural selection o

Imprinting – newborn will follow/attach to the first moving object they see after hatching. Konrad Lorenz (1935)

o Food aversion learning – Garcia and colleagues (1968) – CC is not enough to explain (time delay did not affect learning)

o Innate biases – Seligman (1971) – innate fear of snakes, spiders and angry faces