| Learning: | a relatively permanent change in behaviour, capability, or attitude that is acquired through experience and can't be attributed to illness, injury or maturation |
| Stimulus: | any event or object in the environment to which an organism responds; plural is stimuli |
| Classical Conditioning: | a process through which a response previously made only to a specific stimulus is made to another stimulus that has been paired repeatedly with the original stimulus |
CLASSICAL CONDITIONING
IVAN PAVLOV (1849-1936)
-discovered classical conditioning by accident
-physiologist studying digestion in dogs-measuring saliva
-noticed that dogs
started to salivate before food was tasted
-dogs drooled at sight of food and the sound of his footsteps
-he sounded a bell before presenting food---> dogs salivated -dogs were conditioned
to salivate to new stimulus
| Reflex | an involuntary response to a particular stimulus, like eye blink or salivation to food placed in mouth-made up of both stimulus and response |
| Conditioned reflex | a learned reflex rather than a naturally occurring one |
| Unconditioned Response (UR) | a response that is invariably elicited by the unconditioned stimulus without prior learning |
| Unconditioned Stimulus (US) | a stimulus that elicits a specific response without prior learning |
| Conditioned Stimulus (CS) | a neutral stimulus that after repeated pairing with an unconditioned stimulus, becomes associated with it and elicits a conditioned response |
| Conditioned Response (CR) | a response that comes to be elicited by a conditioned stimulus as a result of its repeated pairing with an unconditioned stimulus |
| Extinction | The weakening and often eventual disappearance of a learned response ( in classical conditioning, the conditioned response (CR) is weakened by repeated presentation without the unconditioned stimulus (US) |
| Spontaneous Recovery | The reappearance of an extinguished response (in a weaker form) when an organism is exposed to the original conditioned stimulus following a rest period |
ELEMENTS OF CLASSICAL CONDITIONING
1. UNCONDITIONED STIMULUS (US) -event that naturally triggers reflex (FOOD)
2. UNCONDITIONED RESPONSE (UR) (SALIVATE)
3. CONDITIONED STIMULUS (CS) - event paired with US (bell)
4. CONDITIONED RESPONSE
(CR) -learned reaction (salivate)
begin CS---> NO EFFECT US+CS--->effect CR -classical conditioning demonstrated
in all animals e.g. -sound of can opener/ sound of dog biscuit box
FACTORS INFLUENCING CLASSICAL CONDITIONING
1. The number of pairings- REPEATED PAIRINGS US+CS,US+CS...>learning -doesn't happen on single pairing -generally more pairings the stronger the conditioned response
2. The intensity of the unconditioned stimulus-if a conditioned stimulus is paired with a very strong unconditioned stimulus, the conditioned response will be stronger and acquired more rapidly compared to pairing with a weaker unconditioned stimulus
3. How reliably the conditioned
stimulus predicts the unconditioned stimulus-the neutral stimulus must reliably
predict the occurrence of the unconditioned stimulus
4. Spacing of Pairing-if pairing CS+US follows too rapidly--->slower
learning
-if pairing CS+US too far apart---->slower learning
-CS and US shouldn't occur alone -intermittent pairing reduces rate and strength
| Generalization | in classical conditioning, the tendency to make a conditioned response to a stimulus that is similar to the original conditioned stimulus; in operant conditioning the tendency to make the learned response to a stimulus that is similar to the one for which it was originally reinforced |
| Discrimination | the learned ability to distinguish between similar stimuli so that the conditioned response occurs only to the original conditioned stimulus but not to similar stimuli |
| Higher-order Conditioning | occurs when a neutral stimulus is paired with an existing conditioned stimulus, becomes associated with it, and gains the power to elicit the same conditioned response |
CLASSICAL CONDITIONING IN HUMANS
-e.g. Phobia through classical
conditioning-intense, irrational fear spider snake--
- Little Albert-fears can be unlearned by classical conditioning-pair rat with
pleasant experience-candy
| Taste Aversion | the dislike and/or avoidance of a particular food that has been associated with nausea or discomfort |
OPERANT CONDITIONING
--at turn of C while Pavlov
with dogs, Thorndike was using a puzzle box to study how
a cat learns -he confined a hungry cat in puzzle box with food just outside-where
cat could see and smell
-cat had to figure out how to open latch
-Thorndike timed escape
-cats learned to make necessary response more rapidly after an increasing number
of trials
ELEMENTS OF OPERANT CONDITIONING
1. Emitted Behaviour (Active)
vs triggered response in classical (Passive)-operating on environment
2. A consequence must follow the behaviour e.g. freedom, food, praise consequence
increases likelihood behaviour repeated
| Operant Conditioning | A type of learning in which the consequences of behaviour tend to modify that behaviour in the future (behaviour that is reinforced tends to be repeated; behavior that is ignored or punished is less likely to be repeated) |
| Reinforcer | anything that strengthens a response or increases the probability that it will occur |
| Shaping | gradually moulding a desired behaviour by reinforcing responses that become progressively closer to it; reinforcing successive approximations of the desired response |
| Skinner Box | invented by BF skinner for conducting experiments in operant conditioning, a soundproof operant conditioning chamber with a device for delivering food and either a bar for rats to press or a disk for pigeons to peck |
| Successive Approximations | a series of gradual training steps with each step becoming more like the desired response |
| Extinction | the weakening and often eventual disappearance of a learned response ( in operant conditioning the conditioned response is weakened by withholding of reinforcement |
Law of Effect -+ve effect-->behaviour repeated -ve effect-->behaviour suppressed -over time Rf could become neutral-eg-too much candy -difficult to know if something is Reinforcer or Punisher
| Generalization: | in operant conditioning it a subject responding to a similar stimulus with an action; the less similar the stimulus the lower the rate of the behaviour |
| Discriminative Stimulus: | a stimulus that signals whether a certain response or behaviour is likely to be followed by reward or punishment |
ESTABLISHING AN OPERANTLY CONDITIONED RESPONSE
-because behaviour is voluntary
not always easy -behaviour must be spontaneous to be rewarded/punished -imagine
waiting for lion to jump through fire hoop
1. Increase motivation-get cat hungry-place food just outside box
2. Reduce opportunities for irrelevant response-eg small bare space like Skinner
Box
3. Rf successive approximations--SHAPING -Shaping with Skinner box-Rf for turning
toward bar-Rf for touching bar -as with classical conditioning reaches point
of diminishing return
A CLOSER LOOK AT REINFORCEMENT
| Reinforcement: | an event that follows a response and increases the strength of the response and increases the strength of the response and/or the likelihood that it will be repeated |
| Positive Reinforcement: | a reward or pleasant consequence that follows a response and increases the probability that the response will be repeated |
| Negative Reinforcement: | the termination of an unpleasant stimulus after a response in order to increase the probability that the response will be repeated |
| Primary Reinforcer: | a reinforcer that fulfills a basic physical need for survival and does not depend on learning (UNLEARNED)(examples: food, water, sleep, termination of pain) |
| Secondary Reinforcer: | a neutral stimulus that becomes reinforcing after repeated pairings with other reinforcers (LEARNED) |
| Continuous Reinforcement: | reinforcement that is administered after every desired or correct response; the most effective method of conditioning a new response |
| Partial Reinforcement: | a pattern of reinforcement in which some portion, rather than 100% of the correct responses are reinforced |
positive RF-- any event that increases the probability of a response
e.g. praise, cookie, work hard for raise or promotion, child throws temper tantrum
to get candy
negative Rf--remove negative stimulus-rock baby so it won't cry, turn on air conditioner to get rid of heat, heroin addict going through withdrawal will do almost anything
Positive reinforcement--> adds something rewarding
Negative reinforcement-->subtracts something noxious
**Reinforcement always strengthens or encourages a behaviour
Schedules of Reinforcement
| Schedule of Reinforcement: | a systematic program for administering reinforcements that has a predictable effect on behaviour |
| Fixed- Ratio Schedule: | a schedule in which a reinforcer is administered after a fixed number of non-reinforced correct responses (e.g. a factory worker who is paid for units produced) effective for maintaining a high response rate-when larger ratios are used subjects tend to pause after each reinforcement |
| Variable -Ratio Schedule: | a schedule in which a reinforcer is administered on the basis of an average ratio after varying number of non-reinforced correct responses ( variable 30 responses rf may be reinforced after 10, 50, 30) Time of reinforcement unpredictable. Results in higher more stable rates versus fixed ratio (e.g. gambling, casino) |
| Fixed-Interval Schedule: | a schedule in which a reinforcer is administered following the first correct response after a fixed period has elapsed. ( person working on salary paid on certain day; study for test prior to test, decrease study, increase prior to next test))-typically a pause after Rf and increase in behaviour prior to time of RF-Produces lowest response rate |
| Variable- Interval Schedule: | a schedule in which a reinforcer is administered on the basis of an average time after the first correct response following a varying time of non-reinforcement- Rf after average time-maintains stable response but lower than ratio (e.g. Pop quiz) |
| Schedule of Reinforcement | Response Rate | Pattern of Responses | Resistance to Extinction |
| Fixed-Ratio | very high | steady response, low ratio, pause after Rf with high ratio | higher the ratio, more resistant to ex |
| Variable-Ratio | highest response rate | constant response no pauses | more resistant to extinction |
| Fixed Interval | lowest response rate | long pause after Rf, followed by gradual acceleration | longer the interval the more resistant to extinction |
| Variable-Interval | moderate | stable, uniform response | more resistant to extinction than fixed-interval |
| Partial-Reinforcement Effect: | the greater resistance to extinction that occurs when a portion, rather than 100 %, of the correct responses have been reinforced- results in greater resistance to extinction ( e.g. nagging child -parent gives in---> tougher to extinguish |
Accidental Reinforcement
-Superstitions -if action is accidentally reinforced we can get superstitious
behaviour
-BF Skinner put pigeon in Skinner box and at random times dropped
food in food cups-whatever bird had been doing at time tended to be repeated
-some psychologists believe that reinforcing with candy money, playtime for
acts that are intrinsically rewarding can undermine- mixed evidence
A Closer Look at Punishment
| Punishment: | the removal of a pleasant stimulus or the application of an unpleasant stimulus which tends to suppress a response |
-punishment can be powerful controller of behaviour
-eg fine for not reporting tax- fine and loss of point for traffic offense -an
unpleasant consequence reduces likelihood of repeat behaviour
-punishment different from -ve reinforcement
--RF strengthens behaviour by removing something--punishment adds something
to weaken behaviour
Is Punishment Effective?
For Punishment to be effective it must
1. be swift-delay won't work as well
2. be sufficient without being cruel
3. be consistent -punishment can be effective where behaviour is dangerous
Drawbacks of Punishment
1. it only suppresses undesired behaviour
-it doesn't prompt person to unlearn-doesn't teach a more desirable behaviour
-if threat of punishment removed, negative behaviour can return (speeder slows
when seeing police car)
2. punishment stirs up negative feelings (child scolded for mispronouncing words)
3. harsh behaviour may increase aggressive behaviour-anger, hostile
Punishment needs to be given carefully and with reinforcement of desirable behaviour
-sometimes after punishment administered a few times threat of punishment is
enough to stop behaviour
-person avoids possibility of punishing consequence ( carry umbrella)-consequence
can also be avoiding issue altogether-shun math
Escape and Avoidance Learning:
| Escape Learning: | learning to perform a behaviour because it terminates an aversive event |
| Avoidance Learning | learning to avoid events or conditions associated with dreaded
or aversive outcomes - involves classical conditioning and operant |
LEARNED HELPLESSNESS
| Learned Helplessness: | learned response of resigning oneself passively to aversive conditions rather than taking action to change, escape, or avoid them; learned through repeated exposure to inescapable or unavoidable aversive events |
-when avoidance of punishment
not possible some give up
LH
-Seligman studied LH in dogs
-placed dogs in chambers that delivered random shocks
-dogs in control group could escape by pushing panel -experimental group helpless
-now both groups put in different situation where they could escape by jumping
hurdle
- a warning light came on before each shock-dogs in control quickly learned
to avoid -experimental group just lay there listless depressed -once established
it can generalize to other situations
FACTORS SHARED BY CLASSICAL
AND OPERANT CONDITIONING
1. Both involve learning associations classical -pairing food and bell US +
CS operant- pairing action with consequence
2. Responses in both are under control of stimuli in environment classical-
triggered by light. bell operant- might be cued by flash of light (see police
car)
3. Both won't last forever if not periodically renewed -but can suddenly recur-EXTINCTION
4. New behaviours can build upon previously established ones
THE IMPORTANCE OF CONTINGENCIES
-both are associative forms of learning that involve contingencies -one thing
depends on another-to get diploma must pass x courses -to get paid must have
job
CONTINGENCIES IN CLASSICAL
CONDITIONING
-CS comes to be viewed as signal for US-must be close in time -if CS comes before
and sometimes after US won't work -can have backward conditioning tone follows
shock animal relax -predictive information is crucial to establish conditioned
response -experiment with rat -noise paired with shock -noise became CS-->CR
fear -a second stimulus-light added before noise-the noise shock contingency
blocked learning of new CS-no new cue
CONTINGENCIES IN OPERANT
CONDITIONING
-subject must learn contingency between act and result to increase behaviour
-fewer rewards are often better-partial or intermittent reinforcement results
in behaviour longer vs continuous reinforcement
WHY? Expectations-subject learns not to expect reward every time-continue in hope of reward-lottery-card game-slot machine -put money in vending machine get food- no food->no more money Fixed Interval- decrease study after test-increase as test nears -
Variable-interval schedule-eg. exams at unpredictable time-steady pace of study Fixed=Ratio schedule- worker paid on piece work
EXTINCTION AND SPONTANEOUS
RECOVERY
-but is learned response lost? No extinction does not erase learning -new learning
may interferes with previously learned response
EXTINCTION AND SPONTANEOUS
RECOVERY IN OPERANT CONDITIONING
-in operant removal of reinforcement results in increase in response initially;
then decline--vending machine pull lever again -response not unlearned, may
recur
Difficulty extinguishing
an operantly conditioned response depends on:
1. strength of original learning-takes longer
2. pattern of reinforcement- occasionally reinforced more resistant
3. variety of settings
4. complexity of behaviour
5. behaviour learned from punishment versus reinforcement harder to extinguish
How to speed up extinction:
-put learner in different situation from one where learned
-habit of home return when home again
STIMULUS CONTROL GENERALIZATION AND DISCRIMINATION
Generalization and Discrimination in Classical Conditioning
-Little Albert-white rat generalized to white rabbit-cotton
-Pavlov's dogs salivated to buzzer -math test anxiety- to anxiety with numbers
Generalization and Discrimination in Operant Conditioning -baby hugged and kissed for saying mama- may call everyone mama or nana-called response generalization
BEHAVIOUR MODIFICATION
AND NEW LEARNING BASED ON ORIGINAL LEARNING
| Behaviour Modification: |
the systematic application of the learning principles of operant conditioning, classical conditioning or observational learning to individuals or groups to eliminate undesirable behaviour/or encourage desirable behaviour (e.g. Time Out) |
| Token Economy: | a program that motivates and reinforces socially acceptable behaviours with tokens that can be exchanges for desired items or privileges |
Why is money such a good reinforcer?
-object with no value can become reinforcer- money-poker chip-token economies
COGNITIVE LEARNING
-some psychologists insist
only way to study learning is through conditioning
-others say mental activity crucial
How do you know how to start car? from watching another
How do we learn abstract concepts? Through Cognitive learning
-impossible to observe and measure directly-but can be inferred
-interest in cognitive learning began in 1930s when Edward Tolman
said we didn't need to show learning-he said there was Latent learning
Tolman experiment:
2 groups of hungry rats in maze
-first group found food at end-2nd group nothing
Group 1 learned maze faster -he took some from second group gave food at end
-they almost immediately ran maze as well as group 1.
He argued they had learned great deal about maze
-learning latent-given motivation change
-in response to Tolman, Thorndike proposed experiment
where rat is carried through maze and then rewarded at end
-he predicted it would not learn -neither did the experiment
-experiment done 20 years later 1956
-they carried rat over correct path -result-->-rats learned as well this
way
-rats had developed a cognitive map -learning involves more than stamped in
reinforcement -it involves formation of new mental images that may be reflected
in future behaviour
INSIGHT AND LEARNING
SETS
-Gestalt psychologist Kohler studied insight into a problem's
solution Kohler's experiment: placed bananas outside chimps cage - chimp
frustrated but then chimp looked in cage to see what he could use (stick) -or
banana hung out of reach-use boxes -previous learning used to provide insight
Harry Harlow experiment-put food under one of coloured box-monkey
quickly learned food was under red
-monkey had established a learning set -in both experiments animals had learned
more than task-they had learned how to learn
LEARNING BY OBSERVING
| Observational Learning: | learning by observing the behaviour of others and the consequences of that behaviour; learning by imitation |
| Modeling: | another name for observational learning |
| Model: | the individual who demonstrates a behaviour or serves as an example in observational learning |
-first time a person tries to drive car, person successfully inserts key and
presses gas
--How learned?-by watching others-
observational or vicarious learning observational examples: how to start lawn
mover, how to love,aggression, smoking
Who/what do we pay attention to? -someone who commands attention: parent, famous,peer
-behaviour has to be memorable -then we must do behaviour -we can learn without
overt change-change depends on motivation Motivation? consequence-teenager sees
adults drinking and having fun
Social Learning Theorist:
Albert Bandura experiment 1965
- he showed people can learn without being reinforced directly
-3 groups of children saw movie where adult walked up to Bozo doll and became
aggressive-hit with mallet-kick
Group 1: saw end where model rewarded with candy
Group 2: saw end where model scolded
Group 3: saw no consequence -after film children put in same room and was observed
-after 10 minutes experimenter offered reward to repeat behaviour
Result-all learned aggressive behaviour
-children who saw punishment suppressed behaviour
-punishment with aggression not a good idea
PLEASE NOTE:
Information on this site is authorized for use only by the students of this course.Students have permission to copy any of the content. For copyright information of the linked sites please see the respective authors.
copyright 2001 Karen E.Hamilton