Dummies Guide to the Learning approach
1) Our environment shapes our behaviour. Environmental factors act as stimuli and we respond
to them. We are born with a ‘blank slate’ upon which our lives are written based on our
experiences of the world. Experiences lead us to behave in particular ways, the role of genetics is seen as relatively
unimportant and does not for example restrict out ability to succeed, we all have equal potential to be anything we want to
2) Behaviour is measurable. We can set up a stimulus and observe and measure the response. Therefore
learning is observable and therefore can be studied scientifically. This can lead to general laws about our
behaviour such as the Law of Effect by Thorndike.
There are 3 types of learning: Classical conditioning, Operant conditioning and Social Learning.
Only applies to reflexive natural responses. E.g. Pavlov’s dogs
The unconditioned stimulus (UCS) (eg food) produces an unconditioned response (UCR) (eg salivation). By pairing a
neutral stimulus (NS) (eg bell) with the UCS an association is established and after several trials the NS becomes the
conditioned stimulus (CS) and now produces the conditioned response (CR) (salivation) in its own right. Once the CR-CS
link has been established it will need occasional links back to the UCS+NS to maintain the response, otherwise extinction
UCS e.g. food --> UCR e.g. salivation
UCS e.g. food + CS e.g. sound of bell --> UCR e.g. salivation
CS e.g. sound of bell --> CR e.g. salivation
Extinction occurs if UCS and CS not paired for a while, dog stops salivating.
Generalisation means associating a similar stimulus with the CR
Discrimination means associating only one stimulus with the CR
Spontaneous recovery is when a CR recurs in response to the CS after extinction
Delayed conditioning is where the NS is presented before the UCS but is still present as the UCS is
presented and is the most effective method in expts. Trace conditioning is where the NS again is present before the UCS but
stops before the UCS is presented. The time gap here is critical – if it is very short it can be moderately effective, the longer
the time lapse the less effective. Simultaneous conditioning is where the NS & UCS are presented simultaneously and is
slightly effective. Backwards conditioning is where the NS comes after the UCS and is considered not to work in nonhuman animals though it does in humans.
Evaluation: reliable, has practical applications such as systematic desensitisation and aversion therapy, valid,
much research only done on animals, does not explain all behaviour.
Classical conditioning in humans: Specific phobias are seen as an example of classical conditioning in humans, for
example it was discovered that a child afraid of sand had a sand pit near a garden gate, a local dog when passing the gate
snarled and snapped at the child causing fear (UCS-UCR) because the child was in the sand pit sand was the NS and the
A study demonstrating classical conditioning is Little Albert, Watson and Rayner (1920)
Systematic desensitisation used to treat phobias. You cannot feel 2 emotions at once so if UCS is relaxation, then
UCR is relaxation, if UCS paired with CS spider leads to UCR relaxation so CS spider leads to CR
Evaluation: works better for small phobias like agoraphobia, therapy works, patients need good
imagination, therapy doesn’t deal with cause of phobia only behaviour.
1) Little Albert, Watson and Rayner 1920
Aim: to explore how classical conditioning could be used to create a phobia in humans using CC principles.
Procedure: 11-month baby Albert, placid and emotionally stable. Banged metal bar to startle Albert and then linked noise
to Albert playing with pet rat. The NS in Watson and Rayner’s experiment was a white rat. Trials before the experiment had
shown that Albert did not mind the rat and certainly did not object to it. The UCS in the experiment was the noise made by
hitting an iron bar with a hammer just behind Albert. This produced a loud noise that Albert found very upsetting. On a
series of occasions, Watson and Rayner presented Albert with the rat and, when he noticed it, struck the metal bar behind his
head. Predictably, this caused Albert to become quite upset. After a few trials, they presented the rat on its own. Even
without the noise, Albert started crying. He had learned to associate the rat with the noise, and this had produced a
Noise (UCS) = Anxiety (UCR)
Noise(UCS) + Rat (NS) = Anxiety (UCR)
Rat (CS) = Anxiety (CR)
Results: After a few trails Albert agitated on seeing rat. it became clear that it wasn’t just rats that made Albert upset. His
anxiety response had generalised to some other objects – white furry ones – that were similar to the white rat.
Conclusion: Watson & Rayner concluded that they had succeeded in conditioning in an infant fear of an animal the child
would not ordinarily be frightened of. Stimulus generalisation also was claimed in that Albert transferred the fear to other
similar stimuli. From the fact that the conditioned response was still present after 31 days, Watson & Rayner concluded it
might last a lifetime. Ivan Pavlov had shown that Classical Conditioning occurs in dogs but Watson & Rayner were the
first to demonstrate it occurred in humans too.
Evaluation: Supported Pavlov’s findings, good controls, The study was carefully documented; witnesses helped to record
the data and there were strict controls. Only one variable was changed at a time. The extensive documentation meant the
study could have been replicated and, therefore, tested for reliability but low ecological validity because it was carried out
in a lab.
Ethical issues – Albert frightened. Albert’s mother appears not to have given fully-informed consent - though there
clearly was some degree of consent and an understanding of when he would be taken back by his mother
The researchers deliberately exposed Albert to psychological harm - causing him distress. They allowed him to rest in
between exposures to frightening stimuli but continued even when it was clear he was distressed. Hard to generalize to
Conditioning and Phobias
Behaviourists believe that phobias are an example of classical conditioning. What is required to produce a phobia is a
UCS that produces a strong emotional reaction, pain, for example, and a situation where that UCS can become
associated with a neutral stimulus. For example, suppose a person got bitten by a dog when they were a child:
Pain (UCS) _ Anxiety (UCR)
Pain (UCS) + Dog (NS) _ Anxiety (UCR)
Dog (CS) _ Anxiety (CR)
If that anxiety response generalises from that particular dog to all dogs then the result would be that the person became
anxious every time they saw a dog. In other words, they would have developed a phobia.
This is where a new behaviour is created or an existing behaviour removed as a result of
selective use of rewards and punishments. Behaviour is shaped by successive
reinforcements until the animal is doing precisely what is wanted. The principles of OC are:
Positive reinforcement to increase behaviour something good is given (rewards, food)
Negative reinforcement to increase behaviour something bad is taken away (electric shock)
Negative Punishment to decrease behaviour something bad is given (smack, detention)
Positive punishment to decrease behaviour something good is taken away (PS3, mobile)
Primary reinforcers: are linked to basic needs (food, water, sex and warmth),
Secondary reinforcers: something that can satisfy a basic need (money, tokens, stickers, praise, extra time)
Continuous reinforcement where the organism is reinforced every time the behaviour is performed is not seen as a very
effective as the behaviour will stop almost immediately if the reinforcer is missed and the rate of responding is not very high
for most reinforcers.
Fixed ratio schedule is where the reinforcer is given for every nth time the action is performed,
Variable ratio there is no predictability about when the reinforcer will occur. This produces the highest level of consistent
performance, is very resistant to extinction and is the reinforcement schedule produced
Fixed interval is where the reinforcer comes after a set amount of time. This produces a
performance pattern where the rate of responding rapidly increases as the time for the
reinforcer is approached; this is followed by inactivity after reinforcement and then a
build up again as the next time point approaches.
Variable interval reinforcement comes after a period of time that varies but averages out
at a certain level; it produces a slow but very steady rate of responding and is resistant to
Extinction, generalisation and discrimination all occur in operant conditioning.
Evaluation: studies use experimental method and controls so they are scientific and cause and effect
conclusions can be drawn. Both CC and OC can be used in therapies so they have practical applications. A lot of
it is common sense and useful when applied to learning in schools etc and token economy
Studies use animals so generalization and credibility is in doubt. Studies are lab experiments and use animals
so validity is questionable too. Many studies use animals generalised to humans, does not explain
genetic/biological ones, adults don’t always reinforce young children so how do children know how to use the
A study demonstrating operant conditioning is Skinners study on superstitious behaviour in pigeons.
SOCIAL LEARNING THEORY
We learn from people we look up to and identify with. SLT recognises that behaviours often occur that have not been
reinforced but merely observed in others. The individual observes a behaviour being performed by a model (another
person) and also notes the consequences of their actions.
The observer then imitates or models (copies) the behaviour that they have seen. There is no cognition or planning
involved, it is just as mechanistic as operant or classical conditioning, however the likelihood of the behaviour observed
being imitated depends on the consequences of the activity for the model. If the model was rewarded for their actions then
the likelihood of imitation is increased. Whereas, if they are punished for a behaviour then reproduction is unlikely.
(Athough research by Bandura showed that they could reproduce the actions if requested to do so). Bandura suggests we are
motivated to imitate in order to also gain the reinforcers that we saw the model receive. Vicarious learning is carrying out a
behaviour that was previously observed and the model was rewarded for it.
For successful SLT to occur the learner needs to pay attention to the important parts of the observed action, retain that
information in memory, this can be difficult or beyond capability if the actions are complex and/or the observer young. The
observer needs to be motivated to both observe and imitate the actions and have the physical ability to copy the behaviour.
• Observing - paying attention to what someone is doing
• Remembering - recording the information in memory
• Motivation – the consequences of the behaviour
• Imitation – whether it is repeated depends on whether the model was
rewarded or punished
The nature of the model and the perceived relationship between the model and the observer will also affect
how likely imitation is to occur. Models are more likely to be imitated if the observer can identify with them, respects,
admires or looks up to them, if the model and their actions are seen as relevant and if the behaviour observed is seen as
consistent with instructions. Observers are more likely to imitate a model if they have relatively low self-esteem and high
dependency on those around them which is why SLT is a more powerful explanation for childrens behaviour.
Evaluation: behaviour can be tested in experimental conditions and so are developed using objective scientific
methods. Animal studies show that SLT can also explain animal behaviour and therefore
reliable. Incorporates operant conditioning and cognition, considers motivation, reliable
studies provide evidence.
It should mean that different cultures show different gender behaviour but developmental
behaviours are similar between cultures. In newborn babies there are gender differences that
cannot be learned so not all differences can be explained by learning theories, studies not
valid, genetic elements not considered.
2) The Bobo Doll study, Bandura, Ross & Ross 1961
Aim: to see whether young children will imitate behaviour they have seen, especially if that behaviour was
rewarded or not
Method: Laboratory Experiment at Stanford University. 8 experimental groups in 4 conditions.(plus control group)
MALE ROLE MODEL
FEMALE ROLE MODEL
MALE ROLE MODEL FEMALE ROLE MODEL
Procedure: 72 children aged 3 –5 yrs matched for aggression before the study started. Some groups watched
aggressive behaviour; some non-aggressive behaviour and control group watched neither.
Children playing in a room when adult entered and either behaved aggressively or non-aggressively. Children
then put in slightly aggressive state by being told they could not play with certain toys. Then behaviour observed
with access to a Bobo doll and child was observed.
Results: Children in non-aggressive state showed almost no aggression, (70%). Those that watched aggressive
models showed physical and verbal aggression imitating model. Male model copied more overall but boys more
Conclusion: Children watching adults behaving aggressively are more likely to imitate aggression so
observational learning does take place. Children also imitated non-aggressive behaviour, which led to less
aggression. A male adult showing aggressive behaviour is copied more than a female adult aggressive model.
Girls are more verbally aggressive.
Evaluation: Controlled experiment with cause and effect conclusions. High reliability because of inter-rater
observation by judges. One judge did not know which condition a child had been in so bias was reduced.
Practical applications of TV viewing, replicable.
Limited sample, not valid because situation was not natural. Children may have thought they had to hit the doll.
Ethical issues of children observing verbal and physical aggressive acts and repeating them. How these were
dealt with was not explained.
Applications and implications of social learning theory
• Copy cat hijackings. Air hijackings were unknown in the US prior to 1961. Then some Cuban airlines planes were
hijacked which sparked off a wave of hijackings culminating in a peak of 87 hijackings in 1969 (Mischel, 1986).
Phobias. Observing somebody else being scared of something is enough to start a phobia. Vicarious modeling can be used to
1) Laboratory experiments. There is an IV manipulated by the experimenter and a DV that can be measured, there is
control of all other variables and the experiment takes place in a controlled artificial environment.
Advantage: allows cause & effect to be establishes because of good control.
Disadvantage: artificial environment means that behaviour may not be realistic so not valid.
2) Animal learning studies. Non-human animals are used to discover the mechanisms involved in learning a new
behaviour. These are usually lab experiments. It is believed that the principles of learning can be extracted from such studies
and applied to humans.
Advantage: animals are less complex and their environments more readily controlled than in humans so easier to establish
what is happening as there are fewer confounding variables, animals mature more quickly so long term effects seen sooner,
animals are relatively cheap and easy to use.
Disadvantage: only valid if believe that can apply animal learning to humans, humans are far more complex and
generalisation not always valid, humans bring emotion, values and more complex understanding to their learning situations,
ethical issues, still an artificial situation.
Token economy is an intensive behaviour modification technique that tends to be used in schools with
behavioural problems or learning difficulties, young offender and mental institutions. In other words, in settings
where behaviour needs to be quickly shaped.
Ayllon & Azrin devised a programme where appropriate behaviours are reinforced with tokens. Tokens are secondary
reinforcers and can be exchanged for primary reinforcers such as food. The tokens can be exchanged for privileges. As the
behaviour becomes more firmly established earning the tokens will get harder. Used in mental institutions and prisons.
Research by Paul and Lentz (1977) found that hospitalized people with schizophrenia showed significant
improvement in social skills. However if a patient/client has become used to an institution that rewards ‘good’ or
appropriate behaviour, when they are re-integrated back into the community they may well expect the same
treatment. Not receiving positive reinforcement may result in a relapse of behaviour.
Evaluation: works in institutions as behaviour needs to be monitored, thus not so good elsewhere. Loss of privacy may be a
problem as is the fact that will get more attention if behaviour monitored. Danger that rights become seen as privileges.
Effective in institutions but will not necessarily transfer out into the community. There is also the problem that individuals
will only do what is needed in order to gain the tokens. Gives power to those administering the system and this power can be
abused. Training of staff to use this programme is easy and cheap.
Some commentators argue that token economies are nothing more than a form of social control – compliance
Gender appropriate behaviour is seen as being operantly conditioned in humans. Children are reinforced for
playing in appropriate ways and praised for gender appropriate behaviour, e.g. girls showing caring and
considerate behaviour, boys being assertive, it could be argued that the acceptance of boys not concentrating on
work but preferring to play football by the term "boys will be boys" is part of the operant conditioning of this type
of behaviour. Shaping is used by reinforcing behaviour gradually e.g. getting a child to tidy his/her room by
rewarding them every time they do it but changing the rewards (praise, chocolate, watching a film, little treat etc).
According to SLT gender is shown by role models and imitated like other behaviour. Gender is one way of
identifying with the model and people are more likely to respond to the child positively if the behaviour was
gender appropriate. Any other child watching the behaviour being positively rewarded for gender appropriate
behaviour is vicariously learning to do the same. For example, imagine three siblings, James (4 years), John (5
years) and Sarah (6 years). Sarah and John play ‘dressing up’ and both put on dresses. Their dad reinforces
Sarah for this, by saying she looks pretty but punishes John by saying he looks silly and boys should not dress that
way. In future, Sarah is more likely to wear dresses and John is less likely. James, who has been watching all this,
is unlikely to imitate the behaviour of wearing a dress because he has seen his brother (who he perceives as
similar to himself) getting punished for doing it.
Social learning theories use two main methods to studying gender development. Some have conducted
laboratory experimental studies, (Bandura) where they have manipulated various features of the behaviour of
models children were shown and then measured the effect on an aspect of the children’s behaviour. Other
researchers have conducted naturalistic observations where they have observed children and adults in their
natural environment and recorded how the adults respond with reinforcement and punishment to the children’s
KEY ISSUE – the influence of advertising on people’s behaviour
An issue of interest is the effect of advertising and how advertising is successful in leading people to want to buy
the product. There is a lot of advertising, for example, on television, as well as in other media, and it costs
companies a lot of money so it is assumed that it is worthwhile. This in turn assumes that the advert leads the
person to buy the product and this idea rests on learning theory. The question is how does advertising work?
Advertising works using classical conditioning principles. An unconditioned stimulus gives an unconditioned
response, so people cannot help but respond that way. If this idea is extended to pair something that a company
wants to sell with the unconditioned stimulus, then after a few trials, that pairing will result in the person giving
that unconditioned response — though it would then be conditioned. For example, if a beautiful girl gives a man a
‘desirable’ response, then advertisements can pair a beautiful girl with their product (perhaps a car) and the car
will then give the ‘desirable’ response. This is what adverts do and responses include salivation to food as well as
sexual responses such as the one outlined here.
Social learning theory can also help to explain how adverts work. Bandura has shown that people imitate role
models. People observe behaviour and if they pay attention and are motivated to repeat the behaviour, they will
imitate it. So if a role model uses a certain shampoo someone watching is likely to buy that shampoo. This is
particularly true if the model is like the person watching in terms, for example, of age and gender. And also if the
model is seen to be rewarded for the behaviour, such as being a celebrity.
Operant conditioning explains that behaviour is repeated if rewarded, so if the person buying the product is not
rewarded, perhaps if the shampoo is not very good, presumably the behaviour would not be repeated, but the
advert is there just to get people to buy the product in the first place. It could be argued that classical and operant
conditioning principles tend to come from using animals in experiments and as humans are different from
animals, such as in their problem-solving skills, it might not be appropriate to generalise the findings from
animals to humans. Also Bandura studied children when he developed the idea of social learning, so perhaps it is
different with adults, though the principles are still used with adults.
KEY ISSUE – the influence of role models on Anorexia
One issue is whether role models have an influence on whether someone develops anorexia or not. Learning
theories can help to explain anorexia. SLT suggests that people imitate role models, especially those they see as
relevant to themselves. One concept from the learning approach is identification. When someone identifies with
a role model they are likely to imitate their behaviour. It is therefore likely that teenage girls will imitate female
models and media celebrities where there is a trend to be very slim. Studies by Bandura have shown that girls
copy female models and boys copy male models, so if female role models are slim then girls are likely to want to
be slim. If someone observes behaviour but does not identify with the role model they are not so likely to
perform the behaviour. Girls who want to be slim are likely to stop eating and can develop eating disorders such
as anorexia. Another concept from the learning approach is reinforcement. If a role model is reinforced for
being slim, such as being praised, paid more or featured a lot in the media, then they might be imitated more.
Studies by Bandura have shown that behaviour that is rewarded is likely to be imitated more, such as in vicarious
learning. There is also negative reinforcement for being fat, through criticism and teasing, to avoid being
teased, fat children might starve themselves to slim down which may turn into anorexia. So not wanting to be fat
to avoid criticism and wanting to be slim to get praise, might be two types of reinforcement that help to explain
anorexia. However, anorexia could also be explained in a different way. The psychodynamic approach suggests
that a girl might starve herself to avoid growing up because she is fixated at a certain psychosexual stage.
Nevertheless anorexia is found around the world between different cultures and cross cultural studies support the
idea that anorexia is learned.