Chapter 4
Learning Theory
Perspectives on
Personality
Key themes
Historical learning theory
approaches to personality
Pavlov and classical
conditioning
Watson and behaviourism
Skinner and operant
conditioning
Integrative personality theory
of Dollard and Miller
Social cognitive approaches of
Bandura
The concept of self-efficacy
Rotter and the locus of control
Mischel and social learning
theory
Learning outcomes
After studying this chapter you should:
Understand the principles underlying the learning theory approach to
personality
Be aware of the opposing views about whether differences in
behaviour are learnt or result from differences in personality
Be able to identify the learning theorists who have contributed to
personality theory
Understand the principles of classical conditioning and some of its
applications
Appreciate Skinner’s approach to psychology
Understand the principles underlying operant conditioning
Be aware of the work of Dollard and Miller, who attempted to
integrate psychodynamic and behavioural concepts within a learning
theory framework
Understand Bandura’s social learning approach to personality
Appreciate the concept of self-efficacy
Have developed an understanding of the concept of locus of control
Understand the contribution that Mischel has made to personality
research
Appreciate the person-situation debate in personality
Know how to critically evaluate learning theory approaches to
explaining personality
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 70
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 71
Do you love parties and never miss one, or do parties
make you anxious so that you avoid them if possible?
What causes these differences? The concept of personality,
as we have seen, is used to help explain such
differences in behaviour between individuals. The theories
we have previously examined suggest that differences
in the personality structures that are said to
exist within each individual interact to produce differences
in behaviour. Our behaviour is driven by inner
motives such as instincts, unconscious drives, feelings
of inferiority and so on, that all shape our personality.
Based on what we have read so far, we might
well claim that the person who enjoys parties and interacting
with others does so because they have an
outgoing, sociable personality. They are driven by an
inner need to be with other people and are not so
comfortable in their own company. There are alternative
explanations for personality, and this chapter is
about a series of theoretical approaches that adopt a
radically different view. These theories reject the idea
of our behaviour being directed by inner motives, suggesting
instead that all our behaviour is learned.
Individual differences in behaviour are the result of the
different learning experiences that people have had
and the situations that they find themselves in. To understand
why someone behaves in a particular way,
you need to examine carefully the situation they are
in and to explore their past experiences in similar
situations, rather than explain differences in behaviour
as resulting from differences in personality. No underlying
personality structure like Jung’s psyche is
thought necessary; rather, individuals have learnt to
behave in certain ways because in the past they have
been rewarded or they have avoided discomfort or
punishment by doing so.
These approaches to understanding the individual
are based on theories of how we learn. The learning
theory explanation of the happy partygoer would suggest
that such individuals have learnt to enjoy parties.
Their first experience of a party as a child was wonderful.
They were given presents, everything went well
and they had a good time. This initial positive party experience
has been followed by others, so that the individual
looks forward to parties as pleasant experiences
because of their learning history of parties. By contrast,
the individual who dreads parties will have had some initial
bad experiences of parties. Perhaps they were made
to share their special toys with other children, or the
other children broke their presents, or they did not get
what they wanted or they were punished for being rude
and so on. Learning theory would suggest that this negative
experience can lead the individual to dread parties,
especially if the negative experience is repeated. These
examples are somewhat simplified, but we hope they
have got the point across. The contention is that your
attitudes to events like socialising are dictated not by
Introduction
Source: Pearson Education Ltd. Jules Selmes
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 71
72 PART 1 PERSONALITY
Introduction to learning theory
We take a slight detour here and look at the history of
learning theory, as many of the crucial concepts that
underpin more current developments emerged early in the
development of psychology as a discipline. You need to
understand these core concepts as they have heavily influenced
many later developments.
Although learning theory developed mainly in the
United States, a major influence was the work of a Russian
physiologist, Ivan Pavlov. At the beginning of the last century,
Pavlov was exploring the digestive system in humans
and other animals. While he was undertaking research
examining the salivary response of dogs, he observed what
appeared to him to be ways in which dogs learnt to
respond to objects and people. When it is given food, a dog
will automatically salivate. This is a naturally occurring
response. In the terminology of learning theory, the food
is called the unconditioned stimulus and the response
of salivating is called the unconditioned response.
Pavlov observed that if a light went on or a bell rang
(unconditioned stimuli) before the dog received food,
after a few trials the dog would salivate when the light
went on or the bell rang. The dog had learnt to associate
what had been a neutral unconditioned stimulus (bell or
light) with the food and salivated at the neutral stimulus
even in the absence of food. This is the basis of what is
known as classical conditioning, and the basic process
is summarised in Figure 4.1. Pavlov (1906, 1927, 1928)
carried out extensive research on the learning associated
with classical conditioning.
Classical conditioning also accounts for some learning
in humans. For example, suppose a parental goal was to
bring up their young child to enjoy books. One scenario for
achieving this, according to classical conditioning, would
be to start with the child being cuddled on a parent’s lap
an experience that makes the child feel good. This is an
unconditioned response (naturally occurring). Reading a
book across a room to a young child will initially be a
neutral stimulus. However, if the parent reads the book to
the child while cuddling the child on their lap, after a few
repeated sessions, being read to will produce pleasant feelings
in the child even when they are not being cuddled by
their parent. In this way, reading books has become a conditioned
stimulus that produces pleasure in the child. Once
reading the book has become a conditioned response,
reading to the child across the room will induce the same
pleasurable response in the child.
Pavlov demonstrated that the conditioned response
could generalise to similar stimuli. In the dog example, it
could be changes in brightness or colour or the light that
would evoke the same response. Similarly with children,
reading while on their parent’s knee may generalise to
being read to across a room and eventually to reading
anywhere, even on their own, and finding it a pleasant
experience. Pavlov showed that there are limits to generalisation.
In the dog example, if the food is delivered to
some sounds but not to others, the dog will learn to discriminate
between the sounds and will only salivate to
the ‘food’ sounds. Finally, Pavlov demonstrated that the
conditioning process could be reversed. If the light is
presented repeatedly with no food following, then the
your personality but by the past experiences of parties
and similar events that you have had. As we mentioned
earlier, these approaches are based on learning theory,
and they vary in terms of how radical they are. We shall
begin by looking briefly at the history of learning theory
and outlining the major concepts that you need to
understand.
Early learning theory developed primarily in the
United States. The roots of the psychoanalytic schools
of personality were firmly based in the European tradition
of psychology, as we have seen, although they
later became established in North America. We shall
examine some of the major approaches to learning
theory in the United States, from the radical approach
of B. F. Skinner to the more moderate views of John
Dollard and Neal Miller. Next, we will examine the work
of Albert Bandura. While still maintaining a learning
theory approach, Bandura introduces cognitive and
emotional variables as factors influencing behaviour.
Next, we will look at two learning theory concepts that
have stimulated enormous amounts of research. These
are Bandura’s concept of self-efficacy and Rotter’s
locus of control. Finally, we will examine the work of
Walter Mischel. Mischel has not developed a fullblown
theory of personality, but his critique of existing
approaches has been enormously influential in personality
psychology, as you will see.
The focus for much of this chapter is the question of
whether you behave as you do because of an inner personality
that drives your behaviour, or whether it is
purely that you have learnt to behave in certain ways in
particular situations.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 72
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 73
dog’s salivary response gets weaker and weaker, till eventually
what is termed extinction is achieved.
At this point, you may wonder what all of this has to do
with personality, but Pavlov went on to show that classical
conditioning could explain many of our emotional reactions.
It could be that I am an anxious person because I
have had experiences where I learnt to be anxious; it is not
simply that I possess a neurotic personality. The crucial difference
is that if you have learnt to be anxious, then you can
unlearn; or, in learning theory terminology, the anxiety
response can be extinguished as it is not a part of your personality.
We will return to this shortly with a detailed
example once we have understood how Pavlov’s work came
to be so influential within psychology.
John B. Watson, an American psychologist, read the
early work of Pavlov and was very impressed by it. He
began to apply some of Pavlov’s observational techniques
in his own research and replicated some of his work in the
United States. As he became established within psychology,
Watson began to call for a change in the direction of
American psychology so that it could become a true science.
He wanted to reject the methods of introspection
and interpretation of patients’ reminiscences that Freud
and the other psychoanalysts had employed. He saw these
methods as unscientific and argued instead for a psychology
that considered only observable aspects of behaviour.
In practice, this means that no assumptions or hypotheses
can be made about what is going on inside someone’s
mind. Stimuli and their effect on behaviour are the subject
matter of the behavioural approach, and rigorous scientific
methods, mainly based in laboratories, are used to
collect data. Watson published his views in 1914 in a book
entitled Behaviour. In 1919, he published Psychology from
the Standpoint of a Behaviourist. This book was influential
in American psychology. It included summaries of
Pavlov’s work on classical conditioning, thereby introducing
Pavlov’s work to a wider audience. Watson is generally
credited as being the founding figure of the School of
Behaviourism, but his career as a psychologist ended with
his withdrawal from academia in 1920 to enter business.
Behaviourism and the popularising of Pavlov’s work had
set the scene for developments in personality theorising
and research. From this perspective variables are manipulated,
ideally in a controlled laboratory setting, and then
the effects of these manipulations on the subject of the
research are carefully observed.
1 Natural response (before conditioning occurs)
Stimulus Response
Food presented to dog
[Food ‫؍‬ Unconditioned stimulus]
Light is switched on
[Light ‫؍‬ Neutral stimulus]
Dog salivates
[Salivation ‫؍‬ Unconditioned response]
No response from the dog
2 Conditioning procedure for several trials
3 After the conditioning trials
g p2 Conditioning procedure for several trials
3 Aft th diti i t i l3 After the conditioning trials
Stimulus Response
Light is switched on and food is presented to dog
Neutral stimulus ؉ Unconditioned stimulus
Dog salivates
Unconditioned response
Stimulus Response
Dog salivates
Conditioned response
Light alone is presented to dog
Conditioned stimulus
Figure 4.1 Summary of classical conditioning.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 73
The clinical perspective within classical
conditioning
If you recall in Chapter 1, a distinction was made between
the clinical and individual differences approaches to the
study of personality. Although the behavioural approach is
a radical departure from the psychoanalytic approaches in
previous chapters, it still maintained a heavy focus on
behaviour change, particularly within a clinical context.
Put simply, if your hypothesis is that behaviour is learnt,
then it is necessary to show that it can be unlearnt. The behavioural
approaches, like the psychoanalytic approaches,
focused on demonstrating that mental health problems
(psychopathology) could be cured using behavioural interventions.
A crucial difference between the psychoanalysts
and the early learning theorists concerns how psychopathology
arises. For the psychoanalysts, as we have
seen, psychopathology arises because of inner causes such
as unresolved developmental crises; or conflict between the
structures within the personality such as the id, ego and
superego; or problems in personality development of some
other kind. The learning theorists rejected this as an explanation
for the cause of mental problems, seeing it as unscientific
to refer to what they saw as unobservable inner
mental processes and/or structures to explain observed differences
in behaviour. For these learning theorists, psychopathology
was a learnt maladaptive response to a
situation that may have generalised to other situations or
similar stimuli, and as such, it could be unlearnt. Normal
development was about learning adaptive responses in a
variety of situations, while abnormal development resulted
from acquiring maladaptive responses.
Pavlov (1927) began this line of research by inducing
what he called experimental neuroses in one of his laboratory
dogs. The dog was conditioned so that he would
salivate to the shape of a circle. He then learnt to distinguish
between circles and ellipses, only salivating to circles.
However, when the distinctions between circles and ellipses
became harder to distinguish, the dog became very distressed;
his behaviour was disorganised, with a preponderance
of neurotic symptoms. The dog barked when taken
into the laboratory, shivered in his harness and tried to bite
the restraining straps. Pavlov interpreted this as demonstrating
that when the dog could no longer cope with what
was being asked of him, he developed neurotic symptoms.
Watson and his colleague Rayner (1920) went on to
demonstrate that human emotional responses could also
be manipulated using classical conditioning. This is the famous
classical conditioning experiment carried out on an
11-month-old infant called Albert. This has come to be
known as the ‘Little Albert’ study and is still regarded as a
classic in psychology. Albert initially did not display any
fear of laboratory rats, but he did produce a startle and fear
response to a loud noise made by banging a hammer on a
metal bar. As Albert began to reach for a rat, the noise was
made behind his head. After a few repetitions, he had been
conditioned to fear the rat in the absence of any noise. This
demonstrates how a child can learn an emotional response.
This condtioned fear of white rats then generalised to other
white, furry objects like a mask of Father Christmas and
even Watson’s own white hair. This work led to other psychologists
exploring ways in which negative emotional reactions
could be unlearned, and a great deal of work was
carried out in this area from the 1920s until the 1980s. A
74 PART 1 PERSONALITY
Stop and think
Treating classically conditioned emotional responses
Systematic desensitisation
This can be used to treat phobias, for example, someone
with a phobia of birds. The aim is to replace the
old association between the feared stimulus (bird) and
the feared response (panic symptoms) with a new
association of relaxation. The client and therapist begin
by ranking bird-related fears from most to the least
feared. Holding a bird might be most feared; a picture
of a bird might be least feared. Next, the client is
taught how to relax. The response of relaxation is
incompatible with the feared response. The client and
therapist move through the list of fears, ensuring that
at each level the feared response becomes conditioned
to the relaxation response, till the client can
comfortably face their worst fear of birds. Many phobias
and other anxieties have been successfully
treated with systematic desensitisation.
Alcoholism
Aversion therapy has been used to treat people with alcohol
addictions. Here the image of a drink could be
paired with images of being sick or other negative
images. This therapy has also been used to assist individuals
stop smoking.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 74
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 75
summary of some of this work on classical conditioning is
given above in Stop and think: Treating classically conditioned
emotional responses, for those of you with a clinical
bent. The principles are still applied in some contexts; but
our next theorist, Skinner, developed this work further.
The radical behaviourism of
B. F. Skinner
Skinner had been influenced by the research of Pavlov and
Watson, among others, and developed it further. (See the
Profiles box on page 77.) He did not claim that unconscious
processes or inner states did not exist, but he
strongly felt that it was unscientific and unnecessary to rely
on these unobservable processes to explain behaviour. He
did not deny that we had ideas and thoughts, but he
strongly believed that these inner thoughts did not cause
our behaviour. Suppose you do not turn up to do a seminar
presentation; you may say that you were so anxious at the
thought of doing it that you could not make yourself
attend. For you, the explanation is that your anxiety prevented
you from attending. You are claiming anxiety as the
inner cause of your not attending behaviour. You may even
go as far as to claim a neurotic personality. Skinner would
not agree with this interpretation. This inner state of anxiety
is not the cause of your non-attendance. He would
argue that you experienced certain aversive behaviours
when preparing to attend; you may have felt nauseous, had
palpitations, sweated and so on, perhaps at the sight of
your presentation or while packing your bag. This resulted
in you altering your preparatory behaviour. The change in
your behaviour and the change in your feelings have the
same cause. Saying that you are an anxious person does not
explain the cause of the anxiety. For Skinner the cause of
your anxiety was located somewhere in your developmental
learning history where you have learned maladaptive
responses. Skinner felt that much of the time, we do not
know the real causes of our behaviour, in terms of what
stimuli in the environment trigger specific behaviour; and
he rejected completely the notion of behaviour being motivated
by inner states. So if you say you feel happy, something
in the environment has triggered a response that you
have previously learnt to label as happiness; it is not some
internally generated feeling for Skinner, but is stimulated
by something in the environment.
Skinner (1948) did not accept the concept of personality,
seeing it as unnecessary and unscientific to postulate
unobservable, inner psychological, personality-generating
structures. He accepted that our genetic inheritance would
have some influence on how we interacted with the environment,
but he played it down, claiming instead that the
situational determinants were crucial in explaining the
cause of behaviour. He made reference to Charles Darwin’s
principles of natural selection, suggesting that over many
generations human beings have evolved particular characteristics
to meet the demands of their particular environment,
and he believed this had led to some genetically
based individual differences. Perhaps being agile had a survival
value for a particular group of people; then these individuals
would have opportunities to express their agility
in their environment, and these responses would be reinforced.
The more agile you were, the greater the reinforcement
and so on. This then would explain observed
individual differences in behaviour. Heredity would only
impose limits on behaviour. For Skinner, it is not the kind
of person you are, but the learning history you have had
and the current demands of your environment that dictate
how you behave.
Skinner accepted the principles of classical conditioning
but felt that it applied to a limited range of learning
situations. He argued that what happened after particular
ways of behaving was a crucially important aspect of
learning that applied to most situations where we learn.
He suggested that the classical conditioning paradigm,
consisting of a stimulus followed by a response, is too
Stop and think
Ethical reflection on the ‘Little Albert’ study
Do you consider it ethical to carry out experiments in
which you make a young child fearful and upset to the
point where the child cries? For most of us, this research
is unethical, as modern codes of ethical conduct
would make clear. The aim is to cause distress in a very
young child. Ethical issues are complex, however, and
you may want to reflect on the following:
Watson carried out much of his conditioning research
on his own children. Does this make any
difference?
Was his research worthwhile? Have we gained useful
knowledge from his work?
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 75
simplistic for most learning situations. He demonstrated
that what happens after the response – the consequences of
the response – is what is crucial, as it affects the probability
of the response being repeated. If you are praised for
your seminar presentation, then you are more likely to
volunteer to do a seminar presentation in future; if you are
heavily criticised, then you are more likely to want to avoid
future presentations.
Skinner refers to this learning process as operant
conditioning. If the consequence of a piece of behaviour is
to encourage the repetition of that behaviour, this is
termed positive reinforcement. Consequences that discourage
repetition of the behaviour are termed negative
reinforcement. Although Skinner’s primary interest was
in human behaviour, most of his research was on animals
in laboratory situations in the now famous Skinner box.
This is illustrated in the photo.
There were slightly different versions of the box for
different animals; but essentially, there is a lever of some
sort that the animal in the course of exploring the box
will press at some point. When this happens, the animal is
rewarded with food. There is an electronic device
attached to the lever to record the animal’s rate of pressing.
What Skinner demonstrated was that after the bar
pressing had resulted in the animal’s being reinforced
with food, the rate at which the animal pressed the bar increased.
The animal did not have to be reinforced every
time for learning to occur, and Skinner studied the effects
of different schedules of reinforcement. Much of the detail
of this work is not particularly relevant in the context
of personality theory, and we will cover only the relevant
concepts.
Skinner demonstrated that random or partial reinforcement
schedules produce behaviours that are very
resistant to change, as an example will show. In one family,
the teenage son was told that his weekend curfew was
11 p.m. However, every Friday night, Tim (the teenage son)
would plead with his mother to be allowed to stay out later,
and an argument would often ensue. The mother could not
understand why Tim always had to argue and could not
just accept that 11 p.m. was the curfew. She said he was
stubborn and argumentative like his dad. In other words, it
was down to his personality. When asked if she ever did
allow Tim to stay out later than 11 p.m. on a Friday, she
said that sometimes he just wore her down; or if she was in
a good mood, she sometimes let him have another hour. In
Skinnerian terms, Tim was on a random/partial reinforcement
schedule. The rule was that his curfew was 11 p.m.
However, Tim had learnt that it was always worth challenging
this as sometimes his mother gave in and he was
rewarded with a later curfew. So, for Skinner it was unnecessary
and unscientific to refer to internal personality attributes
to explain this behaviour, as learning theory
provided an adequate explanation based on observable
events. We are sure that if you reflect on some of the conflicts
that you have experienced over family rules when you
were growing up, you will find that operant conditioning
provides a good explanation.
Another relevant Skinnerian concept is shaping. Skinner
observed that when pigeons first entered a Skinner
box, it might take them some time before they found the
lever and pressed it. To speed up the process, he would
deliver a food reward when they were facing in the
direction of the lever, another reward when they came
close to the lever and so on until the pigeon had actually
achieved the desired response of pressing the lever. Shaping
is applied to many aspects of behaviour where individuals
are initially rewarded for behaviour that approximates
the desired goal, and once that behaviour is established
they are rewarded only for behaviour that comes closer to
the goal and so on. Many of the current television programmes
that help parents develop parenting skills areas
are based on principles of operant conditioning where desired
behaviours are gradually shaped. The children have a
76 PART 1 PERSONALITY
Skinner box.
Source: Nina Leen/Time Life Pictures/Getty Images
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 76
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 77
Profiles
Major figures in learning theory Ivan Petrovich Pavlov, John Broadus Watson
and Burrhus Frederic Skinner
Ivan Petrovich Pavlov
Pavlov was born in Ryazan, a small village in central
Russia, in 1849. He was educated at a church school
followed by a seminary and seemed destined to enter
the priesthood. However, in 1870 he changed direction
and studied chemistry and physiology followed by
medicine at St Petersburg, becoming a skilful surgeon.
After working for two years in Germany, he returned to
St Petersburg and was made professor of physiology in
1890 at the Imperial Medical Academy. In 1904, he was
awarded the Nobel Prize for his work on digestion.
Pavlov was an independent, outspoken man; yet
despite this, he managed to survive the Russian revolution
and was allowed to continue his research although
never becoming communist and openly criticising
aspects of the regime. In 1922, at a time of famine in the
Soviet Union, he asked Stalin for permission to take his
laboratory overseas. This was denied as Stalin felt that
the Soviet Union needed scientists like Pavlov. However
Stalin did allowed Pavlov to visit America, first in 1923
and then in 1929. Although Pavlov was a physiologist,
his research on learning and the methods associated
with it have had, and continue to have, a major influence
on the development of psychology.
John Broadus Watson
Watson was born in 1878, the first son of a poor family
in Greenville, South Carolina. His father was a womaniser
and abandoned his family when Watson was
13 years old. Watson found this difficult and rebelled
against his mother and school. With the support of one
of his teachers, he returned to study and eventually
studied for a doctorate in psychology at the University
of Chicago. In 1902, in the last year of his doctoral studies,
he suffered an emotional breakdown. In his autobiography
(Watson, 1936) he discusses how after his
breakdown, he could accept the validity of much of
Freud’s work. This seems at odds with the individual
who, as we have seen, founded the school of psychology
known as behaviourism. In 1913 Watson lectured
and published the seminal paper on behaviourism,
Psychology as the Behaviourists View It. In 1915 while
professor of psychology at Johns Hopkins University, he
became president of the American Psychological Association
and seemed set for a career as an eminent psychologist.
However, in his private life, like his father, he
had a great number of affairs with women and in 1920,
he was forced to resign from Johns Hopkins University
over a sexual scandal involving his research assistant.
His academic career was over, although he continued to
publish for a few years. He went into the advertising
business and became a successful businessman. However,
his relationships with his family were poor; after his
retirement from business in 1945 and the death of his
second wife, he lived as a recluse on a farm in
Connecticut until his death in 1958.
Burrhus Frederic Skinner
Skinner was born in Susquehanna, a small town in
Pennsylvania, in 1904. He had a middle-class upbringing
in a warm, supportive family. His initial interest was
in literature, and he wanted to become a writer. While
working in a bookstore to support himself, he read
books by Pavlov and Watson. Wanting to know more
about psychology, at age 24 he enrolled at Harvard for a
research degree. This was supposedly jointly supervised
by the physiology and psychology departments,
but in reality, Skinner was allowed a great deal of freedom
to develop his own research and experiment with
equipment, developing the Skinner box illustrated on
page 76. In 1936 he married and left Harvard for a lecturing
post in Minnesota. During the Second World War,
he was funded by the America government to carry out
a project to train pigeons to guide bombs. The pigeons
would keep pecking at a target that kept the missile on
course. A parallel secret project was on the development
of radar, and when that was successful, Skinner’s
research was discontinued. However, he had discovered
that pigeons learnt more quickly than rats, and
from this point onwards he used only pigeons in his research.
In 1945, Skinner became professor of psychology
at the University of Indiana; the following year, the
Society for the Experimental Analysis of Behaviour was
set up. This development reflected the growing influence
of behaviourism in the United States. In 1948,
Skinner was given a chair at Harvard. In the same year,
he wrote his only novel, Walden Two. This describes a
community governed by the principles of learning theory.
It describes a utopian society, which provided a
wealth of experience for individuals to fulfil their potential.
Although the book was fictional, a group of young
people set up a community based on the book in Virginia
(Kinkade, 1973). Skinner continued to work until
his death from leukaemia in 1990. He focused on developing
effective ways of teaching and learning, being an
early proponent of programmed learning. In later life he
became interested in philosophical issues, but he continued
to be upset by the misrepresentation of his work
by sections of psychology. However, the huge number
of publications related to his work testify to his influence.
Indeed the Journal of Experimental Analysis of Behaviour,
set up in 1958, is still dedicated to research in
the Skinnerian tradition.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 77
star chart. They are rewarded for ‘good’ behaviour with a
star, and earning stars ‘buys’ treats. Gradually, as the initial
good behaviour becomes established, the parents up the
ante for the child to earn stars. Skinner’s contention is that
eventually the children’s good behaviour will become selfreinforcing
as their relationships will be better, and this is
rewarding in itself.
As we have seen, one of the big questions for personality
theory is the nature of human motivation. For Skinner
(1971, 1972, 1976) the issue was straightforward. He
believed that human beings aim to produce pleasant events
and to avoid painful events, if possible. All our emotional
states can be understood by analysing the behavioural
events in the environment that preceded them. He does
accept that some behaviour is private, but he refuses to
accept that internal private behaviour causes our emotions.
You don’t get anxious because you have an anxious personality,
but because something in your environment stimulates
the anxious behaviour.
Skinner devoted a lot of his writing to examining
Freudian concepts and dismissed most of them as unscientific,
constructs for which there was no observable evidence.
He agreed with Freud that the early experiences
of the child had long-lasting effects, which could even
continue into adulthood. However, he contended that it
was the early conditioning experiences of the child that
shaped their later behaviour, not the influence of inner
conflicts between hypothesised personality structures.
For Skinner, demanding individuals are not governed by
their id impulses, as Freud would claim; rather, they
have in the past been rewarded for displaying demanding
behaviour by having their demands met and have
therefore learnt to behave in a demanding way. Skinner
(1953) agreed that personality trait names do convey
useful information describing the individual, like how
friendly or enthusiastic they are; but they do not explain,
in any empirical way, how they came to be friendly or
enthusiastic. For him the friendly person has been reinforced
more for being friendly than has the unfriendly
person and so on.
Skinner also denies that human beings are purposeful.
He claims that what we label ‘intentions’ are really
responses to internal stimuli. For example, when you say
that you want to go for a picnic in the park, for Skinner,
you are not setting some mental future goal; rather, you are
responding to some observations – perhaps internal and
external – that in the past were associated with you having
a picnic. It could be that the sun is shining; you observe
that you have nothing else to do, you catch sight of a thermos
flask in your kitchen, or you drive past a park and perhaps
a previous memory of a picnic is triggered. For
Skinner these or variations of them are the stimuli that you
are responding to when you make a statement of intent to
have a picnic.
Attempts to apply learning
theory approaches to personality
All learning theorists are not as radical as Skinner is, and
the theorists that we will explore now all made serious
attempts to apply concepts derived from learning theory to
personality. John Dollard and Neal Miller, two of the earliest
of these theorists, both worked at Yale University. They
are somewhat unique in that their aim was to try to integrate
learning theory principles with Freud’s psychoanalytic
approach. Both Dollard and Miller had trained as
Freudian analysts, Dollard at the Berlin Institute and Miller
at the Vienna Institute. By background, Dollard was a
social scientist, teaching anthropology, sociology and psychology
and only specialising in psychology later in his
career. Miller had trained as an academic psychologist
before his analytic training. Both men were impressed with
the work of the learning theorists while also influenced by
Freud’s theory. They sought to develop a synthesis of the
two concepts to create a theory of personality.
Dollard and Miller collaborated on animal laboratory
studies, mainly using rats, sharing Skinner’s view that animal
learning could be generalised to humans. However,
unlike Skinner, they allowed for inner causation in behaviour.
They believed that due to the higher mental processes
of humans, our behaviour does not consist merely of
responding to stimuli in our environment; instead, we can
also respond to inner stimuli, and thoughts can be reinforcing
for us. This is the first attempt to allow cognitive
processing within a primarily learning theory model. The
principles of learning demonstrated in lower species in
the laboratory still applied to human learning, but because
of their superior mental processes, humans were capable of
more complexity. Thoughts and memories could cue
behaviour within their model. This also allowed humans to
plan ahead and anticipate events. There was even a role for
the unconscious.
Dollard and Miller acknowledged the importance of
unconscious processes in human behaviour, but their definition
of the unconscious is different from Freud’s; he saw
the unconscious as comprising the sex and death instincts,
which were inherited from birth. Dollard and Miller suggest
that we are unaware of some processes because we acquired
our drives and the cues before we learnt to talk and
consequently they are not labelled. Examples might be
some of our secondary drives for social contact, love and so
on that we learnt as infants from our initial contacts with
our parents. We have learnt to associate a particular smell,
perhaps with the good feeling of being fed, but are unconscious
of it. In future when we are exposed to the smell, it
will affect our behaviour at an unconscious level. Other
cues may be unconscious, as they are not labelled in our society.
For example, in Japanese society to lose face (to be
78 PART 1 PERSONALITY
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 78
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 79
embarrassed or humiliated, especially publicly) is an
important concept, and there is a richer vocabulary in
Japanese for labelling the experience than is the case
in English, where the concept is not so important. Whether
labels are readily available also affects how we perceive
cues. The well-known example always cited here is that of
the Inuit people (people who live on the arctic coasts, including
Siberia, Alaska and Greenland) and their wealth of
labels for different types of snow. Consequently, they make
discriminations between types of snow that English speakers
would find difficult or even impossible to do. This then
accounts for material being in the unconscious because it is
unlabelled.
Cues may also be unconscious because they have been
repressed. Dollard and Miller suggest that the defence
mechanism of repression is a learned response, like the rest
of our behaviour. When we discussed repression previously,
as a Freudian defence mechanism, we saw that it involves
suppressing inconvenient or disagreeable feelings or
thoughts. If we cannot remember something, it cannot
upset us. For Dollard and Miller, repression is about a failure
to label the upsetting thoughts or memories so that
they are not easy to recall and making a decision not to
think about it. When you recall unpleasant events, this reinforces
the negative experience you originally had. Repression
avoids this, and not labelling the feelings makes it
harder for them to be recalled to conscious thought.
Dollard and Miller accepted the importance of the effects
of unconscious motivation and Freudian defence mechanism
on behaviour, but they expressed defence mechanisms
in learning theory terminology. The interested
reader can find a very readable account in their book,
Personality and Psychotherapy, published in 1950.
The stimulus–response model of
personality of Dollard and Miller
As expected in a stimulus–response (S–R) theory, the emphasis
was on how behaviour is learnt. From Hull, another
early American learning theory researcher, Dollard and
Miller borrowed the term habit to label the association between
stimulus and response. Within their model, personality
is composed largely of learned habits, and they go on
to explain how these habits are acquired and maintained.
They agreed with Freud that the infant is born with
some innate drives, which they termed primary drives,
but disagreed with Freud about the nature of these drives.
These innate primitive drives are physiological drives associated
with ensuring survival for the individual. They include
hunger, thirst, the need for sleep and the avoidance of
pain. Reduction of these drives provides the most powerful
reinforcement for the individual. Dollard and Miller
(1950) claim that this reinforcement occurs automatically
and unconsciously, and to be maximally effective, it should
immediately follow the response. Like many other personality
theorists with a clinical background, Dollard and
Miller focus mainly on psychopathology in the development
of their theory and then extrapolate from this to normal
development. For example, if an infant is left to
become extremely hungry (primary drive), then it cries
very loudly for attention (response). If the mother then
feeds the infant, what the infant is said to have learnt is that
making a fuss is rewarded. Such a child might then go on to
make an excessive fuss every time they have a drive that requires
satisfaction. In this terminology, making a fuss has
become a habit. The baby whose primary drive of hunger
was quickly met would not have this habit of overreacting
and would develop normal levels of response, in this case
distress. In most Western societies, primary drives are
rarely directly observed, apart from in infant feeding, as societies
have developed means of reducing them before they
become pressing. The process for doing this involves the
acquisition of what Dollard and Miller termed secondary
drives. These secondary drives are learned mainly to help
us cope with our primary drives. An example would be of
setting regular mealtimes so that you are motivated to eat
at particular times before the primary drive of hunger becomes
overwhelming and therefore distressing. Associated
with these primary and secondary drives are different types
of reinforcement. For the innate primary drives, primary
reinforcers are food, water, sleep and so on. Secondary
drives similarly have secondary reinforcers. These secondary
reinforcers are items or events that were originally
neutral but have acquired a value as a reinforcer through
being associated with primary drive reduction. A mother
smiling at her child is a secondary reinforcer as it is associated
with physical well-being. Money is also a secondary
reinforcer as it is associated with being able to buy food,
provide shelter and so on.
Dollard and Miller describe the learning of habits as
being composed of four constituent parts: the initial drive,
the cue to act, the response and reinforcement of the response.
As discussed earlier, the drive stimulates the person
to act. It does not guide them how to act but simply lets the
person know that they want something. A drive might be
hunger. Cues provide guidance about how to act or respond
in S–R terminology. You notice a billboard advertising
a new Chinese takeaway. This might be the cue for you
to respond, by taking a detour past the takeaway, to get
something to eat. If you then pick up a delicious meal that
you enjoy hugely, you will no longer be hungry; your drive
will have been satisfied. In this situation, the Chinese meal
constitutes reinforcement. Reinforcement refers to the
effect that a response has. As the meal was good, it reinforced
your action of going to get it; and next time you are
in a similar position, hungry when walking home, you may
be tempted to repeat the experience. In S–R terms, a habit
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 79
has been formed. If, on the other hand, the meal was disgusting
and the portions were tiny, the experience of visiting
the takeaway would not have been reinforcing and you
are unlikely to visit it again. In S–R terms, if the response
does not satisfy the drive, it will undergo extinction. It does
not mean that you will never again visit the takeaway, but
you are less likely to do so. Remember that habits can be
both positive and negative. They are simply associations
between stimuli and responses.
Dollard and Miller (1941, 1950) were particularly interested
in what happened when we became frustrated in our
attempts to satisfy our drives. They described four types of
conflict situations that we could face. The conflict is caused
by our tendencies to wish to obtain (termed ‘approach’)
certain goals or objects. They developed a simple diagrammatic
system to illustrate these conflicts, as they felt that
this helped them understand exactly what was going on in
any situation (see Figure 4.2).
Approach–approach conflict – This describes the
situation where there are two equally desirable goals,
but they are incompatible. This could be when you are
asked to choose between two equally desirable objects to
have as a gift. You really want both but can only have
one. Both goals are positive but incompatible.
Avoidance–avoidance conflict – This is the situation
where you are faced with what you perceive as two
equally undesirable alternatives. You have a spare hour,
and your partner asks you to go jogging, which you
hate; or you could offer to do the ironing as an excuse
not to go jogging, but you equally hate ironing. Here for
you both goals are undesirable and incompatible in
terms of not having a wish or time to do both.
80 PART 1 PERSONALITY
Approach—approachh h
؉ ؉
Avoidance—avoidanceoiddance—avoiidance
؊؊
Approach—avoidance
؊؊
؉
Double approach—avoidance
؊
؉ ؉
؊
؉
؊
؉
؊
Figure 4.2 The Dollard and Miller system for
analysing conflicts.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 80
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 81
Approach–avoidance conflict – Here there is one
goal, but while an element of it is attractive, an aspect of
it is equally unattractive. For example, you are offered a
place in what seems an ideal house; however, one of
your housemates would be someone you really do not
like.
Double approach–avoidance conflict – Here there
may be multiple goals, some desirable and some undesirable.
This is more like most situations, where there
are a variety of factors, positives and negatives to take
into consideration before being able to make a decision.
Although we have used human examples to illustrate
the analysis of conflict situations, Dollard and Miller used
laboratory animals rather than humans to demonstrate
that this system was accurate at predicting behaviour.
For Dollard and Miller, therefore, behaviour is motivated
by the need to reduce our primary or secondary
drives, and we learn new behaviours in the process. It is a
deterministic account of human development. It is more
complex than the early learning theory models as it allows
for the inner influence of human cognitive processes, and
in this, it is a forerunner of the cognitive models of personality
that we will examine in the next chapter.
One aspect of Dollard and Miller’s model that is subject
to criticism from behaviourists as not being radical enough
is their approach to the treatment of mental disorder. For
them, as with the other learning theorists, psychopathology
consists of learned, unproductive, unhelpful habits or
responses. In their integrative approach, they suggested
that some of these habits might be unconscious because of
the reasons we have discussed earlier and that this factor
added to the complexity. The aim of treatment is to remove
these ineffective habits and replace them with new, more
effective habits. Unlike the earlier behaviourists, Dollard
and Miller did not adopt a purely deconditioning approach
to treatment; they maintain significant elements of psychotherapy,
the ‘talking cure’, in their approach. There are
two phases to their treatment: first is a talking phase,
where the problem habits are identified, explored and
labelled. In the second phase, the patient is encouraged to
learn more adaptive habits and apply them in their life.
They call this phase the performance phase. They
departed from their Freudian psychoanalytic training in
not attending to the problems that patients had experienced
in the past. They felt that past emotional issues do
not have to be relived in therapy for them to be resolved.
The past is only helpful sometimes in helping patients
understand the source of their problems. Their focus was
on current problems in living and future strategies. This
predates the current treatment practice in cognitive therapy
that we will examine in the next chapter.
One other significant contribution made by Dollard
and Miller (1941) was to recognise and outline the process
of observational learning. They demonstrated how performance
on a novel task can be improved by seeing someone
else perform the task. This increases the speed of the
learning process. They suggested that observational learning
is important in development as children learn from
observing adults and other children in situations that are
novel to them. They stressed that observational learning
could explain how both adaptive and maladaptive habits
are learned.
We will now examine the contributions of theorists
Albert Bandura and Michael Rotter, who have further
developed this concept of social and observational learning
in ways that can be usefully applied within personality
theorising.
Albert Bandura and social
learning theory
One of the major questions in personality theorising is
whether inner or outer forces control our behaviour. As we
have seen, the psychoanalysts would have us believe that
inner forces determine who we are and how we behave. The
learning theorists we have examined so far conceptualise
human beings as being at the mercy of outer forces. The
environment determines your opportunities for learning
new behaviour, the interests you are likely to develop and
your history of learning. Dollard and Miller allowed for
some inner influence from the higher cognitive functioning
possessed by humans but said that principles of reinforcement
external to the individual are thought to mainly control
human behaviour. Bandura challenges this view, as we
shall see. (For background on Bandura, see the Profile box.)
Bandura’s work is grounded in the learning theory tradition,
but his focus is on human problems in living. He
moved from animal studies to focusing on purely human
behaviour, although he kept the methodology of undertaking
mainly laboratory-based research. His laboratory techniques
are much more sophisticated, emphasising
observation in situations designed to simulate real-life
experiences. He is interested in developing theory and
applying it to behavioural problems in order to facilitate
positive change in individuals and groups.
The model of the individual in his approach is of an
active player responding both to inner stimuli and the
external environment and moving back and forward in a
dynamic system. Individuals are seen to be influential in
determining their own motivation, development and
behaviour. Bandura (1978, 1989) uses the term reciprocal
determinism to label the processes that drive behaviour.
He sees an individual as being influenced by personal factors,
behaviour and environmental factors. All three factors
interact with one another to influence how individuals
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 81
behave. The direction of these interactions is displayed in
Figure 4.3.
Personal factors include the individual’s cognitions,
emotions and biological variables that contribute to their
inner state. This is a major break from the traditional
learning theory approaches we have examined previously.
These personal factors can impact on both an individual’s
behaviour and on their environment. If you truly think you
will fail at a task in a specific setting, Bandura (1995) has
shown that this greatly increases the likelihood of failure as
you approach the task differently. Your cognitions are
affecting your behaviour. If you do not like opera, then you
are unlikely to choose to go to an opera. Here your cognitions
are influencing the environments you experience.
Similarly, if you hate smoking, you may avoid bars that you
know will be smoky. Here environmental factors and personal
cognitions are impacting on your behaviour.
Bandura also suggests that behavioural factors can
affect the individual’s cognitions, feelings and emotions.
Supposing you go skiing for the first time with some
friends, and you prove to be good at it. You get your balance
quickly, and the instructor is complimentary. You
now have the cognition that ‘This is something I can do’.
Your feelings may also change from apprehension about
whether you could do it, to feeling positive about skiing.
The converse might also be true if your initial experience of
skiing was awful. Your behaviour with regard to the skiing
experience has influenced your cognitions and your feelings,
and both are likely to influence whether you choose to
ski in future. If you take up skiing, you may even become
an expert and your body will develop neurological networks
reflecting your expertise. In this way, Bandura
(2001) demonstrated how our behaviour affects our cognitions,
feelings, emotions, and even our neurobiology in
some instances.
With regard to the environment, we have seen from the
earlier examples how environmental factors like polluted
environments may affect our behavioural choices.
The hole in the ozone layer is another good example that
has resulted in us having to take more care to avoid burning
when it is sunny. We may have to spend time and
money buying sunscreens, or we may avoid sunbathing
(behavioural factors). In addition, we may also worry
about sunburn and plan ways to avoid it (personal factors).
We hope that these examples give you an idea of Bandura’s
model of reciprocal causation in action. You surely can
think of other examples. Doing this makes you aware of the
complexity of learned behaviour that, unlike the earlier
models, Bandura’s model can handle.
Unlike Skinner, Bandura (1995, 1998) believes that individuals
do possess free will and are not at the mercy of their
drives and reinforcement schedules in their learning environment.
For Bandura, our cognitive processes allow us
some control in selecting the situations we operate in and
82 PART 1 PERSONALITY
P
P ‫؍‬ Person factors
B ‫؍‬ Behavioural factors
E ‫؍‬ Environmental factors
E
P P f t
B
Profile
Albert Bandura
Albert Bandura was born in 1925 in Mundare, a small
Canadian town in the province of Alberta. He was an
only son with five older sisters. His parents were farmers
of Polish descent. Bandura went to a small school with
a shortage of teachers. To combat this, the pupils
formed groups to educate themselves in subjects where
no teacher was available. This led to his interest in selfmotivation
and group effects in the process of learning
and motivation. As a young man, he worked with a gang
of labourers repairing the Alaskan highway. He met a
diverse range of individuals, many escaping from the
law and others on the fringes of society, and this is said
to have sparked his interest in psychology and how it
could address the real problems of living (Stokes, 1986).
Bandura studied psychology at the University of British
Colombia and the University of Iowa. Here he was
introduced to researchers working on learning theory.
He was appointed to a teaching post at Stanford University
in 1954 and is still there. Bandura is one of the
most distinguished living American psychologists. He
received the American Psychological Association’s
award for Distinguished Scientific Contributions in 1980.
Figure 4.3 The interacting factors in reciprocal
determinism.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 82
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 83
in creating or transforming situations. We may have to
work, but we can choose what we do in our leisure; or we
may work towards changing a work situation to make it
more amenable to us. We can start businesses or interest
groups or throw parties to provide the experiences and environments
that we need. Bandura has labelled this
personal agency: the belief that you can change things to
make them better for yourself or others. Bandura (1999)
has extended the concept to include proxy agency, where
the individual enlists other people to help change some of
the factors impacting on their life. They may ask a family
member to look after their child so that they can get a job,
or so that they can change their life in some other way.
Bandura points out that there can be a downside to proxy
agency in that people may in the process surrender their
power to the other, who may not have their best interests at
heart, and/or they may become subservient and give up
control of their lives. He prefers the idea of collective
agency, which is where a group of individuals come together
believing that they can make a difference to their
own and/or others’ life circumstances. An example might
be the recent emergence of farmers’ markets. Here groups
of like-minded people who want to be able to continue
earning a living from farming by getting a fair price for
their produce and who share a commitment to fresh local
produce have joined together to sell their produce directly
to the public for a fair price. There are many other examples
of collective agency in community and national
groups and charities.
Learning within Bandura’s model
Within Bandura’s model, personality development is about
how we learn to become the person we are, and this then
explains why we behave as we do. Bandura (1977) suggests
that for learning to be effective, individuals have to be
aware of the consequences of their behaviour. He demonstrated
that people think about the consequences of their
behaviour in learning situations. We think beyond the
immediate situation and anticipate possible outcomes with
an eye to the future. Being aware of the consequences of
our behaviour also allows us forethought, in that we can
anticipate what possible outcomes may follow our behaviour,
and this knowledge can affect how we choose to
behave and what we learn in the situation. Bandura (1995,
1999) sees awareness of the consequences of our behaviour
and foresight as being human attributes as we possess language
and symbolic thought which make it easier to record
the consequences of our actions.
One of the most well-known aspects of Bandura’s work
is observational learning. He points out that more of our
learning occurs by watching and following what other
people do and imitating their behaviour than occurs by
classical or operant conditioning. What happens is that an
individual watches someone perform a novel behaviour,
and when they are required to perform the same behaviour,
they copy what they have previously seen. This is
termed modelling by Bandura.
Bandura and Walters (1963) undertook a series of famous
studies with a doll (Bobo); you may have covered
these studies in your social psychology courses. To summarise,
nursery school children were divided into two
groups. One group, the experimental group, watched an
adult playing aggressively with a plastic doll called ‘Bobo’.
The adult hit and kicked the doll, shouting things like
‘Throw him in the air’. The second group of children, the
control group, did not see the aggressive play. Later, both
groups of children were allowed to play with the doll.
Children in the experimental group displayed twice as
much aggression towards the doll as did those in the
control group.
From variations of this study, Bandura (1977) concluded
that three factors are important in modelling.
Firstly, the characteristics of the model influence how
likely we are to imitate them. The more similar to ourselves
the model is, the more likely we are to imitate them.
Models undertaking simple behaviour are more likely to be
copied than they are if the behaviour is complex. The type
of behaviour being modelled also has an effect, with hostile
and aggressive behaviour more likely to be modelled. Secondly,
the attributes of the observer exert an influence.
Less-confident individuals and those with low self-esteem
or those who feel incompetent in the situation are more
likely to imitate the model. Individuals with a learning history
of being rewarded for conforming behaviour or who
are highly dependent also imitate models more. Finally,
Bandura showed that the consequences of imitating a
behaviour are the most influential factor. If individuals
believe that imitating a behaviour will bring positive
results, then they are more likely to do so.
While we have talked about modelling using the term
‘imitation’, Bandura is insistent that modelling involves
more than passive imitation. It is an active process of learning
through observation, where the observer makes judgements
and constructs symbolic representations of the
behaviours observed. These symbolic representations may
be verbal descriptions or visual images, and they are used
to guide the individuals’ future behaviour in similar situations.
Bandura has studied the factors that may influence
these processes in great detail; interested readers can refer
to the further reading provided at the end of the chapter.
Here we will restrict ourselves to a more aerial view of his
theory as it relates to aspects of personality and its development.
It is sufficient for us to be aware that modelling is not
a passive process of observation, but an active process
where the observer reviews what they have learnt and
makes judgements about it and may decide to keep or
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:02 PM Page 83
discard parts of the behaviour. A distinction is also made
between what we have learned (knowledge acquired) and
what we can do (performance). Performance is seen to
involve trial and error as we gradually shape our behaviour
into the desired format. We also may acquire knowledge
that we do not use, like learning about ways to murder
someone from watching television; fortunately, few of us
ever put this knowledge to use.
While reinforcement is crucial for learning in classical
and operant conditioning, Bandura demonstrates that it is
not always necessary in observational learning. You notice
vivid billboard adverts, or loud noises, because they command
your attention. You may not think about the information
at the time; but when faced with an array of soap
powder at the supermarket, you recognise the one from the
billboard. Bandura also demonstrates that we can and frequently
do reinforce our own behaviour. This selfreinforcement
is where we evaluate our own behaviour;
we may stop doing something we are getting no pleasure
from, or that we judge as harming us in some way, while
continuing to do things that bring positive reinforcement.
The other crucial element for learning Bandura identifies is
an incentive so that we are motivated to learn. Here
forethought plays an important part, as well as the more
traditional cues for learning. Forethought can allow us to
anticipate reinforcements and thus motivate our behaviour.
He suggests that motivation is crucial with observational
learning, as it requires practice for the skills to be
perfected. Motivation and reinforcement are much more
dynamic complex processes in this model.
Bandura (2002) is keen to encourage the application of
his social learning theory to address global problems such as
the AIDS pandemic, population growth and gender inequalities.
He sees social modelling and observational learning
as being core components of behavioural change. The
modelling principles in his famous Bobo study were incorporated
into serial dramas and soap operas by a well-known
writer (Sabido, 1981, 2002). These dramas incorporate positive
role models demonstrating beneficial lifestyles, negative
role models displaying detrimental lifestyles and
individuals who are making the transition from negative to
more positive life roles. Bandura (2002) reports that these
dramas provide individuals with positive role models. They
also provide the inspiration for viewers to make positive
changes in their own lives. To assist in the change process,
supporting resources on linked websites or in postprogramme
information slots are made available to the
viewers. These dramas are tailored for different cultures and
delivered in Africa, Asia and Latin America (Bandura, 2000;
Brown and Cody, 1991; Singhal and Rogers, 1989; Vaughan,
Rogers and Swalehe, 1995). You see examples of postprogramme
information slots closer to home, with helpline
details being provided after popular soap operas when particular
social issues are included in the programme.
Personality development in social
learning theory
It is this emphasis on observational learning that has led to
the term ‘social’ being included in the theory, to stress that
it is about how people learn from other individuals. In
terms of how children develop their personalities, it is a
learning process where parents, peers and others provide
role models for children to learn from through observational
learning mainly. The children learn to model their
behaviour on successful models in their environment. This
might be a sibling who manages to avoid trouble in one situation
and a friend who gets on well in another and so on.
Unfortunately, parents and others are not always consistent
in their reinforcement, as we have seen, so the picture is
more complicated than it might seem at first. Role models
will be more or less effective, as will individual children’s
learning. Children will be exposed to different experiences,
different environments and different cultures, and all of
these influences help account for the observed diversity of
human beings. The child is at the centre of these learning
experiences and actively shaping the process. It is a truly
dynamic, complex process.
Identifying goals to achieve is a crucial part of this process,
and obtaining external feedback from relevant others on
progress made towards achieving these goals plays an important
part in maintaining motivation and ultimate success
(Bandura, 1991). Bandura also demonstrated that goal
achievement depends heavily on self-regulatory processes
(Bandura, 1990, 1991, 1994, 1999, 2002). These internal
self-regulatory processes include self-criticism; selfpraise;
valuation of own personal standards; re-evaluation of
own personal standards if necessary; self-persuasion; evaluation
of attainment and acceptance of challenges. Bandura
(1990) describes these processes as being attempts at selfinfluence,
and he has shown that the more of these factors
involved in achieving a goal, the higher the levels of motivation
to succeed. He identified self-efficacy as one of the most
powerful of the self-regulatory processes, and we shall examine
it next.
Self-efficacy as a self-regulatory
process
Self-efficacy is defined as being your belief that if you perform
some behaviour, it will get you a desired positive outcome
(Bandura, 1989, 1994). It has become a really hot
topic in psychology during the last 10 or so years and has
stimulated a great deal of research, especially in health.
Individuals have been shown to vary greatly in their levels
of self-efficacy related to specific tasks. If we take smokers
who wish to stop smoking as an example, smokers will
vary greatly in whether they believe that they can achieve
84 PART 1 PERSONALITY
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 84
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 85
their goal. It is a special kind of confidence in your ability
to perform. In the smoking example, it might be that the
smoker felt that not smoking at home and at work was
achievable (high self-efficacy) but that not smoking when
out with friends would be more difficult (low selfefficacy).
Their overall judgement would depend on the
relative amount of time they spent in each activity, and
perhaps on their past experiences of similar success in a
relevant area and so on. Bandura (1997) has shown that
high self-efficacy significantly increases the likelihood of
achieving success. Self-efficacy will influence whether a
task will be attempted as well as the effort put into it and
the persistence with which it is pursued in the face of difficulties
or apparent lack of progress. For example, one
recent study of the factors that affected the likelihood of
relapse in a smoking cessation programme found that low
levels of self-efficacy in individuals was a significant predictor
(Segan, Borland and Greenwood, 2006). Another
study looked at factors that predicted heavy drinking in
anxiety-provoking social situations for a student population
in the United States (Gilles, Turk and Fresco, 2006).
The students most likely to drink heavily in these situations
had low self-efficacy for avoiding heavy drinking in
social situations and a correspondingly high belief that alcohol
facilitated social interactions. Halkitis, Kutnick and
Slater (2005) looked at adherence with HIV antiretroviral
treatment in three hundred HIV-positive men. They
found that poorer adherence was associated with low selfefficacy
towards adherence, amongst other factors. Having
confidence in your ability to succeed at something is consistently
shown to be a significant factor in a wide range of
scenarios (Bandura, 1997).
An example will help to clarify the application of selfefficacy.
Let us compare two students, Dan and Stuart, who
have to give assessed seminar presentations.
One student, Dan, is quite looking forward to his presentation.
He knows that if he does well he will get a good
mark, and he really wants to do well this year to get a good
degree. He knows from experience that although he will be
anxious initially, once he gets started he likes public speaking
and will enjoy it. He is interested in the topic, and he is
already quite well informed about it. He knows that he can
organise his work effectively as he has received good marks
previously when he has given himself sufficient time to undertake
the preparatory work well. Not surprisingly then,
his self-efficacy is high with regard to the seminar presentation;
he feels confident about all the component parts that
go into producing a good seminar presentation, and he has
some positive experiences to reinforce this. The one proviso
he has is about ensuring he has enough time to complete
the task. As his motivation to succeed is high and his
self-efficacy is high, he is more likely to devote the time to
the work. The chances of Dan succeeding in delivering a
good presentation are also correspondingly high.
Stuart dreads the event. He hates public speaking, and
he has no confidence in his ability to master the constituent
parts of the task. He knows that he has to do it, but his selfefficacy
in relation to the task is low. As a result, thinking
about it makes him anxious. He tries to put it out of his
mind, and he avoids cues that remind him about it, like
going to the library to prepare and so on. He indulges in
ostrich-like behaviour and consequently, his chances of
success are reduced. His initial low self-efficacy rating has
resulted in him not being motivated to perform the task.
Self-efficacy has been shown to be an important variable in
predicting educational achievement. Lane and Lane (2001)
showed that self-efficacy was a good predictor of British
students’ achievement on sports science courses. Hoy and
Davis (2006) demonstrated that in school situations,
teacher’s ratings of their own self-efficacy in teaching are
associated with the levels of achievement attained by their
pupils.
Increasing self-efficacy ratings
The good news is that Bandura (1997) has demonstrated
that self-efficacy can be modified by several different methods.
Bandura (1999) has shown that the most straightforward
way to improve self-efficacy is to get the person to
‘perform the dreaded task’. If someone can be encouraged
and supported to do something they fear, it has a dramatic
effect on their self-efficacy. If the level of their performance
is an issue, then further self-regulatory processes may be
called into play so that the individual sees it as a gradual
process. The first goal will be to perform, and subsequent
goals may be about making improvements to their performance.
Vicarious experience has also been shown to
have an effect (Bandura, 1994). This is where the individual
sees someone, whom they know shares the same fears as
theirs, actually performing the task. Their cognitions become
more positive, and they may say something like, ‘If
they can learn do it, then so can I’. This then changes their
self-efficacy directly. The final method is termed
participant modelling. In this method, the person with
low self-efficacy shadows a person who is successfully completing
the task. Even this imitative behaviour has been
shown to lower anxiety.
Returning to the student seminar example discussed
earlier, Bandura’s model outlines three possible courses of
action for the student low in self-efficacy. He would suggest
that observing another anxious student perform successfully
and discussing how they prepared for the seminar
would be very helpful in raising self-efficacy. Taking this
action allows the student to change their cognitions, to see
that they can learn to deliver a good presentation as well.
Another technique would be to pair up the anxious student
with someone less anxious. The anxious student would
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 85
follow the confident student through the preparation and
performance. Any possibility for feedback in the process
would increase the chances of success by increasing the
anxious student’s confidence that they were progressing in
the right direction. So obtaining feedback on the content is
valuable, as is rehearsing the presentation with a friendly
audience of family or friends. Once this rehearsal has been
achieved, it again increases confidence. Any steps to improve
confidence will improve the chances of a successful
outcome.
Self-efficacy and the other self-regulatory processes help
us to maintain our motivation and to be resilient even
when faced with setbacks to our progress. Bandura (1990)
quotes interesting examples of such resilience, including
that displayed by the author James Joyce, whose book
Dubliners was rejected by 22 publishers before becoming a
success. Similarly, the artist Van Gogh died a pauper, having
only ever sold one of his many paintings that are now
worth a fortune. There are many more examples of amazing
resilience shown by individuals in the face of rejection
and apparent failure. Bandura sees the self-regulatory
processes such as self-efficacy as being important in helping
us to survive hard knocks and continue striving to
achieve our goals. Benight and Bandura (2004) published
an extensive review of research undertaken on the role of
self-efficacy in helping individuals recover from traumatic
experiences. They looked at natural disasters, war, terrorist
attacks, loss of a spouse and other interpersonal traumas.
They concluded that individuals who believe that they can
overcome their difficulties (high in self-efficacy) are consistently
shown in all the studies they examined to make a
better recovery.
Measuring self-efficacy
There are psychometric tests that have developed to measure
General Self-Efficacy; Sherer et al.’s (1982) General
Self-Efficacy Scale, Schwarzer and Jerusalem’s (1995) General
Perceived Self-Efficacy Scale, and Chen, Gully and
Eden’s (2001) New General Self-Efficacy Scale. Sherer’s
et al.’s General Self-Efficacy Scale aims to measure selfefficacy
via items that measure a general set of expectations
that the respondent brings into new situations they face.
An example item from this scale is ‘If I can’t do a job the
first time, I keep trying until I can’. Schwarzer and
Jerusalem’s General Perceived Self-Efficacy Scale is designed
to measure general Self-Efficacy by examining the
participant’s beliefs around their own capability to handle
new and difficult tasks in a variety of different domains. An
example item from this measure is ‘I can handle whatever
comes my way’. Chen et al.’s New General Self-Efficacy
Scale is designed to measure the respondent’s belief in their
overall competence at being able to produce a necessary
performance across a variety of possible‘achievement’ situations.
An example item from this scale is ‘I will be able to
achieve most of the goals that I have set for myself’.
However, Bandura is critical of attempts to measure
self-efficacy with global scales. He points out that few people
are confident about every aspect of their lives and the
tasks they have to perform. Hence, self-efficacy is measured
in relation to specific tasks. It demands confidence judgements
to be made about the constituent skills or knowledge
elements that make up a task. Researchers need to undertake
this analysis systematically to ensure that all the relevant
components are assessed. Bandura (2006) provides
86 PART 1 PERSONALITY
Children copy adults’
behaviour in many
different and subtle ways.
Source: David R. Frazier Photolibrary,
Inc./Alamy Images
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 86
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 87
detailed guidance on measuring self-efficacy. Consequently
Bandura suggests that new and separate scales need to be
developed for each self-efficacy domain. So for example,
Bandura (2006) suggests:
Children’s Self-Efficacy Scale, which measures children’s
self-efficacy around a number of learning situations that
students may experience at school, e.g. Self-Efficacy for
Academic Achievement (‘Learn reading, writing, and
language skills’), Self-Efficacy for Self-Regulated Learning
(‘Organise my schoolwork’) and Social Self-Efficacy
(‘Carry on conversations with others’),
Teacher Self-Efficacy Scale, which measures teachers’
self-efficacy around a number of teaching situations in
which teachers may face challenges, e.g. Efficacy to Influence
Decision Making (‘Get the instructional materials
and equipment I need’), Disciplinary Self-Efficacy
(‘Get children to follow classroom rules’) and Efficacy to
Create a Positive School Climate (‘Make students enjoy
coming to school’), and the
Parental Self-Efficacy, which measure parents’ selfefficacy
around a number of situations in which parents
may face challenges, e.g. Efficacy in Setting Limits, Monitoring
Activities, and Influencing Peer Affiliations
(‘Keep track of what your children are doing when they
are outside the home’), Efficacy to Exercise Control over
High-Risk Behaviour (‘Prevent your children from
doing things you do not want them to do outside the
home’) and Resiliency of Self-Efficacy (‘Keep up your
spirits when you suffer hardships’).
Julian Rotter and locus of control
We now want to introduce you to an important concept,
locus of control, that has been and is still used extensively
in research in personality and individual differences. This
concept was first described by Julian Rotter (1966), another
American learning theory researcher, who carried out most
of his research at the University of Connecticut where he
still works. We will begin by examining the theoretical
background to the concept of locus of control before going
on to explore how locus of control is measured. There is a
wealth of research on locus of control, as it has been and
continues to be as popular a research tool as are measures
of self-efficacy. For this reason, we present only a brief taste
of some of the research findings here, with an indication of
the areas of research where it has been applied.
Like Bandura, Rotter felt that animal studies were too
simplistic to address the complexity of human behaviour.
Rotter was interested in how you might predict how individuals
would respond in particular situations. Supposing
someone makes a nasty remark about you in front of other
people. You could respond angrily; you could mock them
for doing it; you could get upset; you could go quiet; or you
could walk away. There are a variety of possibilities. Rotter
(1966) aimed to predict which option an individual might
choose in a particular situation. He termed this the
behaviour potential, that is, the likelihood of a specific
behaviour occurring in a particular situation. The response
that you choose will be the one with the strongest behaviour
potential in that situation. However, the crucial question
is, how is the strength of the behaviour potential
determined? Rotter developed a formula to answer this
question:
Behaviour potential ϭ Reinforcement value ϫ
Expectancy
In this formula, Expectancy is our subjective estimate
of the likely outcome of a course of behaviour. It is what we
expect will happen. In learning theory terminology, it is our
estimate of probability of our behaviour receiving a particular
reinforcement in that situation. This is generally based
Stop and think
Your own General Self-Efficacy?
Take the three statements above from the three General
Self-Efficacy Scales described in the text.
If I can’t do a job the first time, I keep trying until
I can.
I can handle whatever comes my way.
I will be able to achieve most of the goals that I have
set for myself.
Rate yourself on each statement using the scale 1 to
5, with 1 being ‘disagree strongly’, 2 being ‘disagree’, 3
being ‘not certain’, 4 being ‘agree’ and 5 being ‘agree
strongly’. Total up your scores and look at your individual
score for each statement. Though this is not necessarily
an accurate measurement of General Self-Efficacy, what
do your total score and your answers to each statement
tell you about your level of General Self-Efficacy?
Remember each statement refers to a different definition
of General Self-Efficacy.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 87
on our experience of the same or similar situations. In the
nasty insult example, it is your estimate of what you expect
will happen if, for example, you mock the person. You may
estimate that they will blush and feel ashamed of having
made the nasty remark. Each option will have a different
expectancy associated with it. This expectancy influences
how you choose to behave in that situation. The final variable
that contributes to predicting our behaviour is
Reinforcement value. This refers to our preferences
amongst the possible reinforcements available. You may be
more inclined to help someone move some furniture if you
know they will buy you a drink as a thank-you.
To summarise, Rotter suggests that to predict behaviour
in a particular situation, we need to know what the options
are and what the individual sees as being the possible outcomes
for each option. The individual then assesses the
likely outcome of each option (Expectancy). Next, they assess
how much they value this outcome. The behaviour
that is likely to occur (Behaviour potential) will be the
behaviour that gets the highest rating. A summary of this
decision-making process for our hypothetical example is
shown in Table 4.1.
In novel situations, where by definition we have no experience
to guide us, Rotter (1966) suggests that we rely on
what he calls generalised expectancies.What he showed
to be important about this concept is that individuals
come to believe on the basis of their other learning experiences
that either reinforcement is controlled by outside
forces, or that their behaviour controls reinforcement
(Rotter, 1966). The question he was interested in was
whether it makes a difference if people believe that the reinforcement
they receive is linked to how they perform, compared
to individuals who believe that the reinforcement
they receive is unrelated to their own behaviour. He labelled
individuals who believe that reinforcement depends on external
forces as externals. The external forces may include
powerful others in the person’s world, luck, God, fate, the
State, and so on. What is crucial is that externals believe that
the locus of control is external to them. What they do does
not influence the outcomes. Individuals who believe that
their behaviour does make a difference to the outcome are
labelled internals. Rotter (1966) demonstrated that locus
of control is a relatively stable personality characteristic and
developed a scale to assess it, the IE Scale. It is assessed via a
30-item forced-choice scale. Scores are on a continuum of
I–E, and Rotter does not suggest a cut-off point to separate
externals from internals. He has published normative scores
for particular groups to allow comparisons to be made.
Although other assessment tools to measure locus of control
have been developed since Rotter’s scale was published,
his IE Scale is still the most widely used in research. Some
sample items from the scale are shown in Figure 4.4.
The impact of locus of control
on behaviour
Rotter (1982) demonstrated that people with an internal
locus of control are more likely to feel in control of their
lives, and to feel empowered to try to change things in their
environment. Individuals with an external locus of control
are more likely to feel powerless and helpless to change
things and to be dependent on others. Research has shown
that internality increases with age. Children become more
internal as they develop into adulthood. Internality becomes
stable in middle age and does not decrease in old
age. Warm, supportive parents who encourage independence
in their offspring have been shown to foster the development
of internality in their children (de Mann, Leduc
and Labrèche-Gauthier, 1992).
Locus of control scores tend to correlate with anxiety,
and there tend be more externals than internals among
people with mental health problems (Lefcourt, 1992). A
major review of studies on depression and locus of control
concluded that external scores correlate positively with
higher levels of depression (Benassi, Sweeney and Dufour,
88 PART 1 PERSONALITY
Stimuli: Someone you know, Angela, makes a nasty remark about you in front of other people.
Table 4.1 Application of Rotter’s equation for predicting behaviour to an insult.
Behavioural option Possible outcome
Rating of expectancy
of outcome
Value of the outcome
to the individual
Behaviour potential
(probability that
option will occur)
Angry reply Argument High Low Low
Mocking comment Angela is embarrassed High High High
Get upset Angela feels remorse Low High Low
Say nothing Feel silly High Low Low
Walk away Feel silly High Low Low
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 88
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 89
Respondents are asked to circle either of the two statements to indicate which
statement they agree with.
Item 2
· Many of the unhappy things in people’s lives are partly due to bad luck.
(external locus of control)
· People’s misfortunes result from the mistakes they make. (internal locus of
control)
Item 9
· I have often found that what is going to happen will happen. (external locus
of control)
· Trusting to fate has never turned out as well for me as making a decision to
take a definite course of action. (internal locus of control)
Item 29
· What happens to me is my own doing. (internal locus of control)
· Sometimes I feel that I don’t have enough control over the direction my life is
taking. (external locus of control)
Figure 4.4 Sample items from Rotter’s Locus of Control Scale.
Source: Rotter (1966).
1988). This link with externality and depression is still
reported currently, and it also links with suicidal behaviour.
Cvengros, Christensen and Lawton (2005) examined
relationships between locus of control and levels of depression
in patients suffering chronic kidney disease. They
compared patient scores on health, locus of control,
depression and progression of the illness over a 22-month
period. Results demonstrated that patients who experienced
an increase in their locus of control scores, demonstrating
that they felt that they had more control over
aspects of their condition, were less likely to be depressed.
Liu, Tein, Zhao and Sandler (2005) surveyed 1,362 adolescents
in five schools in rural China and examined the relationships
between locus of control, suicidal behaviours, life
stressors, depression and family characteristics. They
reported that high scores on the external locus of control
were a risk factor for suicidal ideation and suicide attempts,
along with high life stress, increasing age and depression.
A similar pattern is found for physical health, with
internals becoming better informed about their illness and
coping better with physical illness. Externals are more likely
to adopt a passive patient role, while internals are more
likely to get involved in their treatment by adopting healthier
behaviour (Powell, 1992). Internal locus of control had
been shown to be associated with improved quality of life
in patients undergoing treatment for HIV (Préau and the
APROCO study group, 2005). The study assessed quality of
life, locus of control and demographic and health factors
in 309 HIV-infected patients at the start of their treatment
programme and then monitored the sample over
44 months of treatment. After 44 months of treatment,
internal locus of control was a determinant of higher quality
of both physical and mental health. Similar results,
demonstrating better quality of life for internals, have been
reported for individuals suffering from chronic illnesses
such as epilepsy (Amir, Roziner, Knoll and Neufield, 1999),
diabetes (Aalto, Uutela and Aro, 1997), and migraines
(Allen, Haririfar, Cohen and Henderson, 2000).
Locus of control has also been shown to impact on behaviour
in many other situations. Lerner, Kertes and Zilber
(2005) carried out a study examining risk and protective
factors in psychological distress experienced by six thousand
immigrants who had come to Israel from Russia. In a
survey taken five years after the immigration, the
researchers showed that psychological distress levels in the
participants were linked with having an external locus of
control as well as with other negative health and social factors.
Locus of control is also applied in organisational
research. For example, Allen, Weeks and Moffat (2005)
looked at the role of locus of control among other variables
in predicting whether employees acted on their intention
to change jobs, or whether they simply talked about it.
They found that individuals with an internal locus of control
were more likely to translate their intention to change
jobs into action and change their job.
Locus of control has also been applied in educational
contexts. Martinez (1994) showed that internals tend to
achieve greater academic success than externals do. It is
suggested that when internals do well in examinations or
essays, they tend to attribute their success to their own
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 89
abilities or to having worked hard. Externals, on the other
hand, are more likely to put their success down to luck or
an easy test. These differences in causal attribution will
affect the confidence with which internals and externals
approach academic assessment. Bender (1995) has suggested
that the experience of continued failure despite trying
at school leads to the development of an external locus
of control in schoolchildren. They see that trying hard
brings no reward, so they give up and may come to see failure
as their destiny. Anderman and Midgley (1997) suggest
that in the circumstances of repeated failure, having an
external locus of control protects the individual’s selfesteem.
It is then not their fault that they fail. Internals, on
the other hand, will be more confident and have higher
expectations of themselves, both of which increase their
probability of success. With very few exceptions, it appears
that internals are more successful than externals in most
situations. However, remember that the IE Scale is a continuum,
and scores tend to cluster around the middle of
the scale with few very extreme scores.
Walter Mischel
We debated whether to place Mischel’s theory in Chapter
5 on cognitive theories; however, we decided to include it
in this chapter because of its focus on the importance of
situations and as it is frequently described as a socialcognitive
approach to personality. Mischel was also heavily
influenced by Bandura’s work on self-efficacy and
Rotter’s approach to personality measurement. Mischel’s
theory could equally well sit in Chapter 6 on cognitive
theories.
In 1968 Walter Mischel published Personality and
Assessment, a book that created enormous controversy in
personality psychology. As outlined in the Profile box (see
page 91), Mischel’s own research on the efficacy of global
personality traits to predict performance led him to question
the stability of personality traits across situations. This
became known as the person–situation debate, or the ‘personality
paradox’. The question to be addressed was, do you
behave as you do because of the situation you are in, or is it
because of your personality? Mischel (1968) was concerned
about the way psychologists interpreted personality test
scores and then used them to make decisions about individuals.
He pointed out that traits and other measures of
personality are not good enough as predictive measures of
how an individual will behave in different situations that
they can be used to make important judgements about that
individual, such as whether they are the right person for
the job or if they are likely to violate the conditions of their
parole. He pointed out that there was little evidence that
individuals’ behaviour is consistent in different situations.
Mischel (1973, 1979, 1983a, 1983b) makes it clear that despite
what some critics said, he was not questioning the
existence of personality traits but simply the way they were
interpreted.
Mischel (1968) claimed that the correlations between
personality trait self-report measures and behaviour was
between 0.2 and 0.3, meaning that the trait was accounting
for under 10 per cent of the variance in behaviour. He
termed this correlation between traits measures and behaviour
the personality coefficient. Other researchers questioned
the size he claimed for the personality coefficient,
demonstrating that a more realistic figure is 0.4 (Nisbett
and Ross, 1980) – which is still low.
Personality researchers tried to combat Mischel’s
argument by comparing how well situations and traits
predict behaviour (Endler and Hunt, 1966, 1968). The
conclusion was that knowing about both the situation
and the personality was better than knowing about either
one on their own. However, this approach is impractical
as there are so many possibilities, and researchers have to
make decisions about which traits are likely to be relevant
in particular situations. Mischel (2004) cites a study
undertaken by Newcombe (1929) where 51 boys were
measured on the personality characteristics of extraversion
and introversion and then studied in 21 situations in
a summer camp on a daily basis. Systematic recordings
were made of the amount of time each boy talked at
meals and of how much time he played alone or with
others. Much to his dismay, Newcombe found that the
average correlation of behaviour across situations based
on these daily observations was 0.14. Mischel and others
have continued to examine the consistency in behaviour
that individuals display across situations and have concluded
that there is substantial variation (Mischel, 1968,
1973; Mischel and Peake, 1982; Moskowitz, 1994; Ross
and Nisbett, 1991).
Epstein (1979, 1980) argues that most personality
researchers do not measure the relationships between personality
traits and behaviour correctly. They take a personality
score and then take one measure of behaviour, such as
the likelihood of offering to donate blood, rated on a Likert
Scale. This violates the principles of good measurement
(Epstein, 1980). You need multiple measures to ensure reliability.
For example, if we wanted to compare the different
amounts of time that introverts and extraverts spent studying,
a reliable measure would not be obtained by asking
them how long they studied the previous evening. We
would have to measure their study habits over some more
extended time to get a true picture. This is what researchers
have done to address this issue of the variability across situations
and the associated measurement error. Behaviour
measures from individuals are aggregated. Using this
approach, researchers have demonstrated that there are
90 PART 1 PERSONALITY
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 90
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 91
Profile
Walter Mischel
Walter Mischel was born in Vienna in 1930, in a house
that was a short walk away from where Freud lived. His
family moved to New York when he was 10 years old to
escape from the Nazis. He studied psychology but qualified
as a social worker. He suggested that the early link
with Freud led him to begin his career as an advocate of
Freud and psychoanalysis. However, he found that the
psychoanalytic approach was of little help in his work
with inner-city aggressive youngsters. This led him to
undertake a PhD in psychology at Ohio State University.
Here he worked with George Kelly (Chapter 5) and
Julian Rotter (this chapter). After graduation he worked
at Harvard and then Stanford Universities before moving
to his present post at Columbia University in 1984.
While at Harvard he worked on a project assessing
performance for the Peace Corps and found that global
trait measures of personality were not good predictors
of performance. This led him to question existing
approaches to personality, as we shall see. He received
the Distinguished Scientist Award from the Clinical
Division of the American Psychological Association in
1978.
stable individual differences between individuals on almost
every dimension studied (Epstein, 1979, 1980; Mischel and
Peake, 1982; Pervin, 1994). Epstein (1979) compared extraversion–introversion
scores in students with the number of
social contacts they made, where the contacts were
recorded in daily diaries over a two-week period. They
found a personality coefficient of 0.52, which is a major
improvement on previous figures. What aggregation does
is to minimise the effect of the situation so that the stable
underlying characteristics of the individual become
apparent.
Mischel (2004) points out that the person versus situation
debate caused real divisions between personality psychologists
looking to show consistent differences in
individuals that are independent of the situations they are
in, while social psychologists stressed the importance of the
situation (Nisbett and Ross, 1980; Ross and Nisbett, 1991).
Some personality psychologists did examine the personsituation
interaction in more detail (e.g., Fleeson, 2001;
Moskowitz, 1994; Vansteelandt and Van Mechelen, 1998);
but as Mischel (2004) points out, these were rare
exceptions.
For Mischel, the way forward was to incorporate the
findings from developments in cognitive psychology about
how the mind works. Mischel (1973) outlined a set of
social-cognitive person variables, as opposed to trait
descriptors, to describe individual differences. These variables
described processes that were important in describing
how individuals construed situations (encoding and appraisal),
variables relating to the situation (people and the
self) and the beliefs, behavioural expectancies, goals and
processes of self-regulation. The aim was to discover the
psychological processes in order to determine how individuals
characteristically interpreted the world and how
particular situations produced characteristic behaviour in
individuals. It is an interactional approach that still aims to
uncover individual personality differences; but in Mischel’s
approach, these differences are not encapsulated in situation-free
personality trait terms like ‘optimistic’, ‘considerate’,
‘sociable’, but in situation-related descriptions of how
individuals characteristically behave. Examining some of
the research undertaken by Mischel and his colleagues will
make it clearer.
Mischel and Peake (1982) examined what they called
‘college conscientiousness’ and friendliness in college students.
To begin with, the students themselves specified the
behaviours and situational contexts that they considered
relevant to the traits being examined. This ensured that
the behaviours and situations being measured were personally
meaningful for the college students’ definitions of
‘college conscientiousness’ and friendliness. While the researchers
found behavioural variability across different
situations, so that one person might be very friendly in
one situation but low in another, they also found temporal
stability in individual’s behaviour within similar situations.
Mischel suggested that some situations were
perceived as being highly similar, forming what he describes
as a functional equivalence class of situations.
Individuals perceived themselves as having the
personality characteristic of friendliness, for example,
based on how consistently they behaved within a particular
situation rather than on how they behaved across situations.
In other words, you might well see yourself as
being high in the characteristic of friendliness even
though you do not act in a friendly way in every situation.
From this study, Mischel and Peake (1982) concluded
that there was consistency in how individuals behaved
within a situation, and there appeared to be consistent
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 91
differences between types of situations that were worth
exploring further.
While acknowledging that some of the differences between
situations might be random noise, Mischel and his
colleagues were convinced that there were also systematic
differences in the perception of situations. They set out
to look for some underlying structure that would help
explain where these differences came from. An example
may help your understanding here. Mischel (2004) compares
two individuals who have the same score on a personality
trait measure of aggressiveness. However,
observation of their behaviour shows that they behave
aggressively in very different situations and that these
differences in their behaviour patterns are stable. One is
aggressive to his junior colleagues at work but very
friendly to his superiors, while the other is very friendly
with colleagues but aggressive to his superiors. In this example,
simply describing them as equally aggressive
based on a trait measure does not give a real description
of how they differ in terms of their personality
characteristics.
Searching for invariance in an individual’s behaviour
across situations is a massive undertaking; but Mischel
and his colleagues ran a replication of the Newcombe
(1929) study, using a residential summer camp set up to
treat children with behavioural problems, particularly aggression
(Mischel, Shoda and Mendoza-Denton, 2002;
Shoda, Mischel and Wright, 1993, 1994). The children
were filmed over many weeks and for many hours. Mischel
et al. reported that aggressive behaviour observed in
one type of situation was not a good predictor of how
that individual would behave in another type of situation.
This in itself is not surprising, but Mischel et al. demonstrated
that an individual’s rank-order position on aggression
relative to others in the group changes
predictably and dramatically in different situations. The
conclusion was that individuals might have a similar
mean level of aggression but there are predictable differences
in terms of which situations they behave aggressively
in, and these provide much more insight into the
kind of person they are. It might be that one child is aggressive
to his peers when asked for anything, but another
child with the same aggression trait score might characteristically
only be aggressive to adults when they are
chastising him. Mischel and his colleagues helpfully
characterise these stable situation-behaviour relationships
with the phrase, ‘if . . . then . . . ’, and describe them
as providing a behavioural signature of personality
(Shoda, Mischel and Wright, 1993, 1994). These behavioural
signatures represent our characteristic reactions to
situations. Other researchers have confirmed the existence
of behavioural signatures of personality that provide
distinctive characterisations of individuals
(Andersen and Chen, 2002; Cervone and Shoda, 1999;
Morf and Rodewalt, 2001; Shoda and LeeTiernan, 2002).
To summarise, Mischel and his colleagues have described
the two types of behavioural consistencies. The first is behavioural
consistency, called type 1 consistency, and it
represents the trait ratings describing what individuals
are generally like. Type 2 consistency represents the behavioural
signatures of personality which show distinctive
patterns of behaviour across similar situations, the
if . . . then . . . propositions that encapsulate patterns of
situational effects on personality.
Mischel argues the need for a dynamic personality
system that will incorporate developments from cognitive
science and genetics that are relevant to personality.
He suggests that a dynamic personality system will go
beyond mere descriptions of personality and give us
more information about how the individual mind functions
and personality is organised. He suggests that information
about the individual’s mental and emotional
processes is an essential component of any model of personality.
Mischel and Shoda (1995) outlined a model of
a cognitive-affective processing system (CAPS)
that fulfils some of these criteria. The aim was to
demonstrate how the CAPS model can predict the type 1
and type 2 consistencies in personality that are described
earlier. CAPS is composed of various mental
representations, labelled cognitive-affective units
(CAUs). These CAUs include the individual’s representations
of self, others, situations, expectations, beliefs,
long-term goals, values, emotional states, competencies,
self-regulatory systems and memories of people and past
events. Mischel and Shoda (1995) propose that the CAUs
are organised in an interrelated system within the individual’s
stable networks of cognitions and emotions. A
diagram showing how the CAPS model operates is
shown in Figure 4.5. The yellow box contains developmental
influences, and the green arrows indicate how
these influences affect the system. All the other interactions,
indicated by blue lines and arrows, are envisaged
to happen concurrently.
Higgins (1996) has demonstrated that within one individual,
some representations are more accessible than
others.
This differential accessibility of CAUs, and the differences
in the ways that they are interrelated within each individual,
both contribute to the observed differences in
personality between individuals. Different CAUs will be
activated in different situations and at different times, but
the way that change occurs does not vary, reflecting the
stability of structures within the individual’s CAPS (Mischel
and Shoda, 1995; Shoda and Mischel, 1998). The
CAPS model has been shown to generate type 1 and type
2 behavioural consistencies in computer simulations
(Mischel and Shoda, 1998; Shoda, LeeTiernan and
Mischel, 2002).
92 PART 1 PERSONALITY
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 92
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 93
Behavioural
consequences
Psychological
features of
situations
if . . . then . . .
propositions
(linking
situations and
behaviour)
Developmental influences
Cognitive-affective
processing system
Biological history
Cognitive social
learning history
Genetic endowment Cultural and social factors
Internal feedback
The CAPS model produces descriptions of personality
types based on how individuals organise their CAUs and
how they process situational features. Downey, Feldman
and Ayduk (2000) have provided an example describing
the way that individuals who fear rejection respond to
perceived uncaring behaviour in their partners, such as
when their partner is paying attention to another person.
Rejection-fearful individuals perceive, interpret and evaluate
their partners’ behaviour in terms of potential
rejection. They ruminate about it, and these ruminations
instigate the emotional responses of anger and fear as the
individual becomes more fearful of being abandoned.
They then respond to their partners by activating controlling,
coercive behaviours and blaming their partner for
this. This then creates a self-fulfilling prophecy as their
partner in turn gets angry and may respond with threats
of rejection. This response from the partner then reinforces
the rejection-fearful individual’s feeling that they
are right to fear rejection – oblivious to the role that they
themselves played in generating the rejection threat. From
these observations, the personality signature of a rejection-fearing
individual is apparent. When appraising
interpersonal situations, they anxiously look for evidence
of potential rejection, any evidence of rejection threat is
magnified and they overreact to it with anger and blame
(Downey et al., 2000).
This analysis demonstrates how personality signatures
provide a more in-depth analysis of individual differences
by incorporating situation-specific information or, in this
instance, relation-specific information. Shoda et al. (2002)
point out that in interpersonal situations, the ‘situation’ is
another person, and they have demonstrated that the CAPS
model deals equally well with this case. A great deal of work
has already been accomplished on classifying different
types of situations. Kelley et al. (2003) have published an
Atlas of Interpersonal Situations. The aim was to go beyond
a superficial description of situations and identify the psychologically
important aspects of situations that play a
functional role in generating behaviour. It is about the way
that types of individuals characteristically perceive a situation,
as demonstrated in the rejection-fearful example.
More work is needed to develop a better understanding of
how situation-behaviour signatures work and to link them
with types of individuals.
What Mischel (2004) is arguing for is an approach to
personality research that integrates research findings from
Figure 4.5 Mischel and Shoda’s (1995) cognitive-affective processing system (CAPS).
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 93
other areas of psychology; he is arguing that cognitions,
memories, emotions, perceptual processes, genetic influences,
regulatory systems and memories all play a part in
generating individual differences. As we have seen, Mischel
and his colleagues have already demonstrated, with their
CAPS model, that there is a complex interaction between
situations and enduring individual personality differences.
There are still debates about the details, and the effects of
many variables still have to be examined. Carver and
Scheier (2003) have been examining the self-regulatory
process by looking at the relationships between behavioural
goals and the effects of feedback on goals. It is a
complex undertaking.
The impact of Mischel
Mischel (2004) relates an amusing incident where one of
his students reports to him that, according to a multiplechoice
question in a state licensing exam for psychologists,
he does not believe in personality. After reflection, Mischel
suggests that if personality is defined purely by trait and
state measures (Chapter 7), the answer is true. However, he
now believes that personality research is moving on; a new
era is emerging as researchers are returning to the original
aim of personality theorising, which was to understand the
systems that produce individual differences in behaviour.
Mischel’s work has been a major stimulus in these developments,
bringing closer the possibility of an overarching
explanatory theory of personality.
Mischel’s original paper in 1968 has had a major effect
on personality research. Swann and Seyle (2005) conclude
that initially it led to a decline in research in personality for
about 10 years and that social psychologists began to focus
on the impact of situations and de-emphasise any personality
effects in their research. However, Mischel’s paper did
lead to significant improvements in personality research.
There have been many rebuttals of Mischel’s views that
involved researchers looking very critically at their
methodologies, admitting that measures were often weak
and the selection of which traits to study was sometimes
inappropriate (Baumeister and Tice, 1988; Bem and Allen,
1974; Funder, 1999, 2001).
The concern about the validity of personality tests led
Cronbach and Meehl (1955) to develop a clear procedure
for establishing the construct validity of psychological
tests. There is now widespread adherence to these procedures
in personality test construction, resulting in
improved validity of tests, and more care is taken in the
interpretation of test scores (Swann and Seyle, 2005).
Meyer et al. (2001) have demonstrated that personality
tests now share the same high levels of validity as seen in
medical tests. More attention has also been paid to the design
of studies, with variables being more carefully
selected and operationalised (Block 1977; Funder, 1999,
2001, 2002).
The grand explanatory theory of personality that
Mischel envisages has not yet emerged, but considerable
progress has been made in resolving the person-situation
debate and in developing our understanding of how the
two interact to produce both consistency and change in behaviour.
Situations affect individuals; but individuals also
act to change situations, often in complex ways (Magnusson,
2001). Swann and Seyle (2005), in their review of the
effects of Mischel’s attack on traditional approaches to personality,
conclude that while the new integrative approach
is likely to be fruitful, there are still instances where it is
helpful to make distinctions between personal and situational
determinants of behaviour.
Evaluation of learning theory
approaches
We will now evaluate learning theory approaches using
the eight criteria we identified in Chapter 1: description,
explanation, empirical validity, testable concepts, comprehensiveness,
parsimony, heuristic value and applied
value.
Description
Both classical and operant conditioning provide useful descriptions
of relatively simple behaviour. However Pavlov,
Skinner, and Dollard and Miller all based their research on
observations of pigeons, rats and dogs; and the more developed
and unique qualities of human beings, such as the
effects of language on our behaviour, are largely ignored.
Skinner (1963) did address this criticism by agreeing that
human behaviour was very complex and therefore difficult
to study. However, he strongly felt that the basic principles
governing the way we learn behaviour are the same for humans
and other animals. It was simpler to study animals in
the laboratory, and as the testing conditions could be controlled
very rigorously with animals, it was better science,
as far as Skinner was concerned. These views of Skinner’s
were contentious and generated as much debate as Freud’s
theory. Opponents argued that people are capable of
higher cognitive processing, resulting in more complex
learning than observed in rats and pigeons, and that the
principles of classical and operant conditioning do not really
address that complexity (Bailey and Bailey, 1993;
Garcia, 1993).
Bandura and Rotter addressed this issue by abandoning
animal studies and by allowing for the effects of inner
94 PART 1 PERSONALITY
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 94
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 95
mental processes on human behaviour. In this way, they
both provided a more comprehensive description of
human behaviour although it was nothing like the complexity
of Freud’s work. By examining the importance of
self-efficacy and locus of control, they have provided descriptions
of valuable personality processes. Mischel’s work
follows in the tradition of Bandura and Rotter, but his ultimate
aim is for an overarching, integrative theory of per-
sonality.
Explanation
The principles of learning theory do provide valid explanations
of observed behaviour in specific situations. However,
human beings have a rich mental life, which is ignored
in the behavioural approaches. We are all capable of thinking
and feeling, and these inner mental processes are
ignored. As we have seen, many of the psychoanalytic theorists
suggest that we are not always conscious of the reasons
for our behaviour. Ruling out any idea of unconscious motivation
as it cannot be directly observed seems absurd to
such theorists. Dollard and Miller allowed for a concept of
the unconscious, but they did not explain the role these
processes play in determining behaviour.
These approaches can be criticised as being as deterministic
as Freud’s. The individual has no free will; our
behaviour is determined by how others react to us. With
the exception of the more recent work of Bandura, Rotter
and Mischel, learning theory cannot explain intentional
behaviour. We may have long-term goals that are unconnected
with our prior learning history. An example will
help illustrate this. Imagine that an individual grows up
in a family where her mother and father were both doctors.
The parents had a burning ambition for their
daughter to follow in their footsteps, and she was certainly
intelligent enough to achieve the necessary academic
qualifications. According to the learning theorists,
this example should result in the daughter becoming a
doctor as she was brought up in an environment that fostered
this, and her academic ability did not provide a bar.
However, in this instance the daughter became a librarian.
This is just one case history, but we are sure that if
you ask around among your friends, you will come across
other examples where children do not follow the paths
that parents have wished for.
Sometimes we all do the unexpected in situations, and
learning theory principles cannot easily explain this creativity
in behaviour. Skinner (1972) rejected this criticism
and said that it applies to classical conditioning with its
emphasis purely on the stimulus and the response. In operant
conditioning, we may behave in new and creative ways;
but whether we repeat the behaviour is determined by its
consequences. If we are rewarded in some way for the behaviour,
then it should occur again, whereas if we are not
rewarded, it should extinguish. I am sure you will agree
that this is certainly not the case for much of human behaviour.
Many would-be novelists continue to write, yet no
one will publish their work; similarly, musicians continue
to compose although no one plays their work, inventors
continue to invent despite a lack of success and so on. We
can still maintain goal-directed behaviour in the absence of
positive consequences. Other personality theorists here
might talk to inner drives that motivate us to behave in
certain ways.
Bandura is the only learning theorist who addresses this
issue with his concept of self-regulatory processes and selfreinforcement.
These concepts allow for intentional behaviour
and for behaviour to continue in the absence of any
external reinforcement. The recognition of the role of cognitive
processes and social factors in behaviour result in
Bandura and Rotter’s theories being very different from the
earlier theories, although they still have the same emphasis
on learning being a sufficient explanation for the development
of personality. Mischel’s position goes beyond that of
Bandura and Rotter, although their approaches would be
included as constituent components of an integrative theory
as the effects of learning still need to be explained
within such a theory. Rotter included the effects of memories
of previous situations (prior learning) explicitly in the
description of his theoretical approach.
For these learning theorists, any similarities in the way
that people respond in different situations are down to
environmental factors and prior learning. The environment
that the person occupies is similar to a previous situation
they had experienced; hence, they are responding
in a similar fashion, rather than expressing a particular
character trait that they possess. This rejection of the idea
that people possess individual characteristics that influence
how they behave in different situations flies in the
face of all the empirical studies of stable measured individual
differences in behaviour that are evidenced in
chapters of this book. Again, Bandura, Rotter and Mischel
are exceptions in that they have each identified individual
personality characteristics. Both Bandura and Rotter see
these differences resulting from learning experiences. At
no point do Bandura and Rotter acknowledge a role for
any possible genetic inheritance of personality traits – unlike
Mischel, who is clear that biological factors have a
part to play.
Empirical validity
One strength of the learning theorists is that their work
is based on empirical data collected under controlled
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 95
96 PART 1 PERSONALITY
laboratory conditions. However, researchers such as
Black (1973) have suggested that these theorists sometimes
go beyond the data they have collected and make
assumptions. This is especially the case with regard to
complex human behaviour. Much of the empirical data
is about animal behaviour, but the assumption is that
the principles uncovered with animals will apply to
human beings. They have examined very simple learning
situations in animals and then go on to assume that
somehow, some combination of the same learning principles
can be used to explain much more complex behaviour
in human beings (Skinner, 1973). There is no
empirical evidence for these claims. Bandura and Rotter
do not use animal studies, but even in their human studies,
they too are sometimes guilty of making assumptions
that go beyond their data. The same cannot be said
about Mischel, although the early critiques of his position
claimed that he did. He successfully refuted these
claims, as overviews by Snyder and Ickes (1985) and
Swann and Seyle (2005) make clear.
Certainly the concepts of classical and operant conditioning
can be, and have been, tested quite exhaustively.
The argument is not that we cannot demonstrate the
occurrence of both classical and operant conditioning;
rather, that the concepts are not sufficient in themselves as
an explanation of human behaviour. The animal explanation
applies here also. The concepts may have been adequately
tested in regard to rats, pigeons and dogs; but this is
not the case for much of human behaviour, especially the
more complex human behaviour.
Testable concepts
With regard to Bandura, Rotter and Mischel’s work, the
concepts they have developed have been extensively tested
in a variety of psychological disciplines, and there is a great
deal of supporting evidence. They have provided useful
conceptualisations of elements of the process of acquiring
personality. Mischel has gone further and caused personality
psychologists to improve their methodologies and
measurement tools.
Comprehensiveness
Skinner rejected the idea of personality and did not see
himself as creating any theory; rather, he tackled specific
problems in learning and behaviour. Taken at this level, he
has provided a sound explanation of some aspects of learning
and some specific behaviour, although most of the emphasis
has been on lower animals, not humans. Within his
research, Skinner focused on simple behaviours as they
were easy to control, but this has resulted in his work
failing to address adequately the complexity of human behaviour.
This was also true of Pavlov, who was purely interested
in learning mechanisms.
Dollard and Miller, Bandura and Rotter were interested
in personality and in developing relevant theories.
Dollard and Miller’s attempts were not very comprehensive.
They did provide learning theory descriptions for
how some Freudian defence mechanisms could be acquired.
However, they fell short of developing a comprehensive
theory of personality. Bandura’s theory is
probably the most comprehensive, but the lack of any discussion
of genetic influences on personality development
is a weakness in all the learning theory approaches. While
Rotter has not yet produced a detailed comprehensive
theory of personality, he has outlined the major components
of such a theory.
Parsimony
From what we have discussed so far, it is apparent that
learning theories can be criticised for being too parsimonious
to adequately explain all of human behaviour and
human motivation. The approaches are very parsimonious;
they assume a small number of principles will
apply to all situations, sometimes without empirical evidence.
Towards the end of his life, Skinner did accept that
additional concepts might be necessary to explain the
more complex learning that occurs in humans. This criticism
cannot be applied to Rotter’s work as it aims to incorporate
relevant explanatory and organisational
concepts from all areas of psychology relevant to
personality.
Heuristic value
As we have seen, both classical and operant conditioning
have had an enormous impact on the discipline of psychology.
Firstly, by emphasising the importance of empirical
research evidence in theory development and
hypothesis testing, the learning theorists played a major
role in shaping psychology as an empirical science. They
also demonstrated the importance of attending to situational
and environmental variables that may affect behaviour
in any situation and led to an early emphasis on
laboratory studies where such variables can be more readily
controlled. This early work has generated and continues
to stimulate research within psychology, as evidenced
by the continuation of the Journal of Experimental Analysis
of Behaviour, which is devoted to learning theory approaches
to research. Skinner himself has been a controversial
figure, and his work has created great debates
within psychology, psychiatry, education, philosophy, politics
and the general public.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 96
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 97
Bandura and Rotter, with their concepts of self-efficacy
and locus of control in particular, have stimulated a huge
amount of research. Mischel has also stimulated a great
deal of controversy and research in both personality and
social psychology, and it may well be that Rotter’s work has
created major changes in the discipline, some of which are
yet to become apparent.
Applied value
In terms of applications of psychology, all the learning theorists
have advocated the adoption of very pragmatic approaches
to disturbed behaviour, and this has led to many
new treatments for mental illness. By focusing on the detail
of the ill person’s behaviour, they have provided unique
understanding of how such behaviour may have arisen
in the individual’s previous learning. The concept of disturbed
behaviour as a maladaptive response that has previously
been reinforced immediately opens up the possibility
of that behaviour being extinguished and new responses
being acquired. This concept also helps to demystify mental
illness and consequently, it can be presented as a positive
approach to mental illness.
The concepts of self-efficacy and locus of control have
both been valuable additional factors to consider in behavioural
change programmes. Programmes have been
developed to improve self-efficacy in treatment programmes
ranging from smoking cessation to safe sex
campaigns. Similarly, locus of control has proved a useful
tool in understanding treatment compliance issues in a
variety of areas. Mischel too has always been interested in
clinical aspects of psychology, and his work has led to better
understanding of how personality attributes interact
in situations to amplify disturbed behaviour, as in our example
of the rejection-fearful individual.
This very idea of changing behaviour also leads to concerns
about the potential to apply learning theory in unethical
ways to mould both individual behaviour and that of
societies. One example that we have already examined is the
development of experimental neuroses in the Little Albert
case study. As we have seen in his novel Walden Two, Skinner
also acknowledged this concern. Behavioural approaches
need to be applied ethically, as with all attempts to change
behaviour – hence the importance of research ethics. Ethical
issues in relation to personality research are discussed further
in online Chapter 28 (www.pearsoned.co.uk/maltby).
Final comments
In this chapter you were introduced to the learning theory
approach to personality, notably Pavlov and classical conditioning,
Watson and behaviourism, Skinner and operant
conditioning, the integrative personality theory of Dollard
and Miller and the social cognitive approaches of Bandura.
You were also introduced to the concept of self-efficacy,
Rotter and the locus of control and Mischel and social
learning theory. You should also be able to critically evaluate
each of these theories. You should also be able to
broadly evaluate learning approaches to personality.
Summary
The early learning theories reject the idea of our
behaviour being directed by inner motives. All our
behaviour is learned. Individual differences in behaviour
are the result of the different learning experiences
that people have had and the situations that
they have experienced rather than being due to differences
in personality.
Pavlov demonstrated how behaviour is learnt via
classical conditioning. The process begins with an
unconditioned stimulus (e.g., food), which is something
that automatically produces the response you
are interested in, called unconditioned response
(salivating).
Pavlov demonstrated that the acquisition of many of
our emotional responses can be explained by classical
conditioning.
Watson, influenced by Pavlov’s work, called for the
adoption of rigorous scientific method in psychology
and for theory building to be based on empirical evidence
rather than the introspection, reflection and
anecdotal case study methodologies of the psychoanalytic
school.
For the learning theorists, psychopathology is due to
faulty learning. Normal development is about learning
responses that are adaptive in the individual’s environment.
Abnormal development occurs when
maladaptive responses are learned.
Skinner was a radical behaviourist, and he did not
allow for inner experiences in his account of learning.
As inner experiences could not be observed,
he therefore considered them unscientific. Only
behaviour that could be observed was included in
M04_MALT2908_02_SE_C04.QXD 10/23/09 11:10 AM Page 97
98 PART 1 PERSONALITY
his model of learning. For the same reason, he rejected
the concept of personality as being produced
by the interaction of inner forces. All
behaviour was learnt. He demonstrated three key
concepts important for learning: operant conditioning,
positive reinforcement and negative re-
inforcement.
Skinner’s theory is deterministic. He rejects the concept
of free will and the idea of intention or creativity
in human behaviour. We merely respond to stimuli in
our environment, and the consequences of our responding
determine our learning.
A strength of all the learning theory approaches is
their emphasis on the application of rigorous
methodologies to collect data and the underpinning
of all theory with empirical data. Criticisms
are that they fail to address the complexity of
human behaviour. They are too heavily grounded
in animal studies and have a very limited conceptual
basis.
Dollard and Miller made the first attempt to allow for
cognitive processing in learning theory. They allow
for unconscious influences on motivation but strictly
define what they mean by the unconscious.
Dollard and Miller outlined a stimulus-response (S-R)
theory of learning. This includes the consideration of
primary drives and secondary drives.
Dollard and Miller demonstrated that observational
learning played an important in role in learning. Role
models are observed, and their performance is
imitated.
Bandura was the first learning theorist to allocate a
significant role in learning to inner cognitive
processes. Bandura uses the term reciprocal determinism
to label the processes that drive behaviour.
He sees an individual as being influenced by three
interacting factors: personal factors, behaviour and
environmental factors.
Bandura further develops Dollard and Miller’s concept
of observational learning and has demonstrated its
importance in the acquisition of aggressive behaviour
in particular with the Bobo doll study. He demonstrated
that the characteristics of the model, attributes of the
observer and the consequences of imitating behaviour
are all influential factors in the learning process.
For Bandura, modelling behaviour was an active
process of learning through observation where the
observer makes judgements and constructs symbolic
representations.
Bandura demonstrated that we humans use selfreinforcement
to control our behaviour via internal
self-regulatory processes.
Self-efficacy is identified as one of the most powerful
of the self-regulatory processes.
Rotter demonstrated that the likelihood of a
behaviour occurring, termed behaviour potential,
is predicted by our expectancy and reinforcement
value.
Rotter termed our generalised expectancies in new
situations as locus of control.
Mischel began his major work by criticising traditional
trait and state approaches to measuring personality,
claiming that not enough attention was paid
to situational factors.
Mischel and his colleagues carried out extensive
research to examine the interactions between personality
dispositions and situations. This work
produced behavioural signatures of personality.
Mischel and Shoda (1995) outlined and tested a
model of the Cognitive-Affective Processing System
(CAPS). Individual differences in this system result
from differential accessibility of CAUs and differences
in their interrelationships.
Evaluative criteria are applied to all the theories covered
in this chapter.
Connecting up
In this chapter we started to introduce some cognitive ideas
that overlapped with learning theories. You will learn more
on cognitive ideas of personality in the next chapter in this
book.
In this chapter we emphasised how behaviours are
learnt. Chapter 18 demonstrates a further consideration of
learning theories when we outline the theory of learned
optimism.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 98
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 99
Critical thinking
Discussion questions
A child is having a problem at school. When it comes to
taking her turn in class, rather than raising her hand
and waiting to speak, she shouts out answers to questions
and interrupts other students when they are talking.
What would a behaviourist recommend to solve
this problem?
A friend of yours is having a problem with their university
work. They are unmotivated, fail to turn up to lectures,
and leave doing their coursework to the last
minute, often leading to them getting very low marks.
What would a behaviourist recommend to solve this
problem?
How adequately do learning theorists explain human
motivation?
Skinner argues that humans do not have free will. Critically
discuss.
How necessary is external reinforcement for behaviour?
Can you think of examples where either positive
or negative reinforcement is ineffectual?
Discuss the contribution that learning theories have
made to the treatment of mental illness.
‘Behavioural treatments need to be applied within an
ethical framework’. Critically discuss.
Discuss whether the concept of personality is necessary.
Has the person–situation debate been adequately
resolved?
Earlier in this chapter, we asked you to make some ethical
reflections on the Little Albert study. Nowadays
carrying out such research on your own children would
be likely to bring you into conflict with social services.
It could be construed as child abuse. While we have
gained useful knowledge from this study, it would be
very unlikely to get ethical approval currently. Would
this be a loss to psychology? You may want to reflect on
current ethical codes and consider whether we have
become too stringent to the detriment of scientific
knowledge (for more, see online Chapter 28). The situation
may be straightforward when children are involved,
but how about if it were adults old enough to
freely consent?
Are learning theories only about forms of reward and
punishment?
Essay questions
Critically examine the contribution made by one of the
following psychologists to our understanding of
personality:
• Skinner
• Bandura
• Mischel.
Compare the differences between classical and operant
conditioning.
Critically discuss how Skinner’s and Watson’s behaviourisms
differ.
‘Personality is no more than the sum of our learning experiences.’
Discuss in reference to learning theories.
How adequately does learning theory explain the development
of personality?
Discuss the concept of reward and punishment within
learning theory.
Critically compare two of the following three theorists:
• Skinner
• Bandura
• Mischel.
Going further
Books
Skinner, B. F. (1971). Beyond freedom and dignity.
London: Prentice Hall. This is a controversial book that
sparks debate. Skinner argues in this book that human
beings do not have free will, so there is no real concept
of choice in human behaviour.
Nye, R. (1992). The legacy of B. F. Skinner. Concepts
and perspectives, controversies and misunderstandings.
Belmont, CA: Brooks/Cole. This book provides an
excellent, fair evaluation of Skinner’s work.
Skinner, B. F. (1978). Reflections on behaviourism and
society. Englewood Cliffs, NJ: Prentice Hall. In this book,
M04_MALT2908_02_SE_C04.QXD 10/23/09 10:30 AM Page 99
100 PART 1 PERSONALITY
Skinner discusses the development of his ideas and their
application in the real world.
The classic source for Dollard and Miller is Personality
and psychotherapy: An analysis in terms of learning,
thinking, and culture (1950, McGraw-Hill).
Bandura, A. (1996). Self-efficacy: The exercise of control.
New York: Freeman. This is the most recent publication
by Bandura specifically on self-efficacy.
Rotter, J. B., Chance, J. and Phares, E. J. (eds) (1972).
Application of a social learning theory of personality. New
York: Holt, Rinehart & Winston. This is the book where
Rotter, in collaboration with Chance and Phares,
described a general theory of social learning.
Lefcourt, H. M. (1981, 1983, 1984) has edited three
volumes of the early work on locus of control. These
volumes are titled Research with the locus of control construct
(New York: Academic Press).
Journals
Bandura, A. (1974). Behaviour theory and models of
man. American Psychologist, 29, 859–869. In this article,
Bandura discusses his concept of personality. Published
by the American Psychological Association and available
online via PsycARTICLES.
Bandura,A. (1977). Self-efficacy: Toward a unifying theory
of behavioural change. Psychological Review, 84,
191–215. This is the classic account of self-efficacy. Published
by the American Psychological Association and
available online via PsycARTICLES.
Bandura, A. (2002). Swimming against the mainstream:
The early years from chilly tributary to transformative
mainstream. Behaviour Research and Therapy, 42,
613–630. This is Bandura’s account of the development
of his ideas and their application in therapeutic
contexts. Behaviour Research and Therapy, an international
multidisciplinary journal, is published by Oxford
Elsevier and is available online via Science Direct.
If you are interested in examining applications of locus
of control in health, Journal of Health Psychology (2005),
Vol. 10, No. 5, is a special edition devoted to the
measure, including a short version, and to its new applications
in health.
Mischel, W. (2004). Toward an integrative science of the
person. Annual Review of Psychology, 55, 1–22. This
article by Mischel gives an excellent overview of his
current position and the progress that personality theory
has made in addressing the person–situation controversy.
Annual Review of Psychology is published by
Annual Reviews of Palo Alto, California, and is available
online via Business Source Premier.
Overskeid, G. (2007). Looking for Skinner and finding
Freud. American Psychologist, 62, 590–595. Published by
the American Psychological Association and available
online via PsycARTICLES. In this article Overskeid suggests
that though Sigmund Freud (whose theory we outline
in Chapter 2) and B. F. Skinner are often seen as
opposing theorists, they had many things in common,
including basic assumptions shaped by positivism
and determinism. The article discusses how many of
Skinner’s views may have been influenced by Freud.
If you would like to look through some journals related
to behavioural analysis, there are:
– Journal of Applied Behavior Analysis. A journal
primarily for the original publication of experimental
research involving applications of the experimental
analysis of behaviour to problems of social
importance.
– Journal of the Experimental Analysis of Behavior. A
journal primarily for the original publication of experiments
relevant to the behavior of individual organisms;
also publish review articles and theoretical
contributions.
Web links
The Cambridge Center for Behavioral Studies is a
resource for those interested in behaviour analysis and
its role in education, health and the workplace. Also
provides a comprehensive list of links to other behaviour
analysis resources (http://www.behavior.org/).
This site has a wealth of material on self-efficacy,
including contributions from Bandura himself. It also
has an extensive reference list of research on self-efficacy
(http://www.des.emory.edu/mfp/self-efficacy.html).
This is the link to Walter Mischel’s website. Here you will
get an idea of the work he is currently undertaking and a
list of his most recent publications (http://www.columbia
.edu/cu/psychology/indiv_pages/mischel.html).
Film and literature
Nineteen Eighty-Four (1949, George Orwell, Penguin).
Primarily concerned with the prospect of state control
by behaviourist means. In Nineteen Eighty-Four, rats are
used as a means of shaping Winston’s behaviour to produce
the required response. Nineteen Eighty-Four is also
available online (http://www.online-literature .com/orwell/1984/)
and was made into a film in 1984 (directed
by Michael Radford).
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 100
CHAPTER 4 LEARNING THEORY PERSPECTIVES ON PERSONALITY 101
Explore the website accompanying this text at www.pearsoned.co.uk/maltby for further resources to help
you with your studies. These include multiple-choice questions, essay questions, weblinks and ideas for
advanced reading.
Brave New World (1932, Aldous Huxley, Penguin). Also
a book primarily concerned with the prospect of state
control by behaviourist means.
Two films which depict the principles of learning,
reward, and positive and negative reinforcement very
strongly are The Village (2004, directed by M. Night
Shyamalan) and The Island (2005, directed by Michael
Bay). We have not gone into too much detail of these
films so as not to give the plots away.
Token Economy: Behaviourism Applied (Educational
Resource Film; McGraw-Hill, 1972). Outlines B. F.
Skinner’s ideas on behaviourism and rewards. Concord
Video and Film Council.
Classical and Operant Conditioning (Educational
Resource Film, 1996). The work of Ivan Pavlov and B. F.
Skinner is outlined. The two types of conditioning are
illustrated, including examples of historical laboratory
work and Skinner boxes. Uniview WorldWide.
Discovering Psychology video (Educational Resource
Film, 1990). The theory of self-efficacy. WGBH/Annen-
berg-PCB-Project/CS.
M04_MALT2908_02_SE_C04.QXD 10/21/09 12:03 PM Page 101