Book Item 61140
Book Item 61140
Learning Objectives
x describe how psychologists have sought to define human personality and how it can
be distinguished from other attributes;
x distinguish between the different approaches relating to assessment of individual dif-
ferences in personality;
x describe and evaluate the different theories;
x evaluate critically the use of personality questionnaires, the issues involved and their
limitations.
x Fearfulness.
x Irritability and frustration.
x
Ǥ
x Activity level.
x Attentional persistence (concentration level).
These seem to be related to the major dimensions of adult personality. The tem-
perament of an infant has profound effects on a variety of important developmental
processes, such as learning and relationships with others. Socialization also pre-
sents another source of personality difference, for example through how any one
child is treated within the family group compared to other siblings, and there is the
Ǥ
source of personality change. Lastly, we cannot discount the influence of television,
technology and other media upon behavioural differences.
There are clearly some personality attributes which appear to be more genetic
ϐ
ǡ
and some others which could be the result of an interaction between the two. As
ǡϐ
ϐ
Ǥ
the factors which affect variability in test scores. In addition, factors which affect
variability in the interpretations given to those scores could include some form of
test bias or misunderstanding of the cultural backgrounds of individuals.
We can see, therefore, that the factors of socialization, social environment, fam-
ily (including differences in treatment of different members) and educational dif-
ferences will all have an impact on personality. To these are added other factors
such as race, ethnicity, culture, age and gender. The term ‘race’ relates to the major
division of humans into those having distinct physical characteristics through
Ǥ
ǡǡ
groups, each having a common tradition and origin. Culture is linked to customs,
civilization and group achievements. And lastly, the term ‘gender’ relates to sexual-
ǡǤǤ
ϐ
Ǥ-
ǡ
ϐǤ
There are also trait and behavioural differences between people, and there can be
a gradual change in these over the lifespan. A greater change in personality may
sometimes occur as a result of a traumatic experience, although the environment
ϐ
ϐ
Ǥ
factors which also impact upon different race, ethnicity, culture, age and gender
groups concern socialization (e.g. the family unit, education, television, and other
forms of technology). All of these can create behavioural differences between peo-
ple. Another aspect of interest here is that groups may appear distinct sometimes
because of expressed behavioural differences rather than actual trait differences.
Figure 8.1 The situational view (A) sees personality as being inconsistent, while the
dispositional approach (B) sees it as being consistent and unchanging on a trait
continuum
Ǥ
highly structured situations where behaviour is constrained, people will still dem-
onstrate some aspect of their personality, even if only in a cautious way such as a
sly nod or ‘wink’.
Studies show people do behave consistently across different events, with most
behavioural correlations being above 0.7 and others above 0.3 (Small et al., 1983).
A number of characteristics have also been shown to remain largely invariable after
the age of 30. So the evidence is that personality does exist. This counter-argument
is sometimes called the dispositional approach, and it views personality through
consistent and unchanging dispositions, regardless of circumstances.
inside the person of psychophysical systems that determine the person’s character-
ǯǤϐǤ
There has been a multitude of other attempts as well, and the most common refer-
ǣ
There has often been reference made to personality as arising from a combination
of relatively enduring dimensions of individual differences, which means that other
characteristics, such as intelligence and cognitive abilities, motives, values and atti-
tudes, also have to be considered (Carver & Scheier, 2000). The reason for this is
Ǥ
thus becomes a unique combination of cognitive and affective characteristics which
make up a relatively consistent pattern of behaviour. However they choose to define
it, theorists will generally say that any significant behaviour should distinguish one
person from others and be consistent over situations and time. Most psychologists
ǣ
The characteristic, stable patterns of behaviour and modes of thinking and feeling
that determine a person’s unique way of adjusting to the environment.
The only problem is that we might include within this definition some other aspects
which we want to assess. So how do we distinguish between personality and other
attributes, such as attitudes, values and motivation?
ʹǤǯϐ
ǣǮǡ-
ǡ
ϐ
ǯ
all objects and situations with which it is related’ (Allport, 1937, 1961). So it’s a
learned disposition to respond positively or negatively to any person, object, issue
or event. Some psychologists have viewed this in terms of a tendency to evaluate a
ǡ
ϐ
thoughts, feelings or behaviour (Ajzen & Fishbein, 2000). The common view is that
our attitudes provide a predisposition to behave in a particular way.
Interests
These are thought to be a subset of a person’s attitudes relating to the evaluation
of personal beliefs. If we have a particular belief and feel positive about it, then we
are more likely to want to do more or think more about that belief. Therefore we
will tend to show more interest in it.
Values
Values relate to the usefulness, importance or worth we attach to either activities
or objects and to how people should behave, and can be linked to our interests and
Ǥ ϐ ǡ
ǡ
ȋ
ǡͳͻȌǤ
be ‘happiness’, ‘a healthy body’, ‘fame’, ‘security’, or ‘a loving family’. We are thought
to strive towards our values and to use them as a means of judging our actions and
those of others. It is generally also thought that our values are developed through-
out childhood.
Motivation
Motivation is a characteristic made up of our needs, interests and aspirations. This
focuses upon what drives us to do some things but not others (i.e. our driving force),
its direction and our persistence. A number of theories have been put forward to
explain this, including instincts, drive theories, arousal theories and hedonistic (i.e.
pleasure-seeking) theories (see Chapter 9 for more).
Beliefs
These are the expression, internal or external, of our feelings or thoughts about
something.
Ability
This refers to an underlying capacity to be able to do something or to behave in a
particular way.
On this basis, we can now distinguish between personality and other attributes of a
person. Attitudes, values and ability are more likely to have longer-term stability,
whilst interests, beliefs and motivation could be more transient. Motivation, oddly
enough, could be a state or a trait, given that some people are always highly motivated
whatever they are doing (and therefore, a trait), compared to other people who might
ϐ
other activities (a state). Whenever you read a novel you may well be able to apply
these kinds of attributes to different characters.
SUMMARY
Despite the views of ‘situationalist’ psychologists, who see personality as being ever-
changing, researchers in psychometrics have taken a dispositional approach which
views it as being made up of relatively consistent trait-based internal characteristics.
We have discussed the difference between implicit and explicit theories of person-
ϐ
Ǥ
ϐǡǡ
-
mon aspects refer to patterns of behaviour, thoughts and feelings. It has also been
distinguished from other attributes such as attitudes, values and motivation. As with
ability, personality seems to be founded upon a complex interaction between genetic
and environmental factors, as well as with race, ethnicity, culture, age and gender.
THEORIES OF PERSONALITY
There have been a number of contrasting approaches to understanding this elusive
concept. Not all theories embrace the same subject matter, viewing it instead in
widely disparate ways, and this has had a big impact on assessment measures.
Ǥ
ϐϐ
-
vidual. Unsurprisingly, therefore, different ways of modelling and theorizing about
it have emerged. No single theory encompasses everything. Methods of investigation
have also varied widely, too.
Another important reason for the differences relates to how individuals or
schools of thought have been limited by knowledge and understanding at the time
of their thinking. Therefore, theories have their origins in different paradigms (i.e.
ϐ
time). There is a common aim, however, to develop a model providing a systematic
account of the unique personality structure shared by people. Through this, theo-
rists hope to generate a way of understanding individuals and individual differences.
Some theories have a greater focus on assessment than others, for example psycho-
metric and type approaches have used questionnaires, while social and behavioural
ones tend to use rating scales.
ǡ
ϐ
Ǥ
ϐ
ǡ
make use of personality ‘types’ or not, or whether they link personality to biology
therefore, that the positions of individuals on each dimension are given relative to
their position on every other dimension rather than comparing the scores to norma-
tive data. In Chapter 3 we described this approach as being self-referenced and used
the label ‘ipsative’. The process of making questionnaires this way is often called
‘ipsativity’. As we concluded there, questionnaires having these characteristics are
best reserved for discussion activities rather than for selection decisions.
Traits also have their limitations. For one thing, as with the type approach, there
Ǥ
ͳͻ͵ϐͳͺǡͲͲͲǤ
use of factor analysis and their separation into different forms have helped to reduce
this number considerably. They are more straightforward than types, and are capa-
ble of measurement. They form a normally distributed continuum on a scale, having
Ǥϐ
distributed like this. In general, we identify traits by observing individuals behaving
consistently in response to a variety of environmental conditions. In some instances
traits are interpreted more narrowly as representing biological characteristics, as
ǣ
Ǧ
person consistencies of behaviour.
ǡ
ϐ
-
istics (Cattell, 1973). They are assumed to be generally stable over the lifespan,
especially after reaching adulthood. A trait is, therefore, any persistent characteris-
ǡǡ
ǡ
ϐ
-
ality is demonstrated. Being essentially abstractions, as they do not exist in the real
ǡ
ϐ
-
son which elicits a response. Traits provide some stability and are non-situational
(i.e. the person has these characteristics in all situations although they may be
ϐȌǤ
ǣ-
ple neuroticism, isolated as a fundamental and unique trait, includes behaviours
and thoughts associated with guilt, low self-esteem, depression, phobia, anxiety and
psychosomatic illnesses. Other examples might be how outgoing, assertive or caring
Ǥ
ϐ
traits. (The term ‘trait’, by the way, is derived from the Latin word traho, meaning ‘I
ǯǮǯǤ
ǡϐǮǯ
rather than silent.)
So the key distinction between types and traits is that where the type approach
puts people into discrete categories, the trait viewpoint considers each character-
istic as a continuum and describes personality in terms of where the person is
placed on a number of continuous scales. On this basis, someone may be near the
centre of a scale of intelligence, towards the low end of a scale measuring anxiety,
towards the high end of a scale measuring dominance, and so on until an overall
picture is gained.
We should also consider the distinction between traits and competencies used
in organizations to evaluate job performance and individual development.
Competencies are the set of behaviour patterns which someone needs to bring to a
Person 1 Person 2
Type 1 Type 2
Person 1 Person 2
Trait
Scale
Figure 8.2 Types distinguish between people using categories, while traits
distinguish them using scores on scales
mental processes as possible (Magnusson & Torestad, 1993). You might think the
differences between people outweigh the similarities. On this basis, data collection
will involve qualitative methods such as interviews and observations. Assessments
might include the Rorschach Ink-Blot or the Repertory Grid technique. This approach
ϐ
Ǥ
It is referred to as an idiographic approach, derived from the ancient Greek word
idios, meaning ‘private’ or ‘personal’. Researchers who use it prefer to investigate
people on an individual basis only, for example in a therapeutic environment. Freud
used it to create his theory. The idiographic paradigm is therefore based on an
assumption that people have unique characteristics and cannot be described in
identical terms. Although it provides a richer understanding of individuals, it is dif-
ϐ
ϐǤ
An alternative method would be to identify some attribute which you think is an
important aspect of personality and measure this within a large group, possibly
using standardized questionnaires. By doing so you can identify individual differ-
ences in the extent of a particular trait or set of traits. The focus this time is on
similarities, with the view that each person can be represented in terms of different
degrees of the same thing. This is called the nomothetic approach, the term being
derived again from an ancient Greek word for ‘law’. It assumes the existence of a
ϐ
described, explained and predicted. Trait-based theories do this, and the psycho-
ȋͺǤͳȌǤϐ
they enable predictions to be made of behaviour.
Despite this distinction, psychologists using the nomothetic approach have argued
that they do still accept and can work with uniqueness (Carver & Scheier, 2000). The
uniqueness comes from a person’s particular combination of personality variables,
the degrees to which they exhibit them, and their interactions. Combinations and
Idiographic Nomothetic
interactions make each one unique. Those working within the idiographic paradigm
ϐ
ǡ
ǡ
predictions. So the differences may not be so clear-cut.
We can now review the principal theories of personality and its assessment. The
literature abounds with different taxonomies and labels, so I have opted to use the
term ‘paradigm’ as this enables us to group together theories which relate to similar
patterns of thinking. Some theories will not be accommodated easily into just one
paradigm and may be placed in others as well. We will consider assessment issues,
especially reliability and validity.
Classical typology
A classical typology comes from our old friends, the Greeks, who theorized about
ǮǯǤ
ϐ
attempt at a formal theory to account for differences between people around 400
ǡ
Ǥ
suggested that temperament was determined by relative amounts of certain bodily
ϐǡ Ǥ
ϐǡ
ǡ
phlegm and mucous, as indicated in Table 8.2. The terms he gave to these types
(sanguine, melancholic, choleric and phlegmatic) are still widely used today and
Type label
Predominant body fluid characteristics Temperament
ϐ
Ǥ ǯ
combine in any way and they were thought to be inherited. They have been linked
with the effects of endocrine activity on temperament.
The problem with this approach is that it assumes people have to slot into one
of the categories rather than being made up of mixtures or combinations. It’s per-
fectly possible in real life for many people to demonstrate something of more than
one, possibly of three types, for example someone who is depressed, apathetic and
easy to anger. This is often a problem where people are ‘dropped’ into boxes.
Constitutional typology
In the twentieth century other theorists sought to link personality with individual
physique, an approach described as presenting a constitutional typology. Kretschmer
(1925) began by stating, incorrectly as it happened, that schizophrenia could be
associated with tall thin people, while short fat people were more prone to manic-
depressive psychoses. An alternative was devised by W. H. Sheldon in the 1940s. He
ϐǡ
expressed in relation to extreme types. His somatotypes were based on ‘endomor-
phy’, ‘mesomorphy’ and ‘ectomorphy’ components after he had studied thousands
ȋǡͳͻͲȌǤ
system and level of fatness observed, mesomorphy to the muscles and amount of
musculature, and ectomorphy to the nervous system and brain and thus the body’s
leanness or fragility. These resulted in three somatotypes known as endomorphs,
mesomorphs and ectomorphs, whose differing types of physique are described in
Table 8.3. Sheldon used correlational studies to show that each type could be asso-
ciated with temperament. Accumulating data on 45,000 participants, he also con-
structed tables of male body types using a grading scale to match individuals against
the extremes.
Sheldon’s work gave rise to debate and other studies of the relationship between
personality and biology. He reported strong statistical relationships between body
structure and personality, although he did not take account of measurement error.
ϐ
Ǥ
ϐ-
cant role in the development of more humane treatment of mental patients, there
has been much debate about its therapeutic value. From an assessment point of
ǡ
ϐ
Ǥ ϐ
design objective experiments to test Freud’s hypotheses, which were based on his
interpretations through observations of individuals. His theory is also not conducive
to prediction because different behaviours could be indicators of the same underly-
ing impulse. His views focused much upon the role of the unconscious and his own
rational views of others, though less upon the importance of individual learning or
ϐ
Ǥ
ǡ ϐ
development of projective assessments. (Chapter 9 includes a more detailed discus-
sion of these.)
bimodal (i.e. having two score distributions centred upon the extremes; see
Chapter 4) although this is not the case with the MBTI. The picture is even more
ȋ
ƬǡͳͻͶǢǡͳͻȌǤ
is based upon correlational studies (see Chapter 6) yet the MBTI is partially ipsa-
tive, and therefore these are difficult to do. So there is inadequate evidence for its
validity. Its use is best confined to discussion activities such as development, team,
coaching or counselling activities (Kline, 2000).
Self-actualization
Esteem needs
Identity needs
Safety needs
Physiological needs
events differently. The rep grid focuses upon ‘elements’ in the life of a person, which
could be the self, other people, objects and even events. An example of its use is as
Ǥǣ
elements and what are their relationships?
ϐǮǦ
of a person in relation to a set of statements’. A person may list or be presented with
a list of people, whose names are given in groups of three, and the individual is asked
to indicate how two of the three are similar and the third is different from the oth-
ers. For example, someone might say that two are generous whilst the third is mean.
ǡ
Ǥ
This is repeated for different groups of three people from the list, enabling the
constructs used to organize information to emerge.
In many instances constructs are represented by two terms or phrases which
have opposing meanings, such as decisive/indecisive or friendly/unfriendly.
than physical or other descriptions. Gradually, a matrix can be built up through this
process, having the elements across the top and the constructs elicited listed down
ǡͺǤͶǤϐ
ȋ-
ally about 12), the participant can be asked to rate each element on a scale, such as
ͳ ͷ ͳ ǡ
Ǥ
which dimensions are important.
Although there is no standard method for scoring, a wide range of analyses is
possible, ranging from a simple evaluation of the similarities to factor analysis.
Analysis can focus on the number of constructs elicited, their nature, which attrib-
utes of others are most emphasized, and any differences between elements. High
correlations between two or more constructs may indicate an underlying core com-
ponent. It has been suggested, therefore, that the method can assess the cognitive
complexity of individuals (Bieri, 1955).
Kelly saw the therapist’s role as helping people to become more aware of faulty
Ǥϐ
ϐ-
ences between the personal construct systems of patients suffering from schizo-
phrenia and those experiencing depression, neuroses and mild organic disorders,
and with a healthy control sample. This technique has been adapted for other appli-
cations, for example, becoming an objective method for conducting job analysis in
organizations.
The method can provide a useful awareness of the ways in which people perceive
their world, how they organize attitudes and beliefs, how emotional responses are
ϐ
Ǥ
-
ness whilst also viewing everyone within the same conceptual framework, and is
adaptable to situations. For these reasons it has been popular with therapists, coun-
sellors, psychologists and Human Resources professionals. A good introduction to
its use is provided by Fransella (2003).
ǯ ϐ
person’s constructs and perceptions. But the process of undertaking this however
Generous 3 1 3 4 2 Mean
Decisive 4 2 2 3 1 Indecisive
Friendly 5 1 5 4 3 Unfriendly
Intelligent 3 3 5 3 3 Dim
Gentle 1 5 2 1 5 Violent
Caring 4 1 3 3 5 Uncaring
etc. etc.
is complex, especially if factor analysis is used, demanding expertise and being time-
consuming. It tends also to focus too much on thought processes at the expense of
other aspects of personality and could be seen as mechanistic. Its effective use
depends on expertise in interpreting outcomes and no systematic interpretation is
provided, introducing some subjectivity. There may also be the problem of whether
participant and assessor have a mutual understanding of the constructs elicited.
The participant might seek to sabotage or distort the process by making inappropri-
ate or inadequate disclosures, and this can be controlled by establishing a good
ǡ
ϐ
ǡ
constructs obtained. Any bias in the approach used might be controlled by support-
ing a good understanding and evaluation of the criteria involved.
Being idiographic in nature, the repertory grid cannot enable any standardiza-
tion process or identify individual differences. A new version needs to be con-
structed each time it is employed, meaning that the reliability and validity will have
to be re-checked. It seems to view personality as stable and based upon individuals’
ϐ
Ǥ ǡ
-
ǣǡ
-
ing personal information.
Eysenck’s theory
Ǧ
ȋͳͻǡͳͻͻͲǡͳͻͻͶȌǡ
ϐ
Ǥ
ǡ
ǡ
Ǥϐ
keep us active and awake whilst the other is focused on inactivity, and we must try
to maintain a balance. Sitting at the base of the brain is the ascending reticular
activating system (ARAS) which controls the arousal level of the cortex above, effec-
tively acting as a kind of ‘dimmer switch’.
ǯ
different levels, with the introvert’s cortex being more aroused. Most people prefer
a moderate arousal level and any very high or low level is perceived as unpleasant.
Therefore, in situations where external stimulation is present introverts will expe-
rience greater arousal and try to escape from it. As a result they will need more
effort to adapt, whilst extraverts, having a need for more arousal, will be more com-
fortable. Because of their natural arousability levels, introverts try to avoid intense
stimuli while extraverts seek them out.
limbic system which, he said, accounted for individual differences in neuroticism.
ϐ
ǡǡ
ȋ
-
ing to experience more extreme emotional responses). This was associated with
the arousability of the limbic system which is connected in turn to the autonomic
ȋȌǣ
muscles, heart rate and sweat glands. The ANS is also known as the peripheral
Ǥ
this, the sympathetic nervous system (SNS), appears to be over-active in anxious
ǡ
ȋȌ
and is calming, being stronger in more unemotional people. Neurotic individuals
possess a hyper-arousable limbic system and are more likely to experience emo-
Ǥ
might be linked to a chemical messenger in the brain.
Gray’s theory
ǡ
Ǥ
reinforcement sensitivity theory (Gray, 1970, 1981, 1987) suggests that personality
ǣ
Approach System (BAS) and the Behavioural Inhibition System (BIS). The BAS cen-
tres on motivations to approach the environment, causing people to be sensitive to
rewards and to look for them, while the BIS focuses on motivations to avoid, causing
people to be sensitive towards dangers. These were associated with the character-
istics of impulsivity and anxiety, with individuals high on approach being more
impulsive and those high on inhibition more anxious. A questionnaire based on the
two scales was developed in the USA.
Cloninger’s theory
Ǥǡ
ϐ
psychological, social and medical sciences. His theory is related to seven personal-
ity domains, including ‘temperament’ domains of novelty-seeking, harm avoidance,
reward dependence and persistence, and ‘character’ domains of self-directedness,
cooperativeness and self-transcendence (Cloninger, 1987; Cloninger et al., 1993).
Self-directedness relates to a person’s level of autonomy, cooperativeness to links
with society, and self-transcendence to beliefs about mystical experiences. Cloninger
linked these to neurotransmitters in the brain, as well as to learning through rewards
Ǥ
ǡ
ϐ
ǯǤ
The evidence for these biological theories is inconsistent and contradictory
despite much research (Matthews & Gilliland, 1999) and evidence linking physiology
Ǥϐ
ǯ
terms of arousal, as well as for the approach/inhibition systems of Gray, although
others have not shown support for these and suggest a weak relationship. A link
between neuroticism and arousal is not generally supported. However, it seems that
characteristics are associated with physiological activities such as heart rate, skin
conductance and brain activity. Given that the brain is a complex organism with sub-
ǡ
ϐ
-
tion and that some combination is needed to establish a connection with personality.
The theories have contributed to assessment, including the questionnaire
ǯ
ǯ Ǧ
Questionnaire (revised as the Temperament and Character Inventory), based on
his biological factors and used to assess personality disorder. The Cloninger inven-
ǯǤǡǡ-
ǯ
factor-analytic studies of personality and his development of a psychometric trait-
based measure. There have thus been attempts to relate biology to the empirical
measurement of personality, though it is not clear whether there are any links to
developmental or stable characteristics, genetic or environmental factors, or the
importance of learning.
ǯǣǯ
what the parts of personality might be, not about the aspects of a typical theory
such as how it develops and whether it can be changed. All we have developed is a
system for specifying the components and measuring them, and this provides a
means of describing what any personality is like but does not explain how it devel-
oped. For this reason some people will say we are just labelling and quantifying
personality. Maybe that’s so, but then along any dimension there are different places
for different people, and when we look at one individual’s positions on a number of
dimensions we can combine them in a unique description. Having a normal distribu-
tion enables us to do parametric statistics, and because of this the discovery of traits
has made a major contribution to the study of personality.
That recurring technique called factor analysis has enabled researchers to identify
the underlying basic dimensions or ‘factors’ of many trait labels. A few factors can be
ǣǡ
ǡ
Ǧ
and good-natured might be found under the heading of ‘agreeableness’. This is some-
times called the ‘lexical hypothesis’, going back to Galton (1884), and the approach was
ǡ
Ǥ
‘lexical’ suggests that this approach was founded upon the words which are used in
Ǥ
Ǧ
ǡ
ȋͳͻȌ
ǣ
x ȂǤ
x Neuroticism – Stability.
x
ȂǤ
These major dimensions resulted from correlations between groups of traits. For
example, extraversion derives from traits including sensation-seeking, assertiveness,
activity, liveliness, sociability and others all found to correlate with each other. His
ǡ ǡ ǡ Ǥ
described his third dimension as representing the degree to which a person is tough-
minded and called it ‘psychoticism’, following observations that people suffering from
a psychosis score highly on this. They are impulsive, impersonal, egocentric, cold,
aggressive, anti-social, and lack empathy. They are also creative, which according to
Ǥǡ
ʹͳǤ
ǯ
ǣȋȌǡ
-
ȋȌǡ
ȋȌǤ
ǡǡ
ȋȌǤ
Cattell’s original interest lay in determining whether the unitary patterns of per-
sonality discussed by clinical psychologists could be measured objectively. He
checked out their models by obtaining extensive empirical databases and by devel-
ϐ
ǤϐͳǮ
ǯǡ
ǯϐȋǤǡͳͻͲȌǤ
In order to do this he collected some 17,953 trait names used in everyday lan-
guage to describe behaviour before eliminating synonyms (words or phrases that
mean exactly or nearly the same as other words or phrases), and combined the data
with information from other sources such as observations of behaviour.
Questionnaires were also used to explore the perceptions of individuals. Lastly,
objective data were obtained from observations of what people did, rather than
what they said, in pre-planned structured situations. All of the data were collated
and subjected to correlation and the techniques of factor analysis.
ǡͶͲȂͷͲ
which were factor analysed down to 12 factors. A questionnaire was developed and
the outcomes were once again analysed, leading to four more dimensions, making a
total of 16 including the addition of a general reasoning scale. These form the basis of
ͳ
ȋͳ ȌǤ
ǡ
although they are source traits and represent enduring aspects of behaviour, account-
ing for the variation in more observable surface traits (Cattell, 1950). What we see on
a daily basis are the surface ones corresponding to common generalizations and these
are less stable. Source traits are fundamental units which govern behaviour whilst
interacting together and with other characteristics. When analysis was conducted again
Ǧ
ǡ
ǯǤ
ͳ
also equivalent questionnaires which will measure much the same factors, such as
ǯͳͷ
ȋͳͷ ͳͷ ΪȌǤ
ǡ-
tories appear to measure more of the original surface traits, such as the Occupational
ȋȌǤ
ͳͷ Ϊ
-
ond-order factors are shown in Table 8.5. Looking at these you might ask why are there
no D, K or J factors and why there are Q1 to Q4 factors. The reason for this is that some
ȋǤǤǡ
Ȍͳ
were removed and that Cattell’s original letter labels remained unchanged. Four others
were found in questionnaires only and not in the analysis of language and are distin-
guished as factors Q1 and Q4. There have been debates about Cattell’s work and its
outcomes, and arguments over scale reliabilities, although these have been improved.
ͳͷ Ϊǯ
ǡ
Table 8.5 Descriptions of the 16 source traits and four second-order factors of the 15FQ+
E Introversion Extraversion
N Low aNxiety High aNxiety
O Pragmatism Openness
A Independence Agreeableness
C Low Self-Control High Self-Control
having poor reliability, has been replaced by that of ‘intellectance’, a measure of per-
sonal preference for complex thinking. The letters designated for the second-order
factors relate to the ‘Big Five’ scales which we shall discuss soon.
Development centres provide one of the most useful tools for accurately identifying the
gaps between an individual’s abilities and those needed for jobs. They give information
on individual abilities and provide a mechanism for empowering people to develop them-
selves and to improve organizational performance. Psychometric assessments can make a
major contribution to outcomes and are most effectively used as one element in a range of
procedures (Coaley & Hogg, 1994; Lee & Beard, 1994).
x Ǣ
x Ǣ
x Agreeableness;
x Conscientiousness;
x Openness to experience.
As a result it should be possible to summarize any profile on the basis of these fac-
tors alone. Many psychologists would now agree that when the data are summarized
ǡ
ǯ
Ǯ ǯǤ
ǯ
Ǥ
ȋͳͻͻʹȌ
the five traits (i.e. neuroticism, extraversion, and openness to experience). The sec-
Ǧ
ǣ
ǡ ͳͷ ͳ
ǡ
ͺǤǤ
ǣ
The importance of the Big Five factor model of personality is that it provides
an organizing framework for the overall domain of personality, in the same way
as the concept of ‘g’ does for ability and intelligence. It reduces the wide complex
domain of personality to just five main components. There has been some varia-
tion in the names given by different researchers, yet the scale contents are essen-
tially similar regardless of the names given. The simplicity, conciseness and
straightforward nature of the Big Five make them appealing.
An overall view
Overall, a review of the different theories of personality, from typology to traits,
would suggest that all of the various approaches to personality can offer insights
into and ways of understanding its complexity, although none of these provides an
Ǥ
-
ϐ
and cannot predict behaviour adequately. Type approaches are too generalist and
tend to over-simplify the nature of personality. However trait, biological and social
ϐ
ǡ
of being evaluated in an empirical way. On the one hand, social learning theory has
tended to focus upon behaviour but neglect other aspects of the person, whilst the
trait approach enables us to describe an individual in detail although it fails to con-
Ǥ
ǡ
ǡ
ϐ
ǡ
ǡ
counselling and personal development settings. The trait approach tends to be most
ϐ
‘what’ rather than the ‘why’ of personality.
No single theory is capable of providing an over-arching and comprehensive
description of personality, probably because we are trying to understand the most
ǣǤ
the various areas of human activity, although we should always remember the dan-
Ǥ
ǣ
-
ments of others. This means there will always be a need for evidence of both good
reliability and validity.
Table 8.6 Matching 15FQ and 16PF global factors with the Big Five
SUMMARY
Theory and practice are necessarily linked. We have reviewed a range of personal-
ity theories and their relationships with assessment methods, from classical typol-
Ǥ
Self-report data
In completing questionnaires people will, of course, answer questions about them-
selves. The accuracy of the data gained will therefore depend on personal insight. This
ǣǯǡ
ǡ
be sure. Interpretation should be conducted with caution and include discussion with
Ǥǡϐǡ
considered susceptible to sabotage and distortion by test-takers, especially in selec-
Ǥ
ǣ
Variability in scores can be influenced in two ways. Firstly, the variability can be
introduced by factors which influence responses, for example as sabotage or distor-
tion or a lack of understanding of the language used. Secondly, there can be vari-
ability caused by how the scores are subsequently interpreted by the person giving
the questionnaire to others as a result of a misunderstanding or misrepresentation
of the outcomes, for example through a lack of training in how to understand a
personality profile and give feedback. This can sometimes occur when a feedback-
giver makes statements which are designed to please the person who completed
the questionnaire. (This is often referred to as the ‘Barnum effect’, which will be
explained in Chapter 9.)
The construction of questionnaires is deigned to minimize such problems, but it
is also the responsibility of the administrator to neutralize such tendencies as much
as possible. It is important to develop a good rapport with people undertaking the
ϐ
they cooperate with frank and honest reports. This is clearly easier to do where it
is being used for development purposes. In recruitment, it may be helpful to state
that the questionnaire is only one aspect of the selection procedure being used and
that they will gain helpful feedback. The results can then be linked to other known
aspects of the individual. Applicants will, of course, be less happy in the job if they
gain it through suggesting that they have characteristics they don’t really possess.
Questionnaire items are often chosen to be as ‘neutral’ in value as possible, to
emphasize both desirable and undesirable aspects at both ends of trait scales. In
addition, items which are not ‘face valid’ (i.e. which do not obviously refer to the
trait but which are known to measure it through correlational research) are chosen
wherever possible as an in-built protection against sabotage, distortion or faking.
In any case, this problem of faking is probably less serious than people often
think since they are most likely to complete the questionnaire when they realize
that an accurate result will contribute best to their own welfare (i.e. during develop-
ment projects). If time is taken to ensure that people understand the importance of
careful and truthful responses, a long step towards achieving good measures will
have been taken. Not all personality questionnaires include methods to assess all
of the possible test-taking styles in various assessment situations, although they
will generally include a measure of ‘socially desirable’ responding if the question-
naire is to contribute towards some form of decision making.
Random responding
This is a particular problem when respondents have little motivation to complete a
questionnaire or have been strongly encouraged or coerced to do so. There is a risk
that they may select responses at random and be unlikely to read items. Such an
event is relatively rare in selection settings, although it may arise in forced pro-
Ǥ
ϐ
infrequency scale, consisting of items which have only one correct answer, but unlike
those used in reasoning tests the answer is obvious, for example ‘Babies generally
walk before they can crawl’. A random-responder is unlikely to notice that this is
factually incorrect. The problem is that these types of item tend to be obvious to
people who are reading carefully and can then pose a distraction. Some inventories
ϐ
a low endorsement rate.
Central tendency
Central tendency is the name given to the style of people who constantly choose the
middle option for items, having a response set Ǥ
and don’t want to reveal too much information, or they may be young people who
lack certainty about themselves. Careful wording of items can minimize the effect
and it can be measured also through the use of software (Rust, 1996; Huba, 1997).
ǡϐ-
resent the test-taker. The person may have been disinclined to reveal too much, is
lacking in personal insight, or genuinely holds moderate views and attitudes com-
pared to most other people. A good feedback interview afterwards could help deter-
ϐǤ
Measurement error
Scales used in inventories are subject to measurement error, as Chapter 5 shows.
There can be no 100% accuracy and rating scales should have evidence of good
reliability. Reputable inventories will usually have carefully researched items
Ǥ
ǡ
ǯ ǣ
should be considered as guides to how individuals might behave in many circum-
stances. Having said this, it is still a better guide than the subjective views of an
interviewer.
Personality Scales
A B C D E F G H I J K L M N O
Appraisal ns ns Ns Ns ns ns ns .53 ns ns Ns ns ns ns ns
evaluation
Qualification ns .47 Ns Ns ns ns .5 .57 .42 ns Ns ns ns ns ns
Level
Attendance ns ns Ns Ns ns ns .6 .61 ns ns Ns ns ns .44 ns
Punctuality ns ns Ns Ns ns ns Ns Ns ns .4 Ns ns ns ns ns
Customer ns ns Ns Ns ns ns Ns Ns ns ns Ns .4 ns ns ns
Ratings
Length of ns ns Ns Ns ns ns .6 .58 ns ns Ns ns ns ns ns
Experience
scales, called factor A to factor O, and that I then correlate the scores acquired from
a large sample of employees with their scores on six measures of performance. The
correlation matrix produced is shown in Table 8.7.
This matrix shows 15 x 6 (i.e. 90) correlations. Of these, 12 are given as being
ϐ
ǡǮǯ
ϐ
Ǥ
publisher might, on this basis, state that there is good evidence for the validity of
factors G and H, whilst there is some support, albeit lower, for factors B, I, J, L and N.
But there is a problem here arising from probability theory which says that if lists of
random numbers are continually recorded one or more of them will eventually cor-
relate reasonably well with some of the criteria, suggesting that some of the random
numbers are valid. If the sample size is small a set of random numbers giving a high
correlation will rapidly become clear. In contrast, if the sample size is large then it
will take a longer time before a set of higher correlations becomes available.
Therefore, ‘spurious’ correlations can occur purely by chance depending on the sam-
ȋǮǯϐ
ϐ
ȌǤ
ϐ
into account in any research process and deal with it. There are
corrections which can be applied. But this is often not consid-
ered in the case of validation studies having large matrices. In
ϐ
-
relation occurring by chance is less than 5% or 1 in 20. If a
matrix of 20 correlations is constructed using random numbers,
ϐ
purely by chance. When the matrix has 100 correlations, by the
ϐ
ϐ
Ǥ
ϐ
ǡ
-
ǯ
ǤǨ
of warning for anyone buying a personality inventory.
ͻͲ
ϐ
ǡͳʹϐ
Ǥ
ʹͲ
ϐ
ϐ
ǡͶǤͷΨ
among the total. We can’t have half a correlation, so we will round this up to 5. The
ϐ
ϐ
are potentially genuine, and this is the case. Obviously, the spurious ones occur at
random and therefore don’t appear as a pattern. In our example the factors which
appear most valid would be H and G, although it is worth saying that the matrix is
ϐϐ
ϐ
-
mally be generated. A statistical correction can eliminate the effect.
SUMMARY
The use of self-report questionnaires to assess personality characteristics continues
to grow. Issues concerning their use, such as those relating to response styles,
including sabotage, distortion, central and random responding, as well as measure-
ment error, have been largely met by developments in technology and the applica-
tion of codes of good practice for their administration, data management and
feedback interviews to individuals.
x the distinction between subjective and objective models, and contrasting views
relating to the nature of personality;
x ϐ
Ǣ
x how thinkers and scientists have evolved different models and processes, some
of which have enabled the assessment of individual differences;
x how personality inventories have been constructed, why they are useful, and
ϐǡǤ
How would you distinguish between personality, attitudes, interests, values, motivation,
beliefs and ability?
Is personality simply a result of genetics? What do you think?
Compare the different theoretical approaches to personality. How useful are these?
How would you explain Jungian type theory and its uses?
Why do you think the trait or psychometric approach is widely used for the assessment of
personality?
Explain what is meant by the ‘Big Five’. Why is this thought to be so important?
What are some of the problems relating to the use of personality questionnaires?