What are the salient personality trait models and how do they contribute to the understanding of individual differences in personality and psychopathology?

Personality, like many areas in psychology, has been discussed as far back as the ancient Greeks.  Most notably, it was Hippocrates who devised the very first theory of personality by assuming that fluids within the body were responsible for differences in behaviour; for example, levels of phlegm indicated degrees of calmness.

This theory, although incomplete, has provided the foundations of modern research into personality. Contemporary personality trait models discussed in this paper are Costa & McCrae’s (1991) Five Factor model and Eysenck’s (1992) Giant Three model. 

This paper will first define personality and personality traits before describing each model.  This description will identify each trait, and discuss how it is defined and how it was initially discovered.  It will then attempt, as far as feasible, to describe their differences and evaluate their strengths and weaknesses in a manner which analyses their contribution to understanding individual differences and psychopathology.  Fundamentally, this paper will argue in favour of the Five Factor Model, but will also contend that it is a limited model for describing the totality of personality.

What is personality? There is a general consensus that personality may be defined as the sum total of behavioural and mental characteristics that persist over time (Colman 2006). A personality trait is therefore a single dimension of personality among a finite group of dimensions.  For example, Neuroticism is a single dimension, or trait, that explains how emotional that individual is.

A personality trait model is a psychological model that attempts to amalgamate all traits into a universal system of personality.

The first personality model is Eysenck’s (1992) Giant Three model.  This model states that there are three basic traits that everyone can be classified under.  These are Extraversion (E), Neuroticism (N) and Psychoticism (P). Extraversion determines how the individual relates to the outside world; for example, whether they are outgoing or social.  Neuroticism determines emotional stability; for example, how sad or embarrassed individuals are.  Lastly, Psychoticism reflects a dimension that stretches from psychological normality to psychotic disorders such as schizophrenia (Haslam, 2007).

Much of Eysenck’s model relies heavily on a biological approach to explaining personality, with research findings indicating that genetics (Plomin 1986) and twin studies across cultures (Eaves et al 1989, Martin and Jardine 1986) indicating levels of E, N and P, have a biological foundation, implying individual differences as being partly predetermined.  If such traits are biologically imprinted in our genes, it makes sense, Eysenck thinks, to assume that they are basic dimensions of personality.

The second model of personality is The Five Factor Model (Costa and McCrae 1991). This model of personality describes five basic dimensions of personality.  These five dimensions are: Openness (O), Co-consciousness (C), Extraversion (E), Agreeableness (A) and Neuroticism (N).

Openness reflects how open the individuals are to new experiences, which also indicates how intellectually curious they are.  Conscientiousness reflects how they approach certain tasks: for example, how motivated or organized they are.  This could be a very good indication of how one might approach deadlines at work or school.  Agreeableness reflects interpersonal skills: for example, how warm and cooperative they are.  Scores here can determine their willingness to help.  Extraversion is defined in the same way that Eysenck’s model defines it, but Neuroticism is defined as emotional stability and personal adjustment (Costa and McCrae 1992): that is, how well an individual adjusts to certain situations.

Most of the traits identified by Costa and McCrae were initially found using the lexical hypothesis: that is, words in the language can identify personality traits such as kind and considerate.  Initially, 4,500 traits were found from Allport and Odbert’s work; however, with the use of factor analysis all such traits were reduced to sixteen (Cattell 1970) and then, more recently, to five.

There is constant debate as to which model most accurately identifies a trait as being basic.  The main disagreement centers on the factor P.  Eysenck claims this to be a component of C and A whereas Costa and McCrae believe P to be a combination of A and C (Costa & McCrae 1992c) as they noted that psychoticism tended to be found in individuals with low A and low C.  It is very difficult to argue that both A and C are not important factors in understanding personality and individual differences; but there is a difference of opinion as to whether researchers consider P, or A and C, to be the basic factors. Eysenck’s model appears to be economical and parsimonious and, hence, a more scientific model for describing the basic factors of personality, although P remains its weakest link (Bishop 1977; Block 1977a, 1977b).  One difficulty is perhaps due to its very obscure nature and how it relates to other factors of personality (Zuckerman 1989; Claridge 1981). Splitting P into A and C, as Costa and McCrae do, allows a wider scope to be addressed.

In saying that, however, both models can only contribute a little towards our understanding of P.  This is mainly because the DSM-IV takes a medical approach to diagnosing psychotic abnormalities such as schizophrenia or depression.  Essentially, this is because such abnormalities have an underlying biological foundation that needs to be taken into consideration.  For example, there is a growing consensus that schizophrenia is a genetic disorder, while other categories of P, such as OCD, stress and depression, are heavily influenced by dopamine and other chemical imbalances in the brain.  Such strong biological evidence indicates that personality models cannot explain such abnormalities per se. Without medical examination it is impossible to make accurate diagnoses.  But The Five Factor Model has the advantage over Eysenck’s Giant Three as it can measure psychotic traits by correlating A and C; this is something Eysenck’s model cannot do.  Doing so allows researchers to assess psychotic criteria in a variety of contexts; for example, to examine prison inmates’ attitudes and tendencies in order to review their eligibility for release, or to assess psychiatric clients as part of a general check up, or as part of an admissions procedure (Holden & Troister 2009).

In this respect the model can contribute to our understanding of Psychopathology and help us to make important life changing decisions. 

Further studies have now shown that even Personality Disorder can be predicted from the Five-Factor Model of personality in clinical samples (Reynolds & Clark 2001, Ball et al 1997, Blais 1997).  Additionally, Miller et al 2001 found evidence for personality disorder as an extreme variant of the common dimensions of the Five Factor Model of personality, a claim that was previously hypothesized by Widiger & Lynam (1998).

If we are to take these results as conclusive, the evidence suggests that the Five Factor Model, combined with other forms of testing, is rapidly becoming a more accurate model for identifying psychopathological disorders and individual differences while making positive contributions in the field.

However, there is a problem.  The Five Factor Model doesn’t tell us why people might be more psychotic than others; all they tell us is that certain factors, such as A and C, and at times, N, contribute to psychopathologic behaviour. This is why the field of biology can, potentially, provide a deeper insight into the nature of P as a whole.  But the advantage of such a model is that it allows us to see potential triggers for a variety of psychological abnormalities as well as individual differences.  In this case a Diathesis-Stress Model has been suggested and might, in the near future, be able to explain the link between biological predispositions and environmental stressors via The Five Factor Model or variants of it.

In order to assess both psychopathological and individual differences effectively, both models adopt the questionnaire as a form of self-assessment. 

Eysenck’s model adopts the Eysenck Personality Questionaire (EPQ) while Costa and McCrae adopt the Revised NEO Personality Inventory (NEO PI-R).

The EPQ includes one hundred yes/no questions and also includes a lie scale to determine discrepancies between answers.  There is evidence that this questionnaire produced reliable results for two of the three traits: E and N (Francis et al 2006).  But they found unreliable results when measuring P. Commentators have also noted this unreliability (Maltby 2007). The indication is that if such results vary there might be discrepancies in Eysenck’s definition and measurement of P, which is partly due to its interpretation; for example, Eysenck believed that geniuses and psychotics share a divergent style of thinking when trying to find solutions to problems (Eysenck 1995), whereas others, such as Maslow and Rogers disagree, maintaining that such thinking is the result of optimal health (Simonton 1994).  With such difficulties of interpretation, coupled with a string of unreliable results, the inaccuracy in measuring P in the EPQ could well detract from its contribution to our general understanding of P.

Another weakness with the EPQ, including revised editions (EPQR-S), is that by asking ‘yes’ and ‘no’ questions the individual will feel they have less scope to answer appropriately.  For example, consider this question; “ I consider myself talkative, entertaining, and the “ life and soul of a party ”.  With only ‘yes’ or ‘no’ as possible answers how can one accurately respond?  It is possible to answer ‘yes’ to the first part of the question; that is, I consider myself talkative and entertaining, but I might not consider myself “ the life and soul of a party ”.  It can be equally said about this question: “ I am miserable at times although I cannot really explain the reason for my misery ”; sometimes we do know the reason, at other times we don’t.  It is therefore difficult to say, with any degree of accuracy, how well the EPQ contributes to our understanding of individual differences as participants’ responses are limited solely to agreement or disagreement.

The NEO PI-R questionnaire significantly differs from the EPQ.  The most fundamental differences are that there are no yes/no questions and that it measures five traits rather than three. 

Furthermore, the NEO PI-R adopts a rating scale from one to five, signifying varying degrees of intensity in a participant’s feelings toward a particular question.  This provides participants with a much wider scope in answering compared to the EPQ; because of this, it is possible to analyses the extent to which participants are open or neurotic.  This allows researchers to assess levels of individual differences between participants in a more accurate way than the EPQ. 

Firstly, therefore, the NEO PI-R is a more accurate form of self-assessment than the EPQ with the capacity to increase our understanding of both individual differences and psychopathology as it allows researchers to see how factors vary between individuals.  Secondly, as more and more self-assessments are carried out by the NEO PI-R, emerging patterns can determine which factors are mostly associated with psychopathology and which with personality.  For example, a measurement of N might tell us something about particular individuals and their susceptibility to depression.  Therefore the NEO PI-R provides a better indication of individual differences in psychopathology than the EPQ does.

Indeed, although the Five Factor Model is generally accepted among psychologists, this is not to say it is a perfect theory of personality. 

Some argue the model omits potentially important factors such as honesty and humility.  Cross-cultural studies have suggested evidence for these factors which should therefore be included in the Five Factor Model (Aston et al 2004).  On the other hand, Larsen & Buss (2002) argued that attractiveness should also be an important feature of the model.  This has led some researchers to devise a seven factor model for personality (Tellegen 1993).  Although such inclusions are a welcome improvement to the Five Factor Model, The Seven Factor Model lacks supporting evidence in comparison with the Five Factor Model at present.

To summarize, The Five Factor Model is a more accurate indicator of individual differences in psychopathology for the following reasons.  Firstly, ambiguity in the factor P is dealt with.  Breaking P down to A and C allows a wider scope of measurement which, in turn, provides us with a better understanding of which factors are contributing to certain types of behaviour.

Secondly, the general adoption of the NEO PI-R, with its differential rating scale, provides a more accurate way of determining individual differences in psychopathology.  Indeed, its accuracy of assessment has been such that the Five Factor Model has been useful in describing pathological personality conditions (Goldberg 1993) such as personality disorder in cases where the NEO PI-R significantly predicted 12 out of 13 personality disorders (Reynolds and Clark 2001).

Such strong evidence is an indication that the Five Factor Model, although incomplete, makes a significant contribution to our general understanding of individual differences in personality and psychopathology.