The five factor theory from a social cognitive perspective

Despite many criticisms (Block, 1995) the Five Factor Theory (FFT) is still one of the most researched models of personality. The theory is characterized by the following important attributes of the five factors. Firstly, they are dimensions on which people vary continuously. These factors are stable from young adulthood to old age (Pervin & John, 1990). The factors are at least in part heritable, and it is likely that they had an adaptive significance in a prehistoric environment (Buss, 1996).

Also the factors are proposed to be universal, and knowledge of an individuals position on the five factors is seen as advantageous in designing therapeutic measures for clinical improvements (McCrae & Costa, 1999). This contrasts with the position of Social-Cognitive Theory (SCT) which focuses not on heredity and stability, but instead emphasizes the dynamic and flexible use of multiple cognitive structures. Social-Cognitive theorists point out the importance of using schema tasks, and strategies adaptively (Pervin & John, 1999).

SCT calls for a broader conceptualization of personality to the one proposed in the FFT (Shoda, Mischel & Wright, 1993). They propose that people be defined not by an uncomplicated combination of universals but by a unique cognitive, emotional and behavioural signature. In this essay, an overview of the arguments and criticisms against the validity of the FFT will be presented from the perspective of SCT. The main criticisms which will be briefly summarised focus on the difficulties and disadvantages involved in invoking endogenous basic tendencies in the form of traits as the cause of behaviour.

The elements of FFT most pertinent to this examination are: the limitations of the method of factor analysis as a tool for identifying trait dimensions, the problems which result from the elimination of psychological language and insight in the FFT’s development, the paucity of using single-word adjectives in attempting to describe the variation of personality characteristics over situations, the theory’s lack of explanatory, operative and predictive power and its apparent circularity.

The limitations of the Five Factor Theory and the important areas it neglects will be discussed in light of the broader picture SCT provides in regard to these areas. Some important conceptual differences from which the opposing viewpoints arise will be discussed. These include SCT’s rejection of the dualistic division of theories into either structural or functional as proposed by FFT, and also the view in the FFT of the ‘Big Five’ as ‘causal personality dispositions’ (McCrae & Costa, 1999).

Resolution of these issues of alternative assessment, if possible, may hasten the progress towards an all encompassing theory of personality. The founder of SCT and critic of the FFT Albert Bandura (1999) draws attention to what he believes is a ‘missionary zeal’ on the part of McCrae & Costa (1999) in regard to the FFM. He points out that their claims of consensuality appear to be exaggerated. Statements of their belief that what they have ‘discovered’ in the FFT is ‘an empirical fact’ seem quite unscientific and do little to support the validity of their research.

Dawda (1997) blames the misconstrual of the technique of Factor Analysis (FA) as objective, atheoretical and purely empirical for such claims. There are several ways in which FA is far from objective, the most obvious being the decision as to how many factors best represent the data set. Eysenck (1991) points out the danger of mixing up first and second order factors. He proposes that there are only 3 second order factors which are, not surprisingly, the three of his PEN model. He claims that the three Big Five factors not associated with his PEN model are first order factors.

He does not, however, propose any guidelines whereby it is possible to objectively distinguish first and second order factors. From his study (Eysenck, 1991) it appears that the only determinant of which model emerges is the use of either the Kaiser-Guttman criteria (eigenvalues over 1) or the visual inspection Scree test. The FFT emerges from the former and the PEN model from the later. Jackson et al (1995) found that their data set was best represented by six factors and proposed that different more comprehensive sets of variables would probably yield more factors.

They suggest that researchers should focus on the utility of personality factors, not their salience in factor space. These studies draw attention to the arbitrary nature of factor selection in general. Other studies (Dawda, 1997) have investigated the empirical history of the FFT and found many short-comings. Specifically, in conclusion to his analysis of the emergence of the FFM, Dawda (1997) states that ‘a series of inadequately elaborated decisions, motivated by the need to achieve reliable, orthogonal factor structure, guided the selection of items’.

Such strong claims of prestructuring call into question the apparent scientific nature of the FFT and its supposed validity. Bandura (1999) reiterates these criticisms pointing out that the Big Five often fail to emerge from formal goodness-of-fit tests, that prepruning and extraction method defines what is found from FA. He also points out that large intercorrelations of the five factors oppose claims of their distinctiveness. Probably the greatest limitation of the FFT is its dependence on the uncomplicated relations between adjective descriptors (Dawda, 1997).

The positioning of individuals on the dimensions resulting from the analysis of the adjectival relations and the subsequent investigation into how this positioning correlates with behaviour is seen as by social-cognitive theorists as a deficient notion of personality (Shoda et al, 1993). This contrasts with SCT’s proposed study of how ‘personal factors operate in causal structures in producing and regulating behaviour under the highly contingent conditions of everyday life’ (Bandura, 1999).

On a similar note Mischel (1968) emphasises the variability of personality characteristics over situations, his results providing a strong disparity with FFT’s claims of stability. Other criticisms levied upon the FFT by the proponents of SCT include calls for a fuller incorporation of the interpersonal and intraindividual context, for the individuals developmental history, personal goals, values and motivations to be taken into account (Pervin & John, 1999).

This call for personal determinants over static entities has been answered, albeit inadequately by FFT theorists (Bandura, 1999). They recognise the interaction of social and physical environments with personal dispositions and that individuals selectively influence the environment. However, the details of how this and the ‘universal dynamics’, (the mechanisms which regulate thought, feelings and behaviours) operate are unspecified. It seems the proponents of the FFT are going to leave it to other ‘sub-theories’ to fill in the blanks/specifics in their ‘grand theory’.

These ‘advances’ in the FFT appear to incorporate some of the ideas of SCT, but these additions are constrained by the boundary conditions for individual and situational effects which the dimensions derived from the lexical hypothesis sets upon them. This concept is analogous to the dependent nature of software in the first computers on the punched card bootstrapping programme which had to be loaded before the software could be read.

In a similar way the recent appendages to the FFT are restricted in their application by the ‘decontextualised conglomerate’ of traits on which they are based, and must work through (Dawda, 1997). A debate as to the fundamental meaning of the lexical hypothesis and the subsequently derived factors is central to the analysis of the validity of the FFT. Saucier & Goldberg (1996) claim that studies of personality description are not necessarily concerned with issues of causality.

Lexical studies are thus portrayed as involving important psychological phenomena studied as a descriptive taxonomy (John & Srivastava, 1999). Saucier et al (1996) believe that the concepts central to our cultural, social and personal life experiences are encoded in language and this makes trait dimensions worth studying. This non-causal view suffers from problems of circularity as behaviour in the form of causal dynamic factors is invoked as the cause of behaviour. However, McCrae & Costa (1999), the main proponents of the FFT, view the Big Five not as descriptive, but as causal personality dispositions.

By virtue of this property the problems of circularity aforementioned are avoided. This allows McCrae et al (1999) to put forward an explanatory interpretation of the taxonomy backed up with empirical research on the genetic basis and biological structures on which the Big Five are based. From this perspective the concepts of SCT such as attitudes, motivations and goals are simply outcomes of the interaction between endogenous basic tendencies and the environment (John et al, 1999).

Much of SCT would therefore be classed as characteristic adaptations or dynamic processes, working from the Big Five dimensions, within the FFT. Bandura (1999) argues that this is not the case as FFT is flawed as it endorses a dualistic view which asks if a theory or an aspect of a theory addresses ‘functionalism or structuralism? ‘. He calls this ‘false separateness’ and insists that ‘regulatory processes operate through guiding self-structures rather than disembodied from them’.

There is a conceptual difference here which must be addressed. Bandura (1999) advocates an interdependency of nature and nurture, a potentialist view where experience is produced by regulatory processes, acts in the environment and by this measure shapes self structures. McCrae & Costa (1999) on the other hand propose the more deterministic view that basic tendencies remain stable. Characteristic adaptations by direct or indirect routes are involved in reciprocal causation with themselves, with the environment and with behaviour.

This seems quite similar to the triadic reciprocal causation as proposed by SCT, but FFT does differ in that there is one way causation between biological bases? basic tendencies? characteristic adaptations (containing the self-concept). This lack of reciprocity of change between processes and structures is the fundamental point on which SCT and FFT diverge, and it seems that the problem of deciding which view is more accurate will not be solved by armchair speculation, but by the results of empirical findings (John et al, 1999).

Bandura (1999) claims that global traits tell us little about the nature of personal causation as the determinants of our behaviour operate ‘conditionally at a particular contextualised level, not at a socially detached conglomerate level’. The question as to the extent of the Big Five’s contribution to behaviour in context can best be resolved by investigating genetic and biological structures and processes which may be the underlying mechanisms of these traits. Evidence of the existence of these mechanisms would give the FFT a firmer foothold on which to build its theory.

It appears that this resolution will not occur any time soon as relating traits and behaviours causally to specific biological mechanisms is a very difficult task. At present the large majority of research attempting to relate traits to biological structures is of a correlative rather than a causal nature (John, 1999). It is worth noting that the consensus regarding the Big Five which has arisen of late has only been in regard to the description of individual differences, and has more or less ignored an empirical investigation into the structures from which the main endogenous traits arise.

The research which has been carried out into the existence of such biological mechanism is largely confirmative rather than exploratory in nature (Buss, 1996). The main problem which arises from this approach is the possibility that the five proposed supertraits reflect not the ultimate most parsimonious, replicable and useful set of traits but a premature and overstated social consensus amongst trait theorists (Bandura, 1999).

This method of descriptive taxonomy contrasts with that of the causal theorists who have focused their studies on biological mechanisms and related these (i. emotional reactivity) with traits of individual differences. Fahrenburg (1991) criticizes trait theorists’ acceptance of simple relationships between self report dimensions and underlying biological mechanisms. Their criticism focuses on the complicated nature of the parameters which are necessary to find such results, and the way in which this makes theories overly complicated. He also found that based upon his empirical investigations, that there was little evidence for proposed mechanisms.

Social-Cognitive theorists (Bandura, 1999; Mischel, 1973) insist that FFT, even if backed up by evidence, by its nature only skims ‘the surface of the wetlands of personality’. They propose that the methods used in the study of personality should reflect its complexity. They seek to go beyond the ‘narrow’ factor analytically developed view of personality as proposed by FFT, and to include empirically verified theory and insight in the development of a broader perspective on personality.

Bandura (1999) claims it is an issue which focuses on whether ‘personologists or machines do the conceptualizing’. Similarly, Block (1995) advised that an adequate theory of personality should focus on the terminology of personality theorists not on clusters of adjectives. McAdams (1992) argues that this is the only way we can move beyond the ‘psychology of the stranger’, which is the limited snap-shot of personality the FFT gives us. Bandura (1999) argues that the human mind is ‘generative, creative, proactive and self-reflective’.

He proposes that we are characterised by ‘discriminative forethought’ and are not simply ‘reactors’ to the ‘genetic blueprint for personality’, the view which Costa & McCrae (1996) advocate. On this note Dawda (1997) claims that the inadequate account of the dynamic organisation of personality characteristics and the inability of the FFT to comprehensively address cross situational variations in behaviour reduces the utility of the model for use in a clinical setting where assessment must incorporate both idiographic and dynamic aspects of personality function.

FFT also lacks clinical validity in that if fails to provide a guide as to how to effect personal and social change, which contrasts with the comprehensive guidelines which furnish SCT (e. g guidelines on how to build a strong sense of self-efficacy). SCT theorists thus would advise that the FFT should focus more on the dynamic nature of personality and less on their search for how to ‘extract consistency from variability’, and less on their efforts to explain how global dispositions bring about ‘highly variable conduct’ (Pervin & John, 1999).

Bandura (1999) believes that we can’t expect personality measures which are set in ‘non-conditional generalities’ to aid in our understanding of the ‘contribution of personal factors to psychosocial function’ across many different tasks, circumstances and situations. He instead proposes that researchers study dispositions such as self-beliefs, aspirations and outcome expectancies, as these and not descriptors of habitual behaviour regulate what we do. Bandura (1999) recommends that Five Factor theorists include such descriptors of the ‘sociocognitive belief system’ in their studies rather than the adjectival trait dimensions aforementioned.

It is clear that FFT has a long way to go both methodologically and conceptually before it can become the ‘grand theory’ its proponents ambitiously propose it to be. Issues as to the replicability, operative utility and predictive power of the five factors need to be addressed. The presence of subjectivity in factor analysis, the presumptions of the lexical hypothesis and the psychological implications of describing personality in terms of single word adjectives, must be investigated comprehensively and displayed openly where present.

For instance the extent to which the traits which arise from factor analysis are defined by how we speak needs to be established. Bandura (1999) points out that the possibility of a cultural basis cannot be ignored. He insists that cross-cultural studies are at present far from refuting the hypothesis that the Big Five are at least in part a culturally bound phenomenon. This point indicates the possible errors which may arise from using traits derived from the lexical hypothesis as indicators not simply of how we speak, but of universal endogenous dispositions (Saucier & Goldberg, 1996).

The FFT must also face the problems which arise from using a broad taxonomy of compounded traits to predict behaviour in many situations. This and the aforementioned issues are ones which FFT must address in order for it to develop as a valid, reliable theory. However, some (Bandura, 1999) would insist that for a theory of personality to recognize the vast complexity of the human character, it must do so using complex methods, rather than attempting to analyse it on five finite dimensions.