This dissertation examines the nature of personality in terms of the temperament and character framework identified by Cloninger and colleagues (1 993). This model proposes that personality can be understood in terms of temperament, unlearnt, essentially biological traits, and character, socioculturally learnt traits that reflect the individual ' s goals and values. The nature of these higher order dimensions of personality i s such that temperament motivates action, while character, based on the salience of the situation, intervenes to produce behavioural responses that best suit the individual ' s goals and values (Cloninger et. al., 1993). Two studies were used to test this theory. The first study hypothesised that this model could be used to explain cross-situational variability that has been demonstrated to exist in self-reported personality data. To test this, 1 7 3 first year psychology students completed the Temperament and Character Inventory (TCI) , the Big Five Inventory (BFI), the BIS/BAS Scales (BB S ) , the Appetitive Motivation Scale (AMS) and the Dickman Impulsivity Inventory (DII) indicating whether the items reflected temperament or character, and how true each item was for them across the four situations of: in general, at home, at work and in a perfect world. The results failed to support the hypotheses. Firstly, there were very few significant differences in the number of participants who were able to conceptualise their own behaviour in line with the theory. Secondly, scale level analysis revealed no differences between the temperament or character traits across the situations, suggesting they are underpinned by the same mechanisms . Finally, analysis of the within participants average item standard deviations revealed greater mean standard deviations for the temperament traits than the character traits , with the exception of the scales of the DII. These results were interpreted as an indication that Cloninger et. al.'s (1993) theory of temperament and character possesses several flaws particularly in relation to the operation of character, as no evidence was found for it modifying behaviour based on the salience of the situation and the individual ' s goals and values.
To explore the nature of character further, assessment of the theorised mediation of temperament by character in the prediction of specific behaviour was undertaken . To achieve this, study two employed structural equation modelling of personality inventory data in the assessment of self-reported delinquency. All of the theories analysed in study 1 were again applied to study 2. In addition to these, Eysenck' s theory of personality ( 1 964 : 1 996), including the primary scales identified by Jackson and Lawty-Jones ( 1 996), and emotional control were also analysed. For study two, 210 first year psychology students completed the personality measures and a self-reported delinquency measure. It was hypothesised that character and anxiety would significantly mediate temperament, particularly impulsivity, in the prediction of self-reported delinquency. The results partially supported the hypotheses. Statistically significant models for each theory, with either partial or total mediation of temperament by character, were found. However, only two of the models demonstrated mediation by anxiety in the prediction of self-reported delinquency. Furthermore, impulsivity was found to maintain a direct relationship with delinquency even when character and anxiety were present, suggesting that an aspect of this trait cannot be restrained by character or anxiety. The results were discussed in terms a new understanding of the temperament and character model of personality, and by offering suggestions for clinical interventions for delinquency. The limitations of the present studies and directions for future research are also discussed.