John Duckitt

The Prediction of Violence

Behaviour prediction is an important applied goal of psychology and the prediction of violent behaviour, in particular, has attracted considerable attention. Although the ability of mental health professionals to predict violence adequately was widely accepted till the late 1960s, a number of important studies then seemed to establish irrefutably the conclusion that clinical assessments of dangerousness, or violence proneness, were hopelessly inaccurate. Renewed attempts to predict violent behaviour, particularly in criminal populations, however, have recently culminated in the development of empirically based actuarial systems, which have shown a dramatically improved capacity to predict violent behaviour. These systems have already begun to have important impacts on parole and institutional classification policies. It is argued that these new systems involve not merely a methodological, but also an important conceptual shift in the enterprise of violence prediction, and that actuarial strategies may have been unjustifiably neglected by psychologists. Some suggestions for the integration of such actuarial approaches with contemporary theoretical developments in personality and social psychology are discussed.

Culture, Class, Personality, and Authoritarianism among White South Africans

Summary Although Adorno et al. related F-scale authoritarianism to childhood socialization, substantial F scale/SES correlations have suggested alternative explanations in terms of SES or such concepts as breadth of perspective. Ray has also argued that the F scale only measures attitudinal authoritarianism and has therefore presented his directiveness scale as a behaviorally valid measure of authoritarian personality. The present study attempted to elucidate some of these issues by examining the prediction of F scale authoritarianism from directiveness scores, culture, class, and other demographic variables in a large community sample of white South Africans. The results indicated that culture (language group) was the best predictor of F-scale scores, SES was only weakly predictive, and directiveness had no predictive utility. A sex effect, with higher F scale scores for men, emerged for Afrikaans-speaking but not English-speaking Ss. These findings seemed to be reasonably compatible with a socialization...

Social support, personality and the prediction of psychological distress: An interactionist approach

Theoretical considerations as well as recent research suggest that the concept of social support, instead of being viewed solely as a stress buffer, should be seen as an important etiological factor in symptom development in its own right. Consequently, the present study set out to examine the influence of six personality factors, derived by a principal component analysis of the 16PF, on the relationship between social support and symptoms of psychological distress in a student sample (N = 139). The results indicated a significant interaction between extraversion and social support; extraverts showed a substantially heightened sensitivity to social support variations. This finding appears to emphasize the usefulness of an interactionist approach that integrates person and context variables for the prediction of psychological distress.

Locus of Control and Racial Prejudice

Persons of internal as opposed to external locus of control orientation have been reported to be less threatened by and more tolerant of persons who are different from them ( 6 ) . On this basis Strickland ( 6 ) suggested that internal scorers should be less prejudiced than external ones. Since this hypothesis has not yet received systematic empirical attention, the relationship between externality and anti-black prejudice was examined among white South Africans. To obtain a short-form locus of control scale, the five forced-choice items with the highest item-to-total correlations in Rotters original analysis ( 5 ) were used as a 10-item, Likert-format scale. This scale, together with a 10-item measure of anti-black prejudice, shown to be valid and reliable in South Africa ( 2 ) , and a short socialdesirability scale ( 1 ) were administered in a nation wide survey supported by the Human Sciences Research Council to a reasonably representative sample of 782 adult white South Africans (76% response rate). Missing values reduced this N to 735 for the analyses. The partial correlation controlling for social desirabilicy, sex, age, education, income, occupation, language group and size of town between prejudice and externalicy was .07 ( d f = 726, nonsignificant). However, when partial correlations were computed separately for English and Afrikaans-speakers the former was small but significant ( r ~ s = .18, p < .005) but not the latter (rm = -.0003). . It has been argued that the relationship between predisposing personality-related variables and manifest prejudice should be attenuated the more the prejudiced beliefs are sanctioned by prevailing socio-cultural norms (4 ) . The existence of such norms among white South Africans i n general ( 4 ) and their particular strength among Afrikaans as opposed to English-speakers ( 3 ) could then account for externality predicting anti-black prejudice significantly though weakly for English but not at all for Afrikaansspeakers.

Psychological Characteristics of Over- and Undercontrolled Violent Offenders

Megargee (1966, 1973) has predicted a number of concomitant social, behavioral, and psychological differences between chronically undercontrolled and chronically overcontrolled violent individuals. Attempts to validate his theory, however, have resulted in seriously inconsistent findings, possibly because of the use of a measure to classify violent subjects that is of seemingly dubious validity--the Overcontrolled Hostility (OH) Scale (Megargee, Cook, & Mendelsohn, 1967). The present study used a different strategy--the classification of subjects by expert raters on the basis of extensive case history information--to compare overcontrolled violent, undercontrolled violent, and nonviolent criminal offenders on a number of psychometric measures (16PF, Hostility and Direction of Hostility Questionnaire, Picture Situation Test, and the OH Scale). The findings were broadly consistent with predictions derived from Megargees theory. Contrary to expectation, however, the nonviolent group did not obtain scores intermediate to the other two groups, but was statistically indistinguishable from the undercontrolled group, leaving a question as to the appropriateness of conceptualizing the typology in terms of over- and undercontrol of hostility and suggesting the necessity for some revision of the theory.

Directiveness and Authoritarianism: Some Research Findings and a Critical Reappraisal

Resulting from a cogent critique of the F scale as a measure of personality, Ray (1976) has proposed his ‘directiveness’ scale as a behaviourally valid measure of authoritarian personality. Research using this scale has suggested that authoritarian personality and the endorsement of authoritarian attitudes or ideology may be quite unrelated. This was supported by the present study which found no relationship between directiveness and either intolerant and negative attitudes towards homosexuals or an F scale measure of authoritarian attitudes. The conceptualizations of authoritarianism of Ray and Adorno et al. (1950) were then briefly reviewed and it was suggested that they appear to be quite distinct. It was therefore concluded that research using Rays directiveness scale does not invalidate the theory of the authoritarian personality and that the development of a behaviourally valid measure of authoritarian personality as conceptualized by Adorno et al. remains a major research priority.

Personality Factors as Moderators of the Psychological Impact of Life Stress

Six personality dimensions, derived by factor analysis from the 16 PF, were examined as possible moderators of the impact of recent life changes on psychological strain in a student sample. The results indicated a significant interaction effect for only one of these dimensions, extraversion, with high scorers (extraverts) appearing to be significantly more tolerant of recent life change than low scorers (introverts). This finding, which appears to be consistent with reports from other retrospective studies indicating that measures of arousal-seeking status moderate the psychological impact of life stress, is discussed in terms of Eysencks theory of extraversion.

Prejudice and Neurotic Symptomatology Among White South Africans

Van der Spuy and Shamley (1978) have assembled evidence suggesting chronically elevated levels of neurotic symptomatology among both white and black South Africans. They have argued that these elevated levels could be attributed to the experience of racial discrimination and prejudice. New data obtained from a reasonably representative national sample of white South Africans (N = 782) did reveal symptom scores substantially and significantly higher than scores previously reported for a large community sample in the United States. On the other hand, partial correlational analyses did not indicate any consistent pattern of association between antiblack prejudice and symptom scores among white South Africans. Therefore, intergroup conflict in a sharply stratified society may affect symptomatic levels in individuals by influencing the general quality of social life.

Reply to Ray's Directiveness and Authoritarianism: A Rejoinder to Duckitt

Rays contention that his Directiveness (DR) scale measures one of the component constructs of the syndrome of authoritarianism described by Adorno et al., is questioned. Neither his definition of DR nor the item content of that scale suggests any real correspondence with anyone of the nine components described by the theory of the authoritarian personality (TAP). It is also demonstrated that Rays assertion that the TAP requires the covariation of submissiveness and hostility as generalized behavioural dispositions is mistaken and represents a serious misinterpretation of Adornos concepts of authoritarian submission and authoritarian aggression. Essentially it is argued that in order to test a theorys linkage of two concepts, such as those of authoritarian submission and authoritarian aggression, adequately, their measurement should conform to the manner in which that theory described them. Since Rays concept of DR has very little, if anything, in common with traditional conceptualizations of authoritarianism such as those of Reich, Fromm, Maslow and Adorno, it is suggested that the investigation of correlation between the DR scale and concepts pertinent to the TAP has no theoretical rationale at all and as such becomes an essentially pointless exercise.

Personality Profiles of Homosexual Men and Women

There has been surprisingly little research on possible differences between homosexual and heterosexual personalities, although the few studies that have been conducted suggest some interesting differences. We devised two hypothesis that may account for these differences. First, differences appear to reflect generalized social nonconformity and alienation due to the social stigmatization of the homosexual. Second, these differences seem to express a more specific tendency to deviation from socially normative sex roles. The two hypotheses imply a number of predictions concerning specific personality traits, which were tested by comparing the 16PF (Cattell, Eber, & Tatsuoka, 1970) personality profiles of a group of male (N = 34) and female (N = 31) homosexuals with those obtained from the two large student samples (male, N = 899 and female, N = 912) whose scores are commonly used as norms for the South African version of the 16PF. The findings suggested reasonably good support for the two hypotheses, particularly considering the limitations of the study with respect to the adequacy of the comparison groups used and the relative heterogeneity of the 16PF scale content.