Abstract
M.A. (Psychology)
The present study was conducted to investigate the relationship between personality and risk taking
behaviour in the South African context. Personality was measured with the Basic
Traits Inventory (BTl), an assessment specifically developed to measure the broad
dimensions of the five factor model of personality (John & Srivastava, 1999) in South Africa.
The five dimensions on the BTl have the same names as the well-known five factor model,
namely: Neuroticism, Extraversion, Openness to Experience, Agreeableness, and
Conscientiousness. The primary objective of this study was to investigate the underlying
personality structure across ten different forms of risk-related behaviour. The risk behaviours
included smoking, alcohol consumption,.illegal drug use, sexual promiscuity, thrill-seeking
activities, gambling, physical violence, romantic infidelity and other behaviours that may
have led to a respondent being arrested. Given South Africa's unique population, a further
objective ofthis study was to examine the degree to which the results from the study would
be in line with those reported in so-called Western, Educated, Industrialised, Rich and
Democratic (WEIRD) societies.
The sample consisted of 683 respondents, all second-year students from a bilingual
(Afrikaans and English) university in Johannesburg. There were 142 men and 538 women in
the sample. Three of the respondents' gender was unknown. There were 425 White
respondents, 120 Black respondents, 83 Indian respondents, 46 Coloured respondents and
nine respondents who did not specify any population group. Respondents' mean age was
20.99 years with a standard deviation of5.10 years. The sample was not representative ofthe
South African population, with men being underrepresented and White respondents
overrepresented in comparison to other population groups. A multivariate technique,
Descriptive Discriminant Analysis, was used to analyse personality differences across groups.
The groups were formed based on the frequency with which individuals engaged in the different risk-behaviours. Post-hocanalyses allowed for a close rexamination of group
differences.
The results revealed that a single, statistically significant discriminant functionemergedfor
all ten of the risk variables with the exceptionof one, for whichtwo possible discriminant
functions were identified. This showed that different combinations of the five personality
factors were, to some extent, able to account for group separation on each of the risk
variables. Considering the results as a whole, some interesting findings were revealed: It
became evident that no single personality structureexists across the different risk-variables of
this study. It was clear that some personality factors were more important, whereas others
were less important, depending on the type of risk-behaviour being considered. Despite these
seeming differences, important patterns of personality emergedacross the risk-variables.
Conscientiousness, and in particular, Extraversion were identified as the most salient
predictors of the risk-behaviours in this study, although important contributions were also
made by the remaining personality factors: Conscientiousness was further found to be the
most important predictor of health-risk behaviours such as smoking, alcohol consumption,
and druguse. In general, Opennessto Experience, Agreeableness, and Neuroticism appeared
to be more selectively associated with specific risk-behaviours when compared to
Extraversion and Conscientiousness.
Overall, the findings reported in this study were largely in line with those reported in so called WEIRD countries.
The results of this study further supported the generalisability of
prior research regarding the relationship between personality and risk-taking. It also
demonstrated the utility of the five factor model as a promising predictor of risky behaviour.
For future research it is recommended that the facet-scale level of the BTl be used to further
investigate the personality-risk relationship.