In the last thirty years there have been many research studies which have examined the reasons why adolescents and young adults engage in sexual behaviours. Most of these studies have lacked a theoretical basis, Consequently there are many links made between variables, but no consistency across studies, or attempts to develop an underlying theory to explain the results. However, there have been theoretical models developed to explain adolescents’ and young adults’ sexual decision making. Unfortunately, many of these models have not been empirically validated. This thesis attempts to address these deficiencies in the literature by utilising a theory of behaviour and applying it to adolescent and young adult sexual decision making. This theory is the Theory of Planned Behaviour (TPB). Two longitudinal studies were conducted to examine the utility of an adaptation of the TPB to sexual decision making among adolescents and young adults. In the first study 58 adolescent males, aged between 14 years and 18 years participated in a longitudinal study using a questionnaire adapted from the Depth of Sexual Involvement Scale. In the second study, 194 young adults (156 female, 38 male) aged between 18 years and 21 years participated in a similar study. The first study found that intention to engage in behaviour was well predicted, although some of the variables in the model, did not in fact, contribute significantly to the prediction. The prediction of behaviour was less strong than that of intention. Study two found that intention to engage in behaviours was well predicted by the model. However, the degree to which intention led to behaviour was not well predicted. Overall, the results of these studies suggest that the TPB is a good theoretical basis from which to launch a systematic and theoretically informed explanation of adolescent and young adult sexual decision making. However, other factors may need to be added to the model to fully describe the decision making process and accurately predict behaviour. Suggestions are made for future research, as well as interventions that may arise as more knowledge is gathered using this paradigm.

Using lessor's psychosocial framework of risk behaviour, this study examined the impact of variables within a number of risk domains on condom use at last sex. The sample was 587 sexually active male and female black youth between the ages of 16 and 20 years old from the township areas of Soweto (Johannesburg), Umlazi (Durban) and Khayelitsha (Cape Town). Binary logistic regression models were used in the analysis. Within the biological risk domain, gender was a significant predictor of condom use at last sex. None of the proximate social context variables, viz. parental education, family structure, and parent-adolescent communication, were significant predictors. Health services' promotion of condom use was a significant predictor within the distal social context. Within the perceived environment in relation to the self, perception of risk and perception of barriers were significant predictors of condom use at last sex. None of the variables within the perceived environment in relation to peers domain, viz. perceived peer attitudes to condoms and peers suffering the negative consequences of unprotected sex, were significant. The personality domain contained measures of self-esteem, future time orientation, locus of control and fatalism. Locus of control was the only significant predictor within this domain of condom use at last. Engagement in other risk taking behaviour, such as smoking cigarettes, alcohol and drug use, was a significant predictor within the general behaviour domain. Within the sexual behaviour domain, partner discussion and contraceptive use were significant predictors of condom use at last sex. 2 In the final model, the significant predictors in order of importance, were the perception of risk, the promotion of condom use by a health professional, locus of control, discussion with a partner, the perception of barriers, the use of contraceptives, and risk taking behaviour. The only interaction in the final model of condom use at last sex was between the promotion of condom use by a health professional and concurrent use of other forms of non-barrier contraception. / Thesis (M.A.)- University of Natal, Pietermaritzburg, 2001.