A quasi-experimental comparative cross-sectional study to compare the disclosure rates of sensitive behaviours of University of KwaZulu-Natal students.

Thesis (M.A.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.Firstly, this research aimed to understand what behaviours are considered sensitive or private by university students (N=306) in respect of disclosure in the research context. A total of 71 items were extracted by factor analysis: 20 sensitive items, 26 related non-sensitive items and 25 non-related non-sensitive items. Differences in sensitivity were noted for gender and race, reported below. Furthermore, a three-phase exploration of data collection methods was used in determining which self-report method is most valid and reliable when researching sensitive topics. A quantitative experiment compared the effectiveness of the Unmatched Count Technique (Type I), Self-Report Questionnaires and Audio Computer-Assisted Self-Interviews, in terms of their ability to elicit honest answers when dealing with the sensitive topics (N=410). This section of the study used pairwise tests of proportions by Winks statistical software. The sensitive topics under investigation in this study are condom use,HIV/AIDS as well as relationships such as transactional and multiple and concurrent partners.The results of this study, reported below, indicate pairwise significant differences between the SRQ, ACASI and UCT Type I. Additionally, the Unmatched Count Technique (Type I), Self-Report Questionnaires and Audio Computer-Assisted Self-Interviews were compared in terms of Socially Desirable Responding scores as well as experience of participation. No statistically significant differences were obtained for overall scores across data collection methods for SDR and experience of participation