Transitions to Adulthood in the Context of AIDS 1999-2002
Health Survey [hs]
The Transitions study is conducted by the School of Population and Poverty Studies at the University of Natal, Durban, the Horizons Project, the Policy Research Division of the Population Council, and Focus on Young Adults (FOCUS), and the MEASURE/Evaluation Project of Tulane University. The research is a prospective study of reproductive behavior and sexual health of adolescents in South Africa as well as their education and employment experiences, family and environmental conditions, and other factors in their lives that may influence their sexual behavior and choices.
The study design includes two rounds of data collection from adolescents (ages 14-22), in KwaZulu-Natal (KZN), South Africa, during 1999 and 2001-2002.
Additional data was collected at baseline and follow-up from all schools in the study area regarding the teaching of a Life Skills Programme in those schools. This programme, introduced initially in secondary schools, was a key strategy in the state's response to the HIV/AIDS epidemic in South Africa.
The survey data is complemented by data on communities (collected in May and June 2000) and an exploration of some of the principal results from the survey data based on focus groups and other qualitative approaches (carried out in August and September 2000).
Kind of Data
Longitudinal Survey [ls]
Unit of Analysis
v1: Edited, anonymised dataset, for public distribution
This interviews covered the background characteristics of the respondents, education history, work experience, a diary of their activities in the previous 24 hours, exposure to the school-based Life Skills Programme, sexual relationships, contraceptive and condom knowledge, attitudes and use, connectedness to school, family and community, alcohol and drug use, and reproductive history.
Because the South Africa life skills initiative aimed to be a full coverage program, it was not possible to use a randomized control design to measure program impact. This is because schools initiated or developed life skills education at different speeds and intensities. The study therefore measured young peoples recall of 11 topics from their school training, and assessed the differences in sexual and reproductive health knowledge and behaviors, given that these young people were differentially exposed to the topics. This is called measuring the dose-response relationship between indicators of exposure to education and sexual and reproductive health knowledge and behaviors. All the multivariate models control for unmeasured factors using a variable for survey year. A significant effect for this survey year variable was found for many of the models and may be interpreted as reflecting the combined effects of maturation of youth, exposure to other sources of information other than life skills education, and general secular trends. Exposure to life skills education was measured by changes in recall of 11 key sexual-reproductive health topics based upon student report. Outcomes included: 1) sexual abstinence to either postpone first sex (delay initiation), or for those already initiated to reduce risk through secondary abstinence; 2) reduced number of sex partners in the past 12 months; and 3) changes in condom use (i.e. condom use at first sex; condom use at last sex and consistency of condom use with all partners). Separate analyses were conducted by sex, race, and age.
The survey was carried out in the Durban Metropolitan and Mtunzini Magisterial Districts of Kwazulu-Natal, South Africa.
The lowest level of geographic aggregation in the study in enumerator area.
The study covered adolescents (ages 14-22), in selected households in the Durban Metropolitan and Mtunzini Magisterial Districts of KwaZulu-Natal (KZN), South Africa.
Producers and sponsors
School of Population and Poverty Studies
University of Natal, Durban
Policy Research Division, Population Council
Focus on Young Adults
Tulane University, New Orleans, Louisiana, United States
Two administrative areas within the province - the Durban Metropolitan and Mtunzini Magisterial Districts - wereselected within KwaZulu-Natal for the study.These administrative areas were selected to ensure a variety of urban, transitional and rural regions within the province. Urban respondents (77 percent of the sample) were taken from the Durban Metro sample as well as those living in urban areas within the Mtunzini Magisterial District. Rural respondents (23 percent) were from the rural areas of Mtunzini.
The study used a modified multi-stage cluster sample approach. The first stage required the random selection of 120 enumeration areas (EAs) from a sampling frame of all EAs in the two districts. At the second stage, field supervisors divided EAs into approximately equal segments of a predetermined size (based on an estimate for the average number of adolescents expected per household, derived from census data). The study team then selected one segment randomly, and interviewers tried (in up to three visits) to find all households within that segment and interview every young person between the ages of 14 and 22 reported to live in those households.
For the study, two rounds of surveys of households and youth were undertaken, in 1999 and 2001-2002. Within each area, all youth 14-22 years of age residing in a segmented, probability sample of Census Enumeration Areas (CEAs) were interviewed in the Wave 1 survey (1999). In Wave 2 (2001), all youth 14-24 years of age residing in the same CEAs were included in the survey, including 2,222 of the 3,052 youth also interviewed in Wave 1.
The individual and household survey data were complemented by surveys with school principals undertaken in 1999 and 2001 and a community survey of the areas undertaken in mid-2000.
Interviewers completed interviews with 82.2 percent of the adolescents identified in the selected households. However, response rates varied by population group. Interviewers successfully completed interviews with 90.9 percent of eligible adolescents among Africans in rural areas, 83.6 percent among Africans in urban areas, 69.6 percent among Asians, and 67.5 percent among Whites (the latter two groups were only selected in urban areas). These differences arise from the different patterns of activities among population groups, which keep some youth more than others away from home.
Dates of Data Collection
Community data from intercept questionnaire
Data Collection Mode
The questionnaire included questions about household members, living conditions, economic shocks, expenditure, government assistance, and discussions about HIV in the household.
Data entry and cleaning were done by Policy and Praxis, an independent South African data management organization
University of Natal, Population Council & Tulane University. Transitions to Adulthood in the Context of AIDS in South Africa 1999-2002, Waves 1-2. [dataset]. Version 1. Durban: School of Development Studies, University of Natal, Population Council & Tulane University [producer], 2004. Cape Town: DataFirst [distributor], 2014. DOI: https://doi.org/10.25828/6t07-da67